v1.5.0

never been able to reproduce it intentionally but this should work

.eslintrc.json

@@ -0,0 +1,12 @@
+{
+    "env": {
+        "browser": true,
+        "es2021": true
+    },
+    "extends": "eslint:recommended",
+    "parserOptions": {
+        "ecmaVersion": 12
+    },
+    "rules": {
+    }
+}

.gitattributes

@@ -1,4 +1,6 @@
 * text eol=lf
 
+*.reg text eol=crlf
+
 *.png binary
 *.gif binary

.github/ISSUE_TEMPLATE/bug_report.md

@@ -0,0 +1,40 @@
+---
+name: Bug report
+about: Create a report to help us improve
+title: ''
+labels: bug
+assignees: '9001'
+
+---
+
+NOTE:
+all of the below are optional, consider them as inspiration, delete and rewrite at will, thx md
+
+
+**Describe the bug**
+a description of what the bug is
+
+**To Reproduce**
+List of steps to reproduce the issue, or, if it's hard to reproduce, then at least a detailed explanation of what you did to run into it
+
+**Expected behavior**
+a description of what you expected to happen
+
+**Screenshots**
+if applicable, add screenshots to help explain your problem, such as the kickass crashpage :^)
+
+**Server details**
+if the issue is possibly on the server-side, then mention some of the following:
+* server OS / version: 
+* python version: 
+* copyparty arguments: 
+* filesystem (`lsblk -f` on linux): 
+
+**Client details**
+if the issue is possibly on the client-side, then mention some of the following:
+* the device type and model: 
+* OS version: 
+* browser version: 
+
+**Additional context**
+any other context about the problem here

.github/ISSUE_TEMPLATE/feature_request.md

@@ -0,0 +1,22 @@
+---
+name: Feature request
+about: Suggest an idea for this project
+title: ''
+labels: enhancement
+assignees: '9001'
+
+---
+
+all of the below are optional, consider them as inspiration, delete and rewrite at will
+
+**is your feature request related to a problem? Please describe.**
+a description of what the problem is, for example, `I'm always frustrated when [...]` or `Why is it not possible to [...]`
+
+**Describe the idea / solution you'd like**
+a description of what you want to happen
+
+**Describe any alternatives you've considered**
+a description of any alternative solutions or features you've considered
+
+**Additional context**
+add any other context or screenshots about the feature request here

.github/ISSUE_TEMPLATE/something-else.md

@@ -0,0 +1,10 @@
+---
+name: Something else
+about: "┐(ﾟ∀ﾟ)┌"
+title: ''
+labels: ''
+assignees: ''
+
+---
+
+

.github/branch-rename.md

@@ -0,0 +1,7 @@
+modernize your local checkout of the repo like so,
+```sh
+git branch -m master hovudstraum
+git fetch origin
+git branch -u origin/hovudstraum hovudstraum
+git remote set-head origin -a
+```

.gitignore

@@ -5,12 +5,16 @@ __pycache__/
 MANIFEST.in
 MANIFEST
 copyparty.egg-info/
-buildenv/
-build/
-dist/
-sfx/
 .venv/
 
+/buildenv/
+/build/
+/dist/
+/py2/
+/sfx*
+/unt/
+/log/
+
 # ide
 *.sublime-workspace
 
@@ -18,5 +22,10 @@ sfx/
 *.bak
 
 # derived
+copyparty/res/COPYING.txt
 copyparty/web/deps/
 srv/
+
+# state/logs
+up.*.txt
+.hist/

.vscode/launch.json

@@ -10,12 +10,24 @@
             "cwd": "${workspaceFolder}",
             "args": [
                 //"-nw",
-                "-a",
-                "ed:wark",
-                "-v",
-                "srv::r:aed"
+                "-ed",
+                "-emp",
+                "-e2dsa",
+                "-e2ts",
+                "-mtp",
+                ".bpm=f,bin/mtag/audio-bpm.py",
+                "-aed:wark",
+                "-vsrv::r:rw,ed:c,dupe",
+                "-vdist:dist:r"
             ]
         },
+        {
+            "name": "No debug",
+            "preLaunchTask": "no_dbg",
+            "type": "python",
+            //"request": "attach", "port": 42069
+            // fork: nc -l 42069 </dev/null
+        },
         {
             "name": "Run active unit test",
             "type": "python",
@@ -28,5 +40,13 @@
                 "${file}"
             ]
         },
+        {
+            "name": "Python: Current File",
+            "type": "python",
+            "request": "launch",
+            "program": "${file}",
+            "console": "integratedTerminal",
+            "justMyCode": false
+        },
     ]
 }

.vscode/launch.py

@@ -0,0 +1,49 @@
+#!/usr/bin/env python3
+
+# takes arguments from launch.json
+# is used by no_dbg in tasks.json
+# launches 10x faster than mspython debugpy
+# and is stoppable with ^C
+
+import re
+import os
+import sys
+
+print(sys.executable)
+
+import json5
+import shlex
+import subprocess as sp
+
+
+with open(".vscode/launch.json", "r", encoding="utf-8") as f:
+    tj = f.read()
+
+oj = json5.loads(tj)
+argv = oj["configurations"][0]["args"]
+
+try:
+    sargv = " ".join([shlex.quote(x) for x in argv])
+    print(sys.executable + " -m copyparty " + sargv + "\n")
+except:
+    pass
+
+argv = [os.path.expanduser(x) if x.startswith("~") else x for x in argv]
+
+argv += sys.argv[1:]
+
+if re.search(" -j ?[0-9]", " ".join(argv)):
+    argv = [sys.executable, "-m", "copyparty"] + argv
+    sp.check_call(argv)
+else:
+    sys.path.insert(0, os.getcwd())
+    from copyparty.__main__ import main as copyparty
+
+    try:
+        copyparty(["a"] + argv)
+    except SystemExit as ex:
+        if ex.code:
+            raise
+
+print("\n\033[32mokke\033[0m")
+sys.exit(1)

.vscode/settings.json

@@ -23,7 +23,6 @@
         "terminal.ansiBrightWhite": "#ffffff",
     },
     "python.testing.pytestEnabled": false,
-    "python.testing.nosetestsEnabled": false,
     "python.testing.unittestEnabled": true,
     "python.testing.unittestArgs": [
         "-v",
@@ -35,26 +34,47 @@
     "python.linting.pylintEnabled": true,
     "python.linting.flake8Enabled": true,
     "python.linting.banditEnabled": true,
+    "python.linting.mypyEnabled": true,
+    "python.linting.mypyArgs": [
+        "--ignore-missing-imports",
+        "--follow-imports=silent",
+        "--show-column-numbers",
+        "--strict"
+    ],
     "python.linting.flake8Args": [
         "--max-line-length=120",
-        "--ignore=E722,F405,E203,W503,W293,E402",
+        "--ignore=E722,F405,E203,W503,W293,E402,E501,E128",
     ],
     "python.linting.banditArgs": [
         "--ignore=B104"
     ],
+    "python.linting.pylintArgs": [
+        "--disable=missing-module-docstring",
+        "--disable=missing-class-docstring",
+        "--disable=missing-function-docstring",
+        "--disable=wrong-import-position",
+        "--disable=raise-missing-from",
+        "--disable=bare-except",
+        "--disable=invalid-name",
+        "--disable=line-too-long",
+        "--disable=consider-using-f-string"
+    ],
+    // python3 -m isort --py=27 --profile=black copyparty/
     "python.formatting.provider": "black",
     "editor.formatOnSave": true,
     "[html]": {
         "editor.formatOnSave": false,
     },
+    "[css]": {
+        "editor.formatOnSave": false,
+    },
     "files.associations": {
         "*.makefile": "makefile"
     },
-    "editor.codeActionsOnSaveTimeout": 9001,
-    "editor.formatOnSaveTimeout": 9001,
-    //
-    //  things you may wanna edit:
-    //
-    "python.pythonPath": "/usr/bin/python3",
-    //"python.linting.enabled": true,
+    "python.formatting.blackArgs": [
+        "-t",
+        "py27"
+    ],
+    "python.linting.enabled": true,
+    "python.pythonPath": "/usr/bin/python3"
 }

.vscode/tasks.json

@@ -0,0 +1,18 @@
+{
+    "version": "2.0.0",
+    "tasks": [
+        {
+            "label": "pre",
+            "command": "true;rm -rf inc/* inc/.hist/;mkdir -p inc;",
+            "type": "shell"
+        },
+        {
+            "label": "no_dbg",
+            "type": "shell",
+            "command": "${config:python.pythonPath}",
+            "args": [
+                ".vscode/launch.py"
+            ]
+        }
+    ]
+}

CODE_OF_CONDUCT.md

@@ -0,0 +1,24 @@
+in the words of Abraham Lincoln:
+
+> Be excellent to each other... and... PARTY ON, DUDES!
+
+more specifically I'll paraphrase some examples from a german automotive corporation as they cover all the bases without being too wordy
+
+## Examples of unacceptable behavior
+* intimidation, harassment, trolling
+* insulting, derogatory, harmful or prejudicial comments
+* posting private information without permission
+* political or personal attacks
+
+## Examples of expected behavior
+* being nice, friendly, welcoming, inclusive, mindful and empathetic
+* acting considerate, modest, respectful
+* using polite and inclusive language
+* criticize constructively and accept constructive criticism
+* respect different points of view
+
+## finally and even more specifically,
+* parse opinions and feedback objectively without prejudice
+  * it's the message that matters, not who said it
+
+aaand that's how you say `be nice` in a way that fills half a floppy w

CONTRIBUTING.md

@@ -0,0 +1,3 @@
+* do something cool
+
+really tho, send a PR or an issue or whatever, all appreciated, anything goes, just behave aight

bin/README.md

@@ -1,4 +1,17 @@
-# copyparty-fuse.py
+# [`up2k.py`](up2k.py)
+* command-line up2k client [(webm)](https://ocv.me/stuff/u2cli.webm)
+* file uploads, file-search, autoresume of aborted/broken uploads
+* faster than browsers
+* if something breaks just restart it
+
+
+# [`partyjournal.py`](partyjournal.py)
+produces a chronological list of all uploads by collecting info from up2k databases and the filesystem
+* outputs a standalone html file
+* optional mapping from IP-addresses to nicknames
+
+
+# [`partyfuse.py`](partyfuse.py)
 * mount a copyparty server as a local filesystem (read-only)
 * **supports Windows!** -- expect `194 MiB/s` sequential read
 * **supports Linux** -- expect `117 MiB/s` sequential read
@@ -10,25 +23,61 @@ filecache is default-on for windows and macos;
 
 note that copyparty should run with `-ed` to enable dotfiles (hidden otherwise)
 
+also consider using [../docs/rclone.md](../docs/rclone.md) instead for 5x performance
+
 
 ## to run this on windows:
 * install [winfsp](https://github.com/billziss-gh/winfsp/releases/latest) and [python 3](https://www.python.org/downloads/)
   * [x] add python 3.x to PATH (it asks during install)
 * `python -m pip install --user fusepy`
-* `python ./copyparty-fuse.py n: http://192.168.1.69:3923/`
+* `python ./partyfuse.py n: http://192.168.1.69:3923/`
 
 10% faster in [msys2](https://www.msys2.org/), 700% faster if debug prints are enabled:
 * `pacman -S mingw64/mingw-w64-x86_64-python{,-pip}`
 * `/mingw64/bin/python3 -m pip install --user fusepy`
-* `/mingw64/bin/python3 ./copyparty-fuse.py [...]`
+* `/mingw64/bin/python3 ./partyfuse.py [...]`
 
 you could replace winfsp with [dokan](https://github.com/dokan-dev/dokany/releases/latest), let me know if you [figure out how](https://github.com/dokan-dev/dokany/wiki/FUSE)  
 (winfsp's sshfs leaks, doesn't look like winfsp itself does, should be fine)
 
 
 
-# copyparty-fuse🅱️.py
+# [`partyfuse2.py`](partyfuse2.py)
 * mount a copyparty server as a local filesystem (read-only)
 * does the same thing except more correct, `samba` approves
 * **supports Linux** -- expect `18 MiB/s` (wait what)
 * **supports Macos** -- probably
+
+
+
+# [`partyfuse-streaming.py`](partyfuse-streaming.py)
+* pretend this doesn't exist
+
+
+
+# [`mtag/`](mtag/)
+* standalone programs which perform misc. file analysis
+* copyparty can Popen programs like these during file indexing to collect additional metadata
+
+
+
+# [`dbtool.py`](dbtool.py)
+upgrade utility which can show db info and help transfer data between databases, for example when a new version of copyparty is incompatible with the old DB and automatically rebuilds the DB from scratch, but you have some really expensive `-mtp` parsers and want to copy over the tags from the old db
+
+for that example (upgrading to v0.11.20), first launch the new version of copyparty like usual, let it make a backup of the old db and rebuild the new db until the point where it starts running mtp (colored messages as it adds the mtp tags), that's when you hit CTRL-C and patch in the old mtp tags from the old db instead
+
+so assuming you have `-mtp` parsers to provide the tags `key` and `.bpm`:
+
+```
+cd /mnt/nas/music/.hist
+~/src/copyparty/bin/dbtool.py -ls up2k.db
+~/src/copyparty/bin/dbtool.py -src up2k.*.v3 up2k.db -cmp
+~/src/copyparty/bin/dbtool.py -src up2k.*.v3 up2k.db -rm-mtp-flag -copy key
+~/src/copyparty/bin/dbtool.py -src up2k.*.v3 up2k.db -rm-mtp-flag -copy .bpm -vac
+```
+
+
+
+# [`prisonparty.sh`](prisonparty.sh)
+* run copyparty in a chroot, preventing any accidental file access
+* creates bindmounts for /bin, /lib, and so on, see `sysdirs=`

bin/dbtool.py

@@ -0,0 +1,304 @@
+#!/usr/bin/env python3
+
+import os
+import sys
+import time
+import shutil
+import sqlite3
+import argparse
+
+DB_VER1 = 3
+DB_VER2 = 5
+
+BY_PATH = None
+NC = None
+
+
+def die(msg):
+    print("\033[31m\n" + msg + "\n\033[0m")
+    sys.exit(1)
+
+
+def read_ver(db):
+    for tab in ["ki", "kv"]:
+        try:
+            c = db.execute(r"select v from {} where k = 'sver'".format(tab))
+        except:
+            continue
+
+        rows = c.fetchall()
+        if rows:
+            return int(rows[0][0])
+
+    return "corrupt"
+
+
+def ls(db):
+    nfiles = next(db.execute("select count(w) from up"))[0]
+    ntags = next(db.execute("select count(w) from mt"))[0]
+    print(f"{nfiles} files")
+    print(f"{ntags} tags\n")
+
+    print("number of occurences for each tag,")
+    print(" 'x' = file has no tags")
+    print(" 't:mtp' = the mtp flag (file not mtp processed yet)")
+    print()
+    for k, nk in db.execute("select k, count(k) from mt group by k order by k"):
+        print(f"{nk:9} {k}")
+
+
+def compare(n1, d1, n2, d2, verbose):
+    nt = next(d1.execute("select count(w) from up"))[0]
+    n = 0
+    miss = 0
+    for w1, rd, fn in d1.execute("select w, rd, fn from up"):
+        n += 1
+        if n % 25_000 == 0:
+            m = f"\033[36mchecked {n:,} of {nt:,} files in {n1} against {n2}\033[0m"
+            print(m)
+
+        if rd.split("/", 1)[0] == ".hist":
+            continue
+
+        if BY_PATH:
+            q = "select w from up where rd = ? and fn = ?"
+            hit = d2.execute(q, (rd, fn)).fetchone()
+        else:
+            q = "select w from up where substr(w,1,16) = ? and +w = ?"
+            hit = d2.execute(q, (w1[:16], w1)).fetchone()
+
+        if not hit:
+            miss += 1
+            if verbose:
+                print(f"file in {n1} missing in {n2}: [{w1}] {rd}/{fn}")
+
+    print(f" {miss} files in {n1} missing in {n2}\n")
+
+    nt = next(d1.execute("select count(w) from mt"))[0]
+    n = 0
+    miss = {}
+    nmiss = 0
+    for w1s, k, v in d1.execute("select * from mt"):
+
+        n += 1
+        if n % 100_000 == 0:
+            m = f"\033[36mchecked {n:,} of {nt:,} tags in {n1} against {n2}, so far {nmiss} missing tags\033[0m"
+            print(m)
+
+        q = "select w, rd, fn from up where substr(w,1,16) = ?"
+        w1, rd, fn = d1.execute(q, (w1s,)).fetchone()
+        if rd.split("/", 1)[0] == ".hist":
+            continue
+
+        if BY_PATH:
+            q = "select w from up where rd = ? and fn = ?"
+            w2 = d2.execute(q, (rd, fn)).fetchone()
+        else:
+            q = "select w from up where substr(w,1,16) = ? and +w = ?"
+            w2 = d2.execute(q, (w1s, w1)).fetchone()
+
+        if w2:
+            w2 = w2[0]
+
+        v2 = None
+        if w2:
+            v2 = d2.execute(
+                "select v from mt where w = ? and +k = ?", (w2[:16], k)
+            ).fetchone()
+            if v2:
+                v2 = v2[0]
+
+        # if v != v2 and v2 and k in [".bpm", "key"] and n2 == "src":
+        #    print(f"{w} [{rd}/{fn}] {k} = [{v}] / [{v2}]")
+
+        if v2 is not None:
+            if k.startswith("."):
+                try:
+                    diff = abs(float(v) - float(v2))
+                    if diff > float(v) / 0.9:
+                        v2 = None
+                    else:
+                        v2 = v
+                except:
+                    pass
+
+            if v != v2:
+                v2 = None
+
+        if v2 is None:
+            nmiss += 1
+            try:
+                miss[k] += 1
+            except:
+                miss[k] = 1
+
+            if verbose:
+                print(f"missing in {n2}: [{w1}] [{rd}/{fn}] {k} = {v}")
+
+    for k, v in sorted(miss.items()):
+        if v:
+            print(f"{n1} has {v:7} more {k:<7} tags than {n2}")
+
+    print(f"in total, {nmiss} missing tags in {n2}\n")
+
+
+def copy_mtp(d1, d2, tag, rm):
+    nt = next(d1.execute("select count(w) from mt where k = ?", (tag,)))[0]
+    n = 0
+    ncopy = 0
+    nskip = 0
+    for w1s, k, v in d1.execute("select * from mt where k = ?", (tag,)):
+        n += 1
+        if n % 25_000 == 0:
+            m = f"\033[36m{n:,} of {nt:,} tags checked, so far {ncopy} copied, {nskip} skipped\033[0m"
+            print(m)
+
+        q = "select w, rd, fn from up where substr(w,1,16) = ?"
+        w1, rd, fn = d1.execute(q, (w1s,)).fetchone()
+        if rd.split("/", 1)[0] == ".hist":
+            continue
+
+        if BY_PATH:
+            q = "select w from up where rd = ? and fn = ?"
+            w2 = d2.execute(q, (rd, fn)).fetchone()
+        else:
+            q = "select w from up where substr(w,1,16) = ? and +w = ?"
+            w2 = d2.execute(q, (w1s, w1)).fetchone()
+
+        if not w2:
+            continue
+
+        w2s = w2[0][:16]
+        hit = d2.execute("select v from mt where w = ? and +k = ?", (w2s, k)).fetchone()
+        if hit:
+            hit = hit[0]
+
+        if hit != v:
+            if NC and hit is not None:
+                nskip += 1
+                continue
+
+            ncopy += 1
+            if hit is not None:
+                d2.execute("delete from mt where w = ? and +k = ?", (w2s, k))
+
+            d2.execute("insert into mt values (?,?,?)", (w2s, k, v))
+            if rm:
+                d2.execute("delete from mt where w = ? and +k = 't:mtp'", (w2s,))
+
+    d2.commit()
+    print(f"copied {ncopy} {tag} tags over, skipped {nskip}")
+
+
+def examples():
+    print(
+        """
+# clearing the journal
+./dbtool.py up2k.db
+
+# copy tags ".bpm" and "key" from old.db to up2k.db, and remove the mtp flag from matching files (so copyparty won't run any mtps on it)
+./dbtool.py -ls up2k.db
+./dbtool.py -src old.db up2k.db -cmp
+./dbtool.py -src old.v3 up2k.db -rm-mtp-flag -copy key
+./dbtool.py -src old.v3 up2k.db -rm-mtp-flag -copy .bpm -vac
+
+"""
+    )
+
+
+def main():
+    global NC, BY_PATH
+    os.system("")
+    print()
+
+    ap = argparse.ArgumentParser()
+    ap.add_argument("db", help="database to work on")
+    ap.add_argument("-h2", action="store_true", help="show examples")
+    ap.add_argument("-src", metavar="DB", type=str, help="database to copy from")
+
+    ap2 = ap.add_argument_group("informational / read-only stuff")
+    ap2.add_argument("-v", action="store_true", help="verbose")
+    ap2.add_argument("-ls", action="store_true", help="list summary for db")
+    ap2.add_argument("-cmp", action="store_true", help="compare databases")
+
+    ap2 = ap.add_argument_group("options which modify target db")
+    ap2.add_argument("-copy", metavar="TAG", type=str, help="mtp tag to copy over")
+    ap2.add_argument(
+        "-rm-mtp-flag",
+        action="store_true",
+        help="when an mtp tag is copied over, also mark that file as done, so copyparty won't run any mtps on those files",
+    )
+    ap2.add_argument("-vac", action="store_true", help="optimize DB")
+
+    ap2 = ap.add_argument_group("behavior modifiers")
+    ap2.add_argument(
+        "-nc",
+        action="store_true",
+        help="no-clobber; don't replace/overwrite existing tags",
+    )
+    ap2.add_argument(
+        "-by-path",
+        action="store_true",
+        help="match files based on location rather than warks (content-hash), use this if the databases have different wark salts",
+    )
+
+    ar = ap.parse_args()
+    if ar.h2:
+        examples()
+        return
+
+    NC = ar.nc
+    BY_PATH = ar.by_path
+
+    for v in [ar.db, ar.src]:
+        if v and not os.path.exists(v):
+            die("database must exist")
+
+    db = sqlite3.connect(ar.db)
+    ds = sqlite3.connect(ar.src) if ar.src else None
+
+    # revert journals
+    for d, p in [[db, ar.db], [ds, ar.src]]:
+        if not d:
+            continue
+
+        pj = "{}-journal".format(p)
+        if not os.path.exists(pj):
+            continue
+
+        d.execute("create table foo (bar int)")
+        d.execute("drop table foo")
+
+    if ar.copy:
+        db.close()
+        shutil.copy2(ar.db, "{}.bak.dbtool.{:x}".format(ar.db, int(time.time())))
+        db = sqlite3.connect(ar.db)
+
+    for d, n in [[ds, "src"], [db, "dst"]]:
+        if not d:
+            continue
+
+        ver = read_ver(d)
+        if ver == "corrupt":
+            die("{} database appears to be corrupt, sorry")
+
+        if ver < DB_VER1 or ver > DB_VER2:
+            m = f"{n} db is version {ver}, this tool only supports versions between {DB_VER1} and {DB_VER2}, please upgrade it with copyparty first"
+            die(m)
+
+    if ar.ls:
+        ls(db)
+
+    if ar.cmp:
+        if not ds:
+            die("need src db to compare against")
+
+        compare("src", ds, "dst", db, ar.v)
+        compare("dst", db, "src", ds, ar.v)
+
+    if ar.copy:
+        copy_mtp(ds, db, ar.copy, ar.rm_mtp_flag)
+
+
+if __name__ == "__main__":
+    main()

bin/mtag/README.md

@@ -0,0 +1,53 @@
+standalone programs which take an audio file as argument
+
+**NOTE:** these all require `-e2ts` to be functional, meaning you need to do at least one of these: `apt install ffmpeg` or `pip3 install mutagen`
+
+some of these rely on libraries which are not MIT-compatible
+
+* [audio-bpm.py](./audio-bpm.py) detects the BPM of music using the BeatRoot Vamp Plugin; imports GPL2
+* [audio-key.py](./audio-key.py) detects the melodic key of music using the Mixxx fork of keyfinder; imports GPL3
+
+these invoke standalone programs which are GPL or similar, so is legally fine for most purposes:
+
+* [media-hash.py](./media-hash.py) generates checksums for audio and video streams; uses FFmpeg (LGPL or GPL)
+* [image-noexif.py](./image-noexif.py) removes exif tags from images; uses exiftool (GPLv1 or artistic-license)
+
+these do not have any problematic dependencies at all:
+
+* [cksum.py](./cksum.py) computes various checksums
+* [exe.py](./exe.py) grabs metadata from .exe and .dll files (example for retrieving multiple tags with one parser)
+* [wget.py](./wget.py) lets you download files by POSTing URLs to copyparty
+
+
+# dependencies
+
+run [`install-deps.sh`](install-deps.sh) to build/install most dependencies required by these programs (supports windows/linux/macos)
+
+*alternatively* (or preferably) use packages from your distro instead, then you'll need at least these:
+
+* from distro: `numpy vamp-plugin-sdk beatroot-vamp mixxx-keyfinder ffmpeg`
+* from pypy: `keyfinder vamp`
+
+
+# usage from copyparty
+
+`copyparty -e2dsa -e2ts` followed by any combination of these:
+* `-mtp key=f,audio-key.py`
+* `-mtp .bpm=f,audio-bpm.py`
+* `-mtp ahash,vhash=f,media-hash.py`
+
+* `f,` makes the detected value replace any existing values
+* the `.` in `.bpm` indicates numeric value
+* assumes the python files are in the folder you're launching copyparty from, replace the filename with a relative/absolute path if that's not the case
+* `mtp` modules will not run if a file has existing tags in the db, so clear out the tags with `-e2tsr` the first time you launch with new `mtp` options
+
+
+## usage with volflags
+
+instead of affecting all volumes, you can set the options for just one volume like so:
+
+`copyparty -v /mnt/nas/music:/music:r:c,e2dsa:c,e2ts` immediately followed by any combination of these:
+
+* `:c,mtp=key=f,audio-key.py`
+* `:c,mtp=.bpm=f,audio-bpm.py`
+* `:c,mtp=ahash,vhash=f,media-hash.py`

bin/mtag/audio-bpm.py

@@ -0,0 +1,70 @@
+#!/usr/bin/env python
+
+import os
+import sys
+import vamp
+import tempfile
+import numpy as np
+import subprocess as sp
+
+from copyparty.util import fsenc
+
+"""
+dep: vamp
+dep: beatroot-vamp
+dep: ffmpeg
+"""
+
+
+def det(tf):
+    # fmt: off
+    sp.check_call([
+        b"ffmpeg",
+        b"-nostdin",
+        b"-hide_banner",
+        b"-v", b"fatal",
+        b"-ss", b"13",
+        b"-y", b"-i", fsenc(sys.argv[1]),
+        b"-map", b"0:a:0",
+        b"-ac", b"1",
+        b"-ar", b"22050",
+        b"-t", b"300",
+        b"-f", b"f32le",
+        fsenc(tf)
+    ])
+    # fmt: on
+
+    with open(tf, "rb") as f:
+        d = np.fromfile(f, dtype=np.float32)
+        try:
+            # 98% accuracy on jcore
+            c = vamp.collect(d, 22050, "beatroot-vamp:beatroot")
+            cl = c["list"]
+        except:
+            # fallback; 73% accuracy
+            plug = "vamp-example-plugins:fixedtempo"
+            c = vamp.collect(d, 22050, plug, parameters={"maxdflen": 40})
+            print(c["list"][0]["label"].split(" ")[0])
+            return
+
+        # throws if detection failed:
+        bpm = float(cl[-1]["timestamp"] - cl[1]["timestamp"])
+        bpm = round(60 * ((len(cl) - 1) / bpm), 2)
+        print(f"{bpm:.2f}")
+
+
+def main():
+    with tempfile.NamedTemporaryFile(suffix=".pcm", delete=False) as f:
+        f.write(b"h")
+        tf = f.name
+
+    try:
+        det(tf)
+    except:
+        pass  # mute
+    finally:
+        os.unlink(tf)
+
+
+if __name__ == "__main__":
+    main()

bin/mtag/audio-key-slicing.py

@@ -0,0 +1,123 @@
+#!/usr/bin/env python
+
+import re
+import os
+import sys
+import tempfile
+import subprocess as sp
+
+import keyfinder
+
+from copyparty.util import fsenc
+
+"""
+dep: github/mixxxdj/libkeyfinder
+dep: pypi/keyfinder
+dep: ffmpeg
+
+note: this is a janky edition of the regular audio-key.py,
+  slicing the files at 20sec intervals and keeping 5sec from each,
+  surprisingly accurate but still garbage (446 ok, 69 bad, 13% miss)
+
+  it is fast tho
+"""
+
+
+def get_duration():
+    # TODO provide ffprobe tags to mtp as json
+
+    # fmt: off
+    dur = sp.check_output([
+        "ffprobe",
+        "-hide_banner",
+        "-v", "fatal",
+        "-show_streams",
+        "-show_format",
+        fsenc(sys.argv[1])
+    ])
+    # fmt: on
+
+    dur = dur.decode("ascii", "replace").split("\n")
+    dur = [x.split("=")[1] for x in dur if x.startswith("duration=")]
+    dur = [float(x) for x in dur if re.match(r"^[0-9\.,]+$", x)]
+    return list(sorted(dur))[-1] if dur else None
+
+
+def get_segs(dur):
+    # keep first 5s of each 20s,
+    # keep entire last segment
+    ofs = 0
+    segs = []
+    while True:
+        seg = [ofs, 5]
+        segs.append(seg)
+        if dur - ofs < 20:
+            seg[-1] = int(dur - seg[0])
+            break
+
+        ofs += 20
+
+    return segs
+
+
+def slice(tf):
+    dur = get_duration()
+    dur = min(dur, 600)  # max 10min
+    segs = get_segs(dur)
+
+    # fmt: off
+    cmd = [
+        "ffmpeg",
+        "-nostdin",
+        "-hide_banner",
+        "-v", "fatal",
+        "-y"
+    ]
+
+    for seg in segs:
+        cmd.extend([
+            "-ss", str(seg[0]),
+            "-i", fsenc(sys.argv[1])
+        ])
+    
+    filt = ""
+    for n, seg in enumerate(segs):
+        filt += "[{}:a:0]atrim=duration={}[a{}]; ".format(n, seg[1], n)
+    
+    prev = "a0"
+    for n in range(1, len(segs)):
+        nxt = "b{}".format(n)
+        filt += "[{}][a{}]acrossfade=d=0.5[{}]; ".format(prev, n, nxt)
+        prev = nxt
+
+    cmd.extend([
+        "-filter_complex", filt[:-2],
+        "-map", "[{}]".format(nxt),
+        "-sample_fmt", "s16",
+        tf
+    ])
+    # fmt: on
+
+    # print(cmd)
+    sp.check_call(cmd)
+
+
+def det(tf):
+    slice(tf)
+    print(keyfinder.key(tf).camelot())
+
+
+def main():
+    with tempfile.NamedTemporaryFile(suffix=".flac", delete=False) as f:
+        f.write(b"h")
+        tf = f.name
+
+    try:
+        det(tf)
+    finally:
+        os.unlink(tf)
+        pass
+
+
+if __name__ == "__main__":
+    main()

bin/mtag/audio-key.py

@@ -0,0 +1,55 @@
+#!/usr/bin/env python
+
+import os
+import sys
+import tempfile
+import subprocess as sp
+import keyfinder
+
+from copyparty.util import fsenc
+
+"""
+dep: github/mixxxdj/libkeyfinder
+dep: pypi/keyfinder
+dep: ffmpeg
+"""
+
+
+# tried trimming the first/last 5th, bad idea,
+# misdetects 9a law field (Sphere Caliber) as 10b,
+# obvious when mixing 9a ghostly parapara ship
+
+
+def det(tf):
+    # fmt: off
+    sp.check_call([
+        b"ffmpeg",
+        b"-nostdin",
+        b"-hide_banner",
+        b"-v", b"fatal",
+        b"-y", b"-i", fsenc(sys.argv[1]),
+        b"-map", b"0:a:0",
+        b"-t", b"300",
+        b"-sample_fmt", b"s16",
+        fsenc(tf)
+    ])
+    # fmt: on
+
+    print(keyfinder.key(tf).camelot())
+
+
+def main():
+    with tempfile.NamedTemporaryFile(suffix=".flac", delete=False) as f:
+        f.write(b"h")
+        tf = f.name
+
+    try:
+        det(tf)
+    except:
+        pass  # mute
+    finally:
+        os.unlink(tf)
+
+
+if __name__ == "__main__":
+    main()

bin/mtag/cksum.py

@@ -0,0 +1,89 @@
+#!/usr/bin/env python3
+
+import sys
+import json
+import zlib
+import struct
+import base64
+import hashlib
+
+try:
+    from copyparty.util import fsenc
+except:
+
+    def fsenc(p):
+        return p
+
+
+"""
+calculates various checksums for uploads,
+usage: -mtp crc32,md5,sha1,sha256b=ad,bin/mtag/cksum.py
+"""
+
+
+def main():
+    config = "crc32 md5 md5b sha1 sha1b sha256 sha256b sha512/240 sha512b/240"
+    # b suffix = base64 encoded
+    # slash = truncate to n bits
+
+    known = {
+        "md5": hashlib.md5,
+        "sha1": hashlib.sha1,
+        "sha256": hashlib.sha256,
+        "sha512": hashlib.sha512,
+    }
+    config = config.split()
+    hashers = {
+        k: v()
+        for k, v in known.items()
+        if k in [x.split("/")[0].rstrip("b") for x in known]
+    }
+    crc32 = 0 if "crc32" in config else None
+
+    with open(fsenc(sys.argv[1]), "rb", 512 * 1024) as f:
+        while True:
+            buf = f.read(64 * 1024)
+            if not buf:
+                break
+
+            for x in hashers.values():
+                x.update(buf)
+
+            if crc32 is not None:
+                crc32 = zlib.crc32(buf, crc32)
+
+    ret = {}
+    for s in config:
+        alg = s.split("/")[0]
+        b64 = alg.endswith("b")
+        alg = alg.rstrip("b")
+        if alg in hashers:
+            v = hashers[alg].digest()
+        elif alg == "crc32":
+            v = crc32
+            if v < 0:
+                v &= 2 ** 32 - 1
+            v = struct.pack(">L", v)
+        else:
+            raise Exception("what is {}".format(s))
+
+        if "/" in s:
+            v = v[: int(int(s.split("/")[1]) / 8)]
+
+        if b64:
+            v = base64.b64encode(v).decode("ascii").rstrip("=")
+        else:
+            try:
+                v = v.hex()
+            except:
+                import binascii
+
+                v = binascii.hexlify(v)
+
+        ret[s] = v
+
+    print(json.dumps(ret, indent=4))
+
+
+if __name__ == "__main__":
+    main()

bin/mtag/exe.py

@@ -0,0 +1,96 @@
+#!/usr/bin/env python
+
+import sys
+import time
+import json
+import pefile
+
+"""
+retrieve exe info,
+example for multivalue providers
+"""
+
+
+def unk(v):
+    return "unk({:04x})".format(v)
+
+
+class PE2(pefile.PE):
+    def __init__(self, *a, **ka):
+        for k in [
+            # -- parse_data_directories:
+            "parse_import_directory",
+            "parse_export_directory",
+            # "parse_resources_directory",
+            "parse_debug_directory",
+            "parse_relocations_directory",
+            "parse_directory_tls",
+            "parse_directory_load_config",
+            "parse_delay_import_directory",
+            "parse_directory_bound_imports",
+            # -- full_load:
+            "parse_rich_header",
+        ]:
+            setattr(self, k, self.noop)
+
+        super(PE2, self).__init__(*a, **ka)
+
+    def noop(*a, **ka):
+        pass
+
+
+try:
+    pe = PE2(sys.argv[1], fast_load=False)
+except:
+    sys.exit(0)
+
+arch = pe.FILE_HEADER.Machine
+if arch == 0x14C:
+    arch = "x86"
+elif arch == 0x8664:
+    arch = "x64"
+else:
+    arch = unk(arch)
+
+try:
+    buildtime = time.gmtime(pe.FILE_HEADER.TimeDateStamp)
+    buildtime = time.strftime("%Y-%m-%d_%H:%M:%S", buildtime)
+except:
+    buildtime = "invalid"
+
+ui = pe.OPTIONAL_HEADER.Subsystem
+if ui == 2:
+    ui = "GUI"
+elif ui == 3:
+    ui = "cmdline"
+else:
+    ui = unk(ui)
+
+extra = {}
+if hasattr(pe, "FileInfo"):
+    for v1 in pe.FileInfo:
+        for v2 in v1:
+            if v2.name != "StringFileInfo":
+                continue
+
+            for v3 in v2.StringTable:
+                for k, v in v3.entries.items():
+                    v = v.decode("utf-8", "replace").strip()
+                    if not v:
+                        continue
+
+                    if k in [b"FileVersion", b"ProductVersion"]:
+                        extra["ver"] = v
+
+                    if k in [b"OriginalFilename", b"InternalName"]:
+                        extra["orig"] = v
+
+r = {
+    "arch": arch,
+    "built": buildtime,
+    "ui": ui,
+    "cksum": "{:08x}".format(pe.OPTIONAL_HEADER.CheckSum),
+}
+r.update(extra)
+
+print(json.dumps(r, indent=4))

bin/mtag/file-ext.py

@@ -0,0 +1,9 @@
+#!/usr/bin/env python
+
+import sys
+
+"""
+example that just prints the file extension
+"""
+
+print(sys.argv[1].split(".")[-1])

bin/mtag/guestbook-read.py

@@ -0,0 +1,61 @@
+#!/usr/bin/env python3
+
+"""
+fetch latest msg from guestbook and return as tag
+
+example copyparty config to use this:
+  --urlform save,get -vsrv/hello:hello:w:c,e2ts,mtp=guestbook=t10,ad,p,bin/mtag/guestbook-read.py:mte=+guestbook
+
+explained:
+  for realpath srv/hello (served at /hello), write-only for eveyrone,
+  enable file analysis on upload (e2ts),
+  use mtp plugin "bin/mtag/guestbook-read.py" to provide metadata tag "guestbook",
+  do this on all uploads regardless of extension,
+  t10 = 10 seconds timeout for each dwonload,
+  ad = parse file regardless if FFmpeg thinks it is audio or not
+  p = request upload info as json on stdin (need ip)
+  mte=+guestbook enabled indexing of that tag for this volume
+
+PS: this requires e2ts to be functional,
+  meaning you need to do at least one of these:
+   * apt install ffmpeg
+   * pip3 install mutagen
+"""
+
+
+import json
+import os
+import sqlite3
+import sys
+
+
+# set 0 to allow infinite msgs from one IP,
+# other values delete older messages to make space,
+# so 1 only keeps latest msg
+NUM_MSGS_TO_KEEP = 1
+
+
+def main():
+    fp = os.path.abspath(sys.argv[1])
+    fdir = os.path.dirname(fp)
+
+    zb = sys.stdin.buffer.read()
+    zs = zb.decode("utf-8", "replace")
+    md = json.loads(zs)
+
+    ip = md["up_ip"]
+
+    # can put the database inside `fdir` if you'd like,
+    # by default it saves to PWD:
+    # os.chdir(fdir)
+
+    db = sqlite3.connect("guestbook.db3")
+    with db:
+        t = "select msg from gb where ip = ? order by ts desc"
+        r = db.execute(t, (ip,)).fetchone()
+        if r:
+            print(r[0])
+
+
+if __name__ == "__main__":
+    main()

bin/mtag/guestbook.py

@@ -0,0 +1,111 @@
+#!/usr/bin/env python3
+
+"""
+store messages from users in an sqlite database
+which can be read from another mtp for example
+
+takes input from application/x-www-form-urlencoded POSTs,
+for example using the message/pager function on the website
+
+example copyparty config to use this:
+  --urlform save,get -vsrv/hello:hello:w:c,e2ts,mtp=xgb=ebin,t10,ad,p,bin/mtag/guestbook.py:mte=+xgb
+
+explained:
+  for realpath srv/hello (served at /hello),write-only for eveyrone,
+  enable file analysis on upload (e2ts),
+  use mtp plugin "bin/mtag/guestbook.py" to provide metadata tag "xgb",
+  do this on all uploads with the file extension "bin",
+  t300 = 300 seconds timeout for each dwonload,
+  ad = parse file regardless if FFmpeg thinks it is audio or not
+  p = request upload info as json on stdin
+  mte=+xgb enabled indexing of that tag for this volume
+
+PS: this requires e2ts to be functional,
+  meaning you need to do at least one of these:
+   * apt install ffmpeg
+   * pip3 install mutagen
+"""
+
+
+import json
+import os
+import sqlite3
+import sys
+from urllib.parse import unquote_to_bytes as unquote
+
+
+# set 0 to allow infinite msgs from one IP,
+# other values delete older messages to make space,
+# so 1 only keeps latest msg
+NUM_MSGS_TO_KEEP = 1
+
+
+def main():
+    fp = os.path.abspath(sys.argv[1])
+    fdir = os.path.dirname(fp)
+    fname = os.path.basename(fp)
+    if not fname.startswith("put-") or not fname.endswith(".bin"):
+        raise Exception("not a post file")
+
+    zb = sys.stdin.buffer.read()
+    zs = zb.decode("utf-8", "replace")
+    md = json.loads(zs)
+
+    buf = b""
+    with open(fp, "rb") as f:
+        while True:
+            b = f.read(4096)
+            buf += b
+            if len(buf) > 4096:
+                raise Exception("too big")
+
+            if not b:
+                break
+
+    if not buf:
+        raise Exception("file is empty")
+
+    buf = unquote(buf.replace(b"+", b" "))
+    txt = buf.decode("utf-8")
+
+    if not txt.startswith("msg="):
+        raise Exception("does not start with msg=")
+
+    ip = md["up_ip"]
+    ts = md["up_at"]
+    txt = txt[4:]
+
+    # can put the database inside `fdir` if you'd like,
+    # by default it saves to PWD:
+    # os.chdir(fdir)
+
+    db = sqlite3.connect("guestbook.db3")
+    try:
+        db.execute("select 1 from gb").fetchone()
+    except:
+        with db:
+            db.execute("create table gb (ip text, ts real, msg text)")
+            db.execute("create index gb_ip on gb(ip)")
+
+    with db:
+        if NUM_MSGS_TO_KEEP == 1:
+            t = "delete from gb where ip = ?"
+            db.execute(t, (ip,))
+
+        t = "insert into gb values (?,?,?)"
+        db.execute(t, (ip, ts, txt))
+
+        if NUM_MSGS_TO_KEEP > 1:
+            t = "select ts from gb where ip = ? order by ts desc"
+            hits = db.execute(t, (ip,)).fetchall()
+
+            if len(hits) > NUM_MSGS_TO_KEEP:
+                lim = hits[NUM_MSGS_TO_KEEP][0]
+                t = "delete from gb where ip = ? and ts <= ?"
+                db.execute(t, (ip, lim))
+
+    print(txt)
+
+
+if __name__ == "__main__":
+    main()

bin/mtag/image-noexif.py

@@ -0,0 +1,95 @@
+#!/usr/bin/env python3
+
+"""
+remove exif tags from uploaded images
+
+dependencies:
+  exiftool
+
+about:
+  creates a "noexif" subfolder and puts exif-stripped copies of each image there,
+  the reason for the subfolder is to avoid issues with the up2k.db / deduplication:
+
+  if the original image is modified in-place, then copyparty will keep the original
+  hash in up2k.db for a while (until the next volume rescan), so if the image is
+  reuploaded after a rescan then the upload will be renamed and kept as a dupe
+
+  alternatively you could switch the logic around, making a copy of the original
+  image into a subfolder named "exif" and modify the original in-place, but then
+  up2k.db will be out of sync until the next rescan, so any additional uploads
+  of the same image will get symlinked (deduplicated) to the modified copy
+  instead of the original in "exif"
+
+  or maybe delete the original image after processing, that would kinda work too
+
+example copyparty config to use this:
+  -v/mnt/nas/pics:pics:rwmd,ed:c,e2ts,mte=+noexif:c,mtp=noexif=ejpg,ejpeg,ad,bin/mtag/image-noexif.py
+
+explained:
+  for realpath /mnt/nas/pics (served at /pics) with read-write-modify-delete for ed,
+  enable file analysis on upload (e2ts),
+  append "noexif" to the list of known tags (mtp),
+  and use mtp plugin "bin/mtag/image-noexif.py" to provide that tag,
+  do this on all uploads with the file extension "jpg" or "jpeg",
+  ad = parse file regardless if FFmpeg thinks it is audio or not
+
+PS: this requires e2ts to be functional,
+  meaning you need to do at least one of these:
+   * apt install ffmpeg
+   * pip3 install mutagen
+  and your python must have sqlite3 support compiled in
+"""
+
+
+import os
+import sys
+import filecmp
+import subprocess as sp
+
+try:
+    from copyparty.util import fsenc
+except:
+
+    def fsenc(p):
+        return p.encode("utf-8")
+
+
+def main():
+    cwd, fn = os.path.split(sys.argv[1])
+    if os.path.basename(cwd) == "noexif":
+        return
+
+    os.chdir(cwd)
+    f1 = fsenc(fn)
+    f2 = os.path.join(b"noexif", f1)
+    cmd = [
+        b"exiftool",
+        b"-exif:all=",
+        b"-iptc:all=",
+        b"-xmp:all=",
+        b"-P",
+        b"-o",
+        b"noexif/",
+        b"--",
+        f1,
+    ]
+    sp.check_output(cmd)
+    if not os.path.exists(f2):
+        print("failed")
+        return
+
+    if filecmp.cmp(f1, f2, shallow=False):
+        print("clean")
+    else:
+        print("exif")
+
+    # lastmod = os.path.getmtime(f1)
+    # times = (int(time.time()), int(lastmod))
+    # os.utime(f2, times)
+
+
+if __name__ == "__main__":
+    try:
+        main()
+    except:
+        pass

bin/mtag/install-deps.sh

@@ -0,0 +1,301 @@
+#!/bin/bash
+set -e
+
+
+# install dependencies for audio-*.py
+#
+# linux/alpine: requires gcc g++ make cmake patchelf {python3,ffmpeg,fftw,libsndfile}-dev py3-{wheel,pip} py3-numpy{,-dev}
+# linux/debian: requires libav{codec,device,filter,format,resample,util}-dev {libfftw3,python3,libsndfile1}-dev python3-{numpy,pip} vamp-{plugin-sdk,examples} patchelf cmake
+# linux/fedora: requires gcc gcc-c++ make cmake patchelf {python3,ffmpeg,fftw,libsndfile}-devel python3-numpy vamp-plugin-sdk qm-vamp-plugins
+# win64: requires msys2-mingw64 environment
+# macos: requires macports
+#
+# has the following manual dependencies, especially on mac:
+#   https://www.vamp-plugins.org/pack.html
+#
+# installs stuff to the following locations:
+#   ~/pe/
+#   whatever your python uses for --user packages
+#
+# does the following terrible things:
+#   modifies the keyfinder python lib to load the .so in ~/pe
+
+
+linux=1
+
+win=
+[ ! -z "$MSYSTEM" ] || [ -e /msys2.exe ] && {
+	[ "$MSYSTEM" = MINGW64 ] || {
+		echo windows detected, msys2-mingw64 required
+		exit 1
+	}
+	pacman -S --needed mingw-w64-x86_64-{ffmpeg,python,python-pip,vamp-plugin-sdk}
+	win=1
+	linux=
+}
+
+mac=
+[ $(uname -s) = Darwin ] && {
+	#pybin="$(printf '%s\n' /opt/local/bin/python* | (sed -E 's/(.*\/[^/0-9]+)([0-9]?[^/]*)$/\2 \1/' || cat) | (sort -nr || cat) | (sed -E 's/([^ ]*) (.*)/\2\1/' || cat) | grep -E '/(python|pypy)[0-9\.-]*$' | head -n 1)"
+	pybin=/opt/local/bin/python3.9
+	[ -e "$pybin" ] || {
+		echo mac detected, python3 from macports required
+		exit 1
+	}
+	pkgs='ffmpeg python39 py39-wheel'
+	ninst=$(port installed | awk '/^  /{print$1}' | sort | uniq | grep -E '^('"$(echo "$pkgs" | tr ' ' '|')"')$' | wc -l)
+	[ $ninst -eq 3 ] || {
+		sudo port install $pkgs
+	}
+	mac=1
+	linux=
+}
+
+hash -r
+
+[ $mac ] || {
+	command -v python3 && pybin=python3 || pybin=python
+}
+
+$pybin -m pip install --user numpy
+
+
+command -v gnutar && tar() { gnutar "$@"; }
+command -v gtar && tar() { gtar "$@"; }
+command -v gsed && sed() { gsed "$@"; }
+
+
+need() {
+	command -v $1 >/dev/null || {
+		echo need $1
+		exit 1
+	}
+}
+need cmake
+need ffmpeg
+need $pybin
+#need patchelf
+
+
+td="$(mktemp -d)"
+cln() {
+	rm -rf "$td"
+}
+trap cln EXIT
+cd "$td"
+pwd
+
+
+dl_text() {
+	command -v curl >/dev/null && exec curl "$@"
+	exec wget -O- "$@"
+}
+dl_files() {
+	local yolo= ex=
+	[ $1 = "yolo" ] && yolo=1 && ex=k && shift
+	command -v curl >/dev/null && exec curl -${ex}JOL "$@"
+	
+	[ $yolo ] && ex=--no-check-certificate
+	exec wget --trust-server-names $ex "$@"
+}
+export -f dl_files
+
+
+github_tarball() {
+	rm -rf g
+	mkdir g
+	cd g
+	dl_text "$1" |
+	tee ../json |
+	(
+		# prefer jq if available
+		jq -r '.tarball_url' ||
+
+		# fallback to awk (sorry)
+		awk -F\" '/"tarball_url": "/ {print$4}'
+	) |
+	tee /dev/stderr |
+	head -n 1 |
+	tr -d '\r' | tr '\n' '\0' |
+	xargs -0 bash -c 'dl_files "$@"' _
+	mv * ../tgz
+	cd ..
+}
+
+
+gitlab_tarball() {
+	dl_text "$1" |
+	tee json |
+	(
+		# prefer jq if available
+		jq -r '.[0].assets.sources[]|select(.format|test("tar.gz")).url' ||
+
+		# fallback to abomination
+		tr \" '\n' | grep -E '\.tar\.gz$' | head -n 1
+	) |
+	tee /dev/stderr |
+	head -n 1 |
+	tr -d '\r' | tr '\n' '\0' |
+	tee links |
+	xargs -0 bash -c 'dl_files "$@"' _
+}
+
+
+install_keyfinder() {
+	# windows support:
+	#   use msys2 in mingw-w64 mode
+	#   pacman -S --needed mingw-w64-x86_64-{ffmpeg,python}
+	
+	[ -e $HOME/pe/keyfinder ] && {
+		echo found a keyfinder build in ~/pe, skipping
+		return
+	}
+
+	cd "$td"
+	github_tarball https://api.github.com/repos/mixxxdj/libkeyfinder/releases/latest
+	ls -al
+
+	tar -xf tgz
+	rm tgz
+	cd mixxxdj-libkeyfinder*
+	
+	h="$HOME"
+	so="lib/libkeyfinder.so"
+	memes=(-DBUILD_TESTING=OFF)
+
+	[ $win ] &&
+		so="bin/libkeyfinder.dll" &&
+		h="$(printf '%s\n' "$USERPROFILE" | tr '\\' '/')" &&
+		memes+=(-G "MinGW Makefiles")
+	
+	[ $mac ] &&
+		so="lib/libkeyfinder.dylib"
+
+	cmake -DCMAKE_INSTALL_PREFIX="$h/pe/keyfinder" "${memes[@]}" -S . -B build
+	cmake --build build --parallel $(nproc || echo 4)
+	cmake --install build
+
+	libpath="$h/pe/keyfinder/$so"
+	[ $linux ] && [ ! -e "$libpath" ] &&
+		so=lib64/libkeyfinder.so
+	
+	libpath="$h/pe/keyfinder/$so"
+	[ -e "$libpath" ] || {
+		echo "so not found at $sop"
+		exit 1
+	}
+	
+	# rm -rf /Users/ed/Library/Python/3.9/lib/python/site-packages/*keyfinder*
+	CFLAGS="-I$h/pe/keyfinder/include -I/opt/local/include -I/usr/include/ffmpeg" \
+	LDFLAGS="-L$h/pe/keyfinder/lib -L$h/pe/keyfinder/lib64 -L/opt/local/lib" \
+	PKG_CONFIG_PATH=/c/msys64/mingw64/lib/pkgconfig \
+	$pybin -m pip install --user keyfinder
+
+	pypath="$($pybin -c 'import keyfinder; print(keyfinder.__file__)')"
+	for pyso in "${pypath%/*}"/*.so; do
+		[ -e "$pyso" ] || break
+		patchelf --set-rpath "${libpath%/*}" "$pyso" ||
+			echo "WARNING: patchelf failed (only fatal on musl-based distros)"
+	done
+	
+	mv "$pypath"{,.bak}
+	(
+		printf 'import ctypes\nctypes.cdll.LoadLibrary("%s")\n' "$libpath"
+		cat "$pypath.bak"
+	) >"$pypath"
+
+	echo
+	echo libkeyfinder successfully installed to the following locations:
+	echo "  $libpath"
+	echo "  $pypath"
+}
+
+
+have_beatroot() {
+	$pybin -c 'import vampyhost, sys; plugs = vampyhost.list_plugins(); sys.exit(0 if "beatroot-vamp:beatroot" in plugs else 1)'
+}
+
+
+install_vamp() {
+	# windows support:
+	#   use msys2 in mingw-w64 mode
+	#   pacman -S --needed mingw-w64-x86_64-{ffmpeg,python,python-pip,vamp-plugin-sdk}
+	
+	$pybin -m pip install --user vamp
+
+	cd "$td"
+	echo '#include <vamp-sdk/Plugin.h>' | gcc -x c -c -o /dev/null - || [ -e ~/pe/vamp-sdk ] || {
+		printf '\033[33mcould not find the vamp-sdk, building from source\033[0m\n'
+		(dl_files yolo https://code.soundsoftware.ac.uk/attachments/download/2588/vamp-plugin-sdk-2.9.0.tar.gz)
+		sha512sum -c <(
+			echo "7ef7f837d19a08048b059e0da408373a7964ced452b290fae40b85d6d70ca9000bcfb3302cd0b4dc76cf2a848528456f78c1ce1ee0c402228d812bd347b6983b  -"
+		) <vamp-plugin-sdk-2.9.0.tar.gz
+		tar -xf vamp-plugin-sdk-2.9.0.tar.gz
+		rm -- *.tar.gz
+		ls -al
+		cd vamp-plugin-sdk-*
+		./configure --prefix=$HOME/pe/vamp-sdk
+		make -j1 install
+	}
+
+	cd "$td"
+	have_beatroot || {
+		printf '\033[33mcould not find the vamp beatroot plugin, building from source\033[0m\n'
+		(dl_files yolo https://code.soundsoftware.ac.uk/attachments/download/885/beatroot-vamp-v1.0.tar.gz)
+		sha512sum -c <(
+			echo "1f444d1d58ccf565c0adfe99f1a1aa62789e19f5071e46857e2adfbc9d453037bc1c4dcb039b02c16240e9b97f444aaff3afb625c86aa2470233e711f55b6874  -"
+		) <beatroot-vamp-v1.0.tar.gz
+		tar -xf beatroot-vamp-v1.0.tar.gz 
+		rm -- *.tar.gz
+		cd beatroot-vamp-v1.0
+		[ -e ~/pe/vamp-sdk ] &&
+			sed -ri 's`^(CFLAGS :=.*)`\1 -I'$HOME'/pe/vamp-sdk/include`' Makefile.linux
+		make -f Makefile.linux -j4 LDFLAGS=-L$HOME/pe/vamp-sdk/lib
+		# /home/ed/vamp /home/ed/.vamp /usr/local/lib/vamp
+		mkdir ~/vamp
+		cp -pv beatroot-vamp.* ~/vamp/
+	}
+	
+	have_beatroot &&
+		printf '\033[32mfound the vamp beatroot plugin, nice\033[0m\n' ||
+		printf '\033[31mWARNING: could not find the vamp beatroot plugin, please install it for optimal results\033[0m\n'
+}
+
+
+# not in use because it kinda segfaults, also no windows support
+install_soundtouch() {
+	cd "$td"
+	gitlab_tarball https://gitlab.com/api/v4/projects/soundtouch%2Fsoundtouch/releases
+	
+	tar -xvf soundtouch-*
+	rm -- *.tar.gz
+	cd soundtouch-*
+	
+	# https://github.com/jrising/pysoundtouch
+	./bootstrap
+	./configure --enable-integer-samples CXXFLAGS="-fPIC" --prefix="$HOME/pe/soundtouch"
+	make -j$(nproc || echo 4)
+	make install
+	
+	CFLAGS=-I$HOME/pe/soundtouch/include/ \
+	LDFLAGS=-L$HOME/pe/soundtouch/lib \
+	$pybin -m pip install --user git+https://github.com/snowxmas/pysoundtouch.git
+	
+	pypath="$($pybin -c 'import importlib; print(importlib.util.find_spec("soundtouch").origin)')"
+	libpath="$(echo "$HOME/pe/soundtouch/lib/")"
+	patchelf --set-rpath "$libpath" "$pypath"
+
+	echo
+	echo soundtouch successfully installed to the following locations:
+	echo "  $libpath"
+	echo "  $pypath"
+}
+
+
+[ "$1" = keyfinder ] && { install_keyfinder; exit $?; }
+[ "$1" = soundtouch ] && { install_soundtouch; exit $?; }
+[ "$1" = vamp ] && { install_vamp; exit $?; }
+
+echo no args provided, installing keyfinder and vamp
+install_keyfinder
+install_vamp

bin/mtag/media-hash.py

@@ -0,0 +1,73 @@
+#!/usr/bin/env python
+
+import re
+import sys
+import json
+import time
+import base64
+import hashlib
+import subprocess as sp
+
+try:
+    from copyparty.util import fsenc
+except:
+
+    def fsenc(p):
+        return p.encode("utf-8")
+
+
+"""
+dep: ffmpeg
+"""
+
+
+def det():
+    # fmt: off
+    cmd = [
+        b"ffmpeg",
+        b"-nostdin",
+        b"-hide_banner",
+        b"-v", b"fatal",
+        b"-i", fsenc(sys.argv[1]),
+        b"-f", b"framemd5",
+        b"-"
+    ]
+    # fmt: on
+
+    p = sp.Popen(cmd, stdout=sp.PIPE)
+    # ps = io.TextIOWrapper(p.stdout, encoding="utf-8")
+    ps = p.stdout
+
+    chans = {}
+    for ln in ps:
+        if ln.startswith(b"#stream#"):
+            break
+
+        m = re.match(r"^#media_type ([0-9]): ([a-zA-Z])", ln.decode("utf-8"))
+        if m:
+            chans[m.group(1)] = m.group(2)
+
+    hashers = [hashlib.sha512(), hashlib.sha512()]
+    for ln in ps:
+        n = int(ln[:1])
+        v = ln.rsplit(b",", 1)[-1].strip()
+        hashers[n].update(v)
+
+    r = {}
+    for k, v in chans.items():
+        dg = hashers[int(k)].digest()[:12]
+        dg = base64.urlsafe_b64encode(dg).decode("ascii")
+        r[v[0].lower() + "hash"] = dg
+
+    print(json.dumps(r, indent=4))
+
+
+def main():
+    try:
+        det()
+    except:
+        pass  # mute
+
+
+if __name__ == "__main__":
+    main()

bin/mtag/mousepad.py

@@ -0,0 +1,38 @@
+#!/usr/bin/env python3
+
+import os
+import sys
+import subprocess as sp
+
+
+"""
+mtp test -- opens a texteditor
+
+usage:
+  -vsrv/v1:v1:r:c,mte=+x1:c,mtp=x1=ad,p,bin/mtag/mousepad.py
+
+explained:
+  c,mte: list of tags to index in this volume
+  c,mtp: add new tag provider
+     x1: dummy tag to provide
+     ad: dontcare if audio or not
+      p: priority 1 (run after initial tag-scan with ffprobe or mutagen)
+"""
+
+
+def main():
+    env = os.environ.copy()
+    env["DISPLAY"] = ":0.0"
+
+    if False:
+        # open the uploaded file
+        fp = sys.argv[-1]
+    else:
+        # display stdin contents (`oth_tags`)
+        fp = "/dev/stdin"
+
+    p = sp.Popen(["/usr/bin/mousepad", fp])
+    p.communicate()
+
+
+main()

bin/mtag/rclone-upload.py

@@ -0,0 +1,76 @@
+#!/usr/bin/env python
+
+import json
+import os
+import subprocess as sp
+import sys
+import time
+
+try:
+    from copyparty.util import fsenc
+except:
+
+    def fsenc(p):
+        return p.encode("utf-8")
+
+
+_ = r"""
+first checks the tag "vidchk" which must be "ok" to continue,
+then uploads all files to some cloud storage (RCLONE_REMOTE)
+and DELETES THE ORIGINAL FILES if rclone returns 0 ("success")
+
+deps:
+  rclone
+
+usage:
+  -mtp x2=t43200,ay,p2,bin/mtag/rclone-upload.py
+
+explained:
+t43200: timeout 12h
+    ay: only process files which contain audio (including video with audio)
+    p2: set priority 2 (after vidchk's suggested priority of 1),
+          so the output of vidchk will be passed in here
+
+complete usage example as vflags along with vidchk:
+  -vsrv/vidchk:vidchk:r:rw,ed:c,e2dsa,e2ts,mtp=vidchk=t600,p,bin/mtag/vidchk.py:c,mtp=rupload=t43200,ay,p2,bin/mtag/rclone-upload.py:c,mte=+vidchk,rupload
+
+setup: see https://rclone.org/drive/
+
+if you wanna use this script standalone / separately from copyparty,
+either set CONDITIONAL_UPLOAD False or provide the following stdin:
+  {"vidchk":"ok"}
+"""
+
+
+RCLONE_REMOTE = "notmybox"
+CONDITIONAL_UPLOAD = True
+
+
+def main():
+    fp = sys.argv[1]
+    if CONDITIONAL_UPLOAD:
+        zb = sys.stdin.buffer.read()
+        zs = zb.decode("utf-8", "replace")
+        md = json.loads(zs)
+
+        chk = md.get("vidchk", None)
+        if chk != "ok":
+            print(f"vidchk={chk}", file=sys.stderr)
+            sys.exit(1)
+
+    dst = f"{RCLONE_REMOTE}:".encode("utf-8")
+    cmd = [b"rclone", b"copy", b"--", fsenc(fp), dst]
+
+    t0 = time.time()
+    try:
+        sp.check_call(cmd)
+    except:
+        print("rclone failed", file=sys.stderr)
+        sys.exit(1)
+
+    print(f"{time.time() - t0:.1f} sec")
+    os.unlink(fsenc(fp))
+
+
+if __name__ == "__main__":
+    main()

bin/mtag/res/twitter-unmute.user.js

@@ -0,0 +1,21 @@
+// ==UserScript==
+// @name         twitter-unmute
+// @namespace    http://ocv.me/
+// @version      0.1
+// @description  memes
+// @author       ed <irc.rizon.net>
+// @match        https://twitter.com/*
+// @icon         https://www.google.com/s2/favicons?domain=twitter.com
+// @grant        GM_addStyle
+// ==/UserScript==
+
+function grunnur() {
+    setInterval(function () {
+        //document.querySelector('div[aria-label="Unmute"]').click();
+        document.querySelector('video').muted = false;
+    }, 200);
+}
+
+var scr = document.createElement('script');
+scr.textContent = '(' + grunnur.toString() + ')();';
+(document.head || document.getElementsByTagName('head')[0]).appendChild(scr);

bin/mtag/res/yt-ipr.conf

@@ -0,0 +1,39 @@
+# example config file to use copyparty as a youtube manifest collector,
+# use with copyparty like:  python copyparty.py -c yt-ipr.conf
+#
+# see docs/example.conf for a better explanation of the syntax, but
+# newlines are block separators, so adding blank lines inside a volume definition is bad
+# (use comments as separators instead)
+
+
+# create user ed, password wark
+u ed:wark
+
+
+# create a volume at /ytm which stores files at ./srv/ytm
+./srv/ytm
+/ytm
+# write-only, but read-write for user ed
+w
+rw ed
+# rescan the volume on startup
+c e2dsa
+# collect tags from all new files since last scan
+c e2ts
+# optionally enable compression to make the files 50% smaller
+c pk
+# only allow uploads which are between 16k and 1m large
+c sz=16k-1m
+# allow up to 10 uploads over 5 minutes from each ip
+c maxn=10,300
+# move uploads into subfolders: YEAR-MONTH / DAY-HOUR / <upload>
+c rotf=%Y-%m/%d-%H
+# delete uploads when they are 24 hours old
+c lifetime=86400
+# add the parser and tell copyparty what tags it can expect from it
+c mtp=yt-id,yt-title,yt-author,yt-channel,yt-views,yt-private,yt-manifest,yt-expires=bin/mtag/yt-ipr.py
+# decide which tags we want to index and in what order
+c mte=yt-id,yt-title,yt-author,yt-channel,yt-views,yt-private,yt-manifest,yt-expires
+
+
+# create any other volumes you'd like down here, or merge this with an existing config file

bin/mtag/res/yt-ipr.user.js

@@ -0,0 +1,47 @@
+// ==UserScript==
+// @name    youtube-playerdata-hub
+// @match   https://youtube.com/*
+// @match   https://*.youtube.com/*
+// @version 1.0
+// @grant   GM_addStyle
+// ==/UserScript==
+
+function main() {
+    var server = 'https://127.0.0.1:3923/ytm?pw=wark',
+        interval = 60; // sec
+
+    var sent = {};
+    function send(txt, mf_url, desc) {
+        if (sent[mf_url])
+            return;
+
+        fetch(server + '&_=' + Date.now(), { method: "PUT", body: txt });
+        console.log('[yt-pdh] yeet %d bytes, %s', txt.length, desc);
+        sent[mf_url] = 1;
+    }
+
+    function collect() {
+        try {
+            var pd = document.querySelector('ytd-watch-flexy');
+            if (!pd)
+                return console.log('[yt-pdh] no video found');
+
+            pd = pd.playerData;
+            var mu = pd.streamingData.dashManifestUrl || pd.streamingData.hlsManifestUrl;
+            if (!mu || !mu.length)
+                return console.log('[yt-pdh] no manifest found');
+
+            var desc = pd.videoDetails.videoId + ', ' + pd.videoDetails.title;
+            send(JSON.stringify(pd), mu, desc);
+        }
+        catch (ex) {
+            console.log("[yt-pdh]", ex);
+        }
+    }
+    setInterval(collect, interval * 1000);
+}
+
+var scr = document.createElement('script');
+scr.textContent = '(' + main.toString() + ')();';
+(document.head || document.getElementsByTagName('head')[0]).appendChild(scr);
+console.log('[yt-pdh] a');

bin/mtag/sleep.py

@@ -0,0 +1,8 @@
+#!/usr/bin/env python
+
+import time
+import random
+
+v = random.random() * 6
+time.sleep(v)
+print(f"{v:.2f}")

bin/mtag/very-bad-idea.py

@@ -0,0 +1,139 @@
+#!/usr/bin/env python3
+
+"""
+use copyparty as a chromecast replacement:
+  * post a URL and it will open in the default browser
+  * upload a file and it will open in the default application
+  * the `key` command simulates keyboard input
+  * the `x` command executes other xdotool commands
+  * the `c` command executes arbitrary unix commands
+
+the android app makes it a breeze to post pics and links:
+  https://github.com/9001/party-up/releases
+  (iOS devices have to rely on the web-UI)
+
+goes without saying, but this is HELLA DANGEROUS,
+  GIVES RCE TO ANYONE WHO HAVE UPLOAD PERMISSIONS
+
+example copyparty config to use this:
+  --urlform save,get -v.::w:c,e2d,e2t,mte=+a1:c,mtp=a1=ad,kn,c0,bin/mtag/very-bad-idea.py
+
+recommended deps:
+  apt install xdotool libnotify-bin
+  https://github.com/9001/copyparty/blob/hovudstraum/contrib/plugins/meadup.js
+
+and you probably want `twitter-unmute.user.js` from the res folder
+
+
+-----------------------------------------------------------------------
+-- startup script:
+-----------------------------------------------------------------------
+
+#!/bin/bash
+set -e
+
+# create qr code
+ip=$(ip r | awk '/^default/{print$(NF-2)}'); echo http://$ip:3923/ | qrencode -o - -s 4 >/dev/shm/cpp-qr.png
+/usr/bin/feh -x /dev/shm/cpp-qr.png &
+
+# reposition and make topmost (with janky raspbian support)
+( sleep 0.5
+xdotool search --name cpp-qr.png windowactivate --sync windowmove 1780 0
+wmctrl -r :ACTIVE: -b toggle,above || true
+
+ps aux | grep -E 'sleep[ ]7\.27' ||
+while true; do
+  w=$(xdotool getactivewindow)
+  xdotool search --name cpp-qr.png windowactivate windowraise windowfocus
+  xdotool windowactivate $w
+  xdotool windowfocus $w
+  sleep 7.27 || break
+done &
+xeyes  # distraction window to prevent ^w from closing the qr-code
+) &
+
+# bail if copyparty is already running
+ps aux | grep -E '[3] copy[p]arty' && exit 0
+
+# dumb chrome wrapper to allow autoplay
+cat >/usr/local/bin/chromium-browser <<'EOF'
+#!/bin/bash
+set -e
+/usr/bin/chromium-browser --autoplay-policy=no-user-gesture-required "$@"
+EOF
+chmod 755 /usr/local/bin/chromium-browser
+
+# start the server  (note: replace `-v.::rw:` with `-v.::w:` to disallow retrieving uploaded stuff)
+cd ~/Downloads; python3 copyparty-sfx.py --urlform save,get -v.::rw:c,e2d,e2t,mte=+a1:c,mtp=a1=ad,kn,very-bad-idea.py
+
+"""
+
+
+import os
+import sys
+import time
+import subprocess as sp
+from urllib.parse import unquote_to_bytes as unquote
+
+
+def main():
+    fp = os.path.abspath(sys.argv[1])
+    with open(fp, "rb") as f:
+        txt = f.read(4096)
+
+    if txt.startswith(b"msg="):
+        open_post(txt)
+    else:
+        open_url(fp)
+
+
+def open_post(txt):
+    txt = unquote(txt.replace(b"+", b" ")).decode("utf-8")[4:]
+    try:
+        k, v = txt.split(" ", 1)
+    except:
+        open_url(txt)
+
+    if k == "key":
+        sp.call(["xdotool", "key"] + v.split(" "))
+    elif k == "x":
+        sp.call(["xdotool"] + v.split(" "))
+    elif k == "c":
+        env = os.environ.copy()
+        while " " in v:
+            v1, v2 = v.split(" ", 1)
+            if "=" not in v1:
+                break
+
+            ek, ev = v1.split("=", 1)
+            env[ek] = ev
+            v = v2
+
+        sp.call(v.split(" "), env=env)
+    else:
+        open_url(txt)
+
+
+def open_url(txt):
+    ext = txt.rsplit(".")[-1].lower()
+    sp.call(["notify-send", "--", txt])
+    if ext not in ["jpg", "jpeg", "png", "gif", "webp"]:
+        # sp.call(["wmctrl", "-c", ":ACTIVE:"])  # closes the active window correctly
+        sp.call(["killall", "vlc"])
+        sp.call(["killall", "mpv"])
+        sp.call(["killall", "feh"])
+        time.sleep(0.5)
+        for _ in range(20):
+            sp.call(["xdotool", "key", "ctrl+w"])  # closes the open tab correctly
+    # else:
+    #    sp.call(["xdotool", "getactivewindow", "windowminimize"])  # minimizes the focused windo
+
+    # close any error messages:
+    sp.call(["xdotool", "search", "--name", "Error", "windowclose"])
+    # sp.call(["xdotool", "key", "ctrl+alt+d"])  # doesnt work at all
+    # sp.call(["xdotool", "keydown", "--delay", "100", "ctrl+alt+d"])
+    # sp.call(["xdotool", "keyup", "ctrl+alt+d"])
+    sp.call(["xdg-open", txt])
+
+
+main()

bin/mtag/vidchk.py

@@ -0,0 +1,131 @@
+#!/usr/bin/env python3
+
+import json
+import re
+import os
+import sys
+import subprocess as sp
+
+try:
+    from copyparty.util import fsenc
+except:
+
+    def fsenc(p):
+        return p.encode("utf-8")
+
+
+_ = r"""
+inspects video files for errors and such
+plus stores a bunch of metadata to filename.ff.json
+
+usage:
+  -mtp vidchk=t600,ay,p,bin/mtag/vidchk.py
+
+explained:
+t600: timeout 10min
+  ay: only process files which contain audio (including video with audio)
+   p: set priority 1 (lowest priority after initial ffprobe/mutagen for base tags),
+       makes copyparty feed base tags into this script as json
+
+if you wanna use this script standalone / separately from copyparty,
+provide the video resolution on stdin as json:  {"res":"1920x1080"}
+"""
+
+
+FAST = True  # parse entire file at container level
+# FAST = False  # fully decode audio and video streams
+
+
+# warnings to ignore
+harmless = re.compile(
+    r"Unsupported codec with id |Could not find codec parameters.*Attachment:|analyzeduration"
+    + r"|timescale not set"
+)
+
+
+def wfilter(lines):
+    return [x for x in lines if x.strip() and not harmless.search(x)]
+
+
+def errchk(so, se, rc, dbg):
+    if dbg:
+        with open(dbg, "wb") as f:
+            f.write(b"so:\n" + so + b"\nse:\n" + se + b"\n")
+
+    if rc:
+        err = (so + se).decode("utf-8", "replace").split("\n", 1)
+        err = wfilter(err) or err
+        return f"ERROR {rc}: {err[0]}"
+
+    if se:
+        err = se.decode("utf-8", "replace").split("\n", 1)
+        err = wfilter(err)
+        if err:
+            return f"Warning: {err[0]}"
+
+    return None
+
+
+def main():
+    fp = sys.argv[1]
+    zb = sys.stdin.buffer.read()
+    zs = zb.decode("utf-8", "replace")
+    md = json.loads(zs)
+
+    fdir = os.path.dirname(os.path.realpath(fp))
+    flag = os.path.join(fdir, ".processed")
+    if os.path.exists(flag):
+        return "already processed"
+
+    try:
+        w, h = [int(x) for x in md["res"].split("x")]
+        if not w + h:
+            raise Exception()
+    except:
+        return "could not determine resolution"
+
+    # grab streams/format metadata + 2 seconds of frames at the start and end
+    zs = "ffprobe -hide_banner -v warning -of json -show_streams -show_format -show_packets -show_data_hash crc32 -read_intervals %+2,999999%+2"
+    cmd = zs.encode("ascii").split(b" ") + [fsenc(fp)]
+    p = sp.Popen(cmd, stdout=sp.PIPE, stderr=sp.PIPE)
+    so, se = p.communicate()
+
+    # spaces to tabs, drops filesize from 69k to 48k
+    so = b"\n".join(
+        [
+            b"\t" * int((len(x) - len(x.lstrip())) / 4) + x.lstrip()
+            for x in (so or b"").split(b"\n")
+        ]
+    )
+    with open(fsenc(f"{fp}.ff.json"), "wb") as f:
+        f.write(so)
+
+    err = errchk(so, se, p.returncode, f"{fp}.vidchk")
+    if err:
+        return err
+
+    if max(w, h) < 1280 and min(w, h) < 720:
+        return "resolution too small"
+
+    zs = (
+        "ffmpeg -y -hide_banner -nostdin -v warning"
+        + " -err_detect +crccheck+bitstream+buffer+careful+compliant+aggressive+explode"
+        + " -xerror -i"
+    )
+
+    cmd = zs.encode("ascii").split(b" ") + [fsenc(fp)]
+
+    if FAST:
+        zs = "-c copy -f null -"
+    else:
+        zs = "-vcodec rawvideo -acodec pcm_s16le -f null -"
+
+    cmd += zs.encode("ascii").split(b" ")
+
+    p = sp.Popen(cmd, stdout=sp.PIPE, stderr=sp.PIPE)
+    so, se = p.communicate()
+    return errchk(so, se, p.returncode, f"{fp}.vidchk")
+
+
+if __name__ == "__main__":
+    print(main() or "ok")

bin/mtag/wget.py

@@ -0,0 +1,85 @@
+#!/usr/bin/env python3
+
+"""
+use copyparty as a file downloader by POSTing URLs as
+application/x-www-form-urlencoded (for example using the
+message/pager function on the website)
+
+example copyparty config to use this:
+  --urlform save,get -vsrv/wget:wget:rwmd,ed:c,e2ts,mtp=title=ebin,t300,ad,bin/mtag/wget.py
+
+explained:
+  for realpath srv/wget (served at /wget) with read-write-modify-delete for ed,
+  enable file analysis on upload (e2ts),
+  use mtp plugin "bin/mtag/wget.py" to provide metadata tag "title",
+  do this on all uploads with the file extension "bin",
+  t300 = 300 seconds timeout for each dwonload,
+  ad = parse file regardless if FFmpeg thinks it is audio or not
+
+PS: this requires e2ts to be functional,
+  meaning you need to do at least one of these:
+   * apt install ffmpeg
+   * pip3 install mutagen
+"""
+
+
+import os
+import sys
+import subprocess as sp
+from urllib.parse import unquote_to_bytes as unquote
+
+
+def main():
+    fp = os.path.abspath(sys.argv[1])
+    fdir = os.path.dirname(fp)
+    fname = os.path.basename(fp)
+    if not fname.startswith("put-") or not fname.endswith(".bin"):
+        raise Exception("not a post file")
+
+    buf = b""
+    with open(fp, "rb") as f:
+        while True:
+            b = f.read(4096)
+            buf += b
+            if len(buf) > 4096:
+                raise Exception("too big")
+
+            if not b:
+                break
+
+    if not buf:
+        raise Exception("file is empty")
+
+    buf = unquote(buf.replace(b"+", b" "))
+    url = buf.decode("utf-8")
+
+    if not url.startswith("msg="):
+        raise Exception("does not start with msg=")
+
+    url = url[4:]
+    if "://" not in url:
+        url = "https://" + url
+
+    os.chdir(fdir)
+
+    name = url.split("?")[0].split("/")[-1]
+    tfn = "-- DOWNLOADING " + name
+    open(tfn, "wb").close()
+
+    cmd = ["wget", "--trust-server-names", "--", url]
+
+    try:
+        sp.check_call(cmd)
+
+        # OPTIONAL:
+        #   on success, delete the .bin file which contains the URL
+        os.unlink(fp)
+    except:
+        open("-- FAILED TO DONWLOAD " + name, "wb").close()
+
+    os.unlink(tfn)
+    print(url)
+
+
+if __name__ == "__main__":
+    main()

bin/mtag/yt-ipr.py

@@ -0,0 +1,198 @@
+#!/usr/bin/env python
+
+import re
+import os
+import sys
+import gzip
+import json
+import base64
+import string
+import urllib.request
+from datetime import datetime
+
+"""
+youtube initial player response
+
+it's probably best to use this through a config file; see res/yt-ipr.conf
+
+but if you want to use plain arguments instead then:
+  -v srv/ytm:ytm:w:rw,ed
+       :c,e2ts,e2dsa
+       :c,sz=16k-1m:c,maxn=10,300:c,rotf=%Y-%m/%d-%H
+       :c,mtp=yt-id,yt-title,yt-author,yt-channel,yt-views,yt-private,yt-manifest,yt-expires=bin/mtag/yt-ipr.py
+       :c,mte=yt-id,yt-title,yt-author,yt-channel,yt-views,yt-private,yt-manifest,yt-expires
+
+see res/yt-ipr.user.js for the example userscript to go with this
+"""
+
+
+def main():
+    try:
+        with gzip.open(sys.argv[1], "rt", encoding="utf-8", errors="replace") as f:
+            txt = f.read()
+    except:
+        with open(sys.argv[1], "r", encoding="utf-8", errors="replace") as f:
+            txt = f.read()
+
+    txt = "{" + txt.split("{", 1)[1]
+
+    try:
+        pd = json.loads(txt)
+    except json.decoder.JSONDecodeError as ex:
+        pd = json.loads(txt[: ex.pos])
+
+    # print(json.dumps(pd, indent=2))
+
+    if "videoDetails" in pd:
+        parse_youtube(pd)
+    else:
+        parse_freg(pd)
+
+
+def get_expiration(url):
+    et = re.search(r"[?&]expire=([0-9]+)", url).group(1)
+    et = datetime.utcfromtimestamp(int(et))
+    return et.strftime("%Y-%m-%d, %H:%M")
+
+
+def parse_youtube(pd):
+    vd = pd["videoDetails"]
+    sd = pd["streamingData"]
+
+    et = sd["adaptiveFormats"][0]["url"]
+    et = get_expiration(et)
+
+    mf = []
+    if "dashManifestUrl" in sd:
+        mf.append("dash")
+    if "hlsManifestUrl" in sd:
+        mf.append("hls")
+
+    r = {
+        "yt-id": vd["videoId"],
+        "yt-title": vd["title"],
+        "yt-author": vd["author"],
+        "yt-channel": vd["channelId"],
+        "yt-views": vd["viewCount"],
+        "yt-private": vd["isPrivate"],
+        # "yt-expires": sd["expiresInSeconds"],
+        "yt-manifest": ",".join(mf),
+        "yt-expires": et,
+    }
+    print(json.dumps(r))
+
+    freg_conv(pd)
+
+
+def parse_freg(pd):
+    md = pd["metadata"]
+    r = {
+        "yt-id": md["id"],
+        "yt-title": md["title"],
+        "yt-author": md["channelName"],
+        "yt-channel": md["channelURL"].strip("/").split("/")[-1],
+        "yt-expires": get_expiration(list(pd["video"].values())[0]),
+    }
+    print(json.dumps(r))
+
+
+def freg_conv(pd):
+    # based on getURLs.js v1.5 (2021-08-07)
+    # fmt: off
+    priority = {
+        "video": [
+            337, 315, 266, 138,  # 2160p60
+            313, 336,  # 2160p
+            308,  # 1440p60
+            271, 264,  # 1440p
+            335, 303, 299,  # 1080p60
+            248, 169, 137,  # 1080p
+            334, 302, 298,  # 720p60
+            247, 136  # 720p
+        ],
+        "audio": [
+            251, 141, 171, 140, 250, 249, 139
+        ]
+    }
+
+    vid_id = pd["videoDetails"]["videoId"]
+    chan_id = pd["videoDetails"]["channelId"]
+
+    try:
+        thumb_url = pd["microformat"]["playerMicroformatRenderer"]["thumbnail"]["thumbnails"][0]["url"]
+        start_ts = pd["microformat"]["playerMicroformatRenderer"]["liveBroadcastDetails"]["startTimestamp"]
+    except:
+        thumb_url = f"https://img.youtube.com/vi/{vid_id}/maxresdefault.jpg"
+        start_ts = ""
+
+    # fmt: on
+
+    metadata = {
+        "title": pd["videoDetails"]["title"],
+        "id": vid_id,
+        "channelName": pd["videoDetails"]["author"],
+        "channelURL": "https://www.youtube.com/channel/" + chan_id,
+        "description": pd["videoDetails"]["shortDescription"],
+        "thumbnailUrl": thumb_url,
+        "startTimestamp": start_ts,
+    }
+
+    if [x for x in vid_id if x not in string.ascii_letters + string.digits + "_-"]:
+        print(f"malicious json", file=sys.stderr)
+        return
+
+    basepath = os.path.dirname(sys.argv[1])
+
+    thumb_fn = f"{basepath}/{vid_id}.jpg"
+    tmp_fn = f"{thumb_fn}.{os.getpid()}"
+    if not os.path.exists(thumb_fn) and (
+        thumb_url.startswith("https://img.youtube.com/vi/")
+        or thumb_url.startswith("https://i.ytimg.com/vi/")
+    ):
+        try:
+            with urllib.request.urlopen(thumb_url) as fi:
+                with open(tmp_fn, "wb") as fo:
+                    fo.write(fi.read())
+
+            os.rename(tmp_fn, thumb_fn)
+        except:
+            if os.path.exists(tmp_fn):
+                os.unlink(tmp_fn)
+
+    try:
+        with open(thumb_fn, "rb") as f:
+            thumb = base64.b64encode(f.read()).decode("ascii")
+    except:
+        thumb = "/9j/4AAQSkZJRgABAQEASABIAAD/2wBDAAMCAgICAgMCAgIDAwMDBAYEBAQEBAgGBgUGCQgKCgkICQkKDA8MCgsOCwkJDRENDg8QEBEQCgwSExIQEw8QEBD/yQALCAABAAEBAREA/8wABgAQEAX/2gAIAQEAAD8A0s8g/9k="
+
+    metadata["thumbnail"] = "data:image/jpeg;base64," + thumb
+
+    ret = {
+        "metadata": metadata,
+        "version": "1.5",
+        "createTime": datetime.utcnow().strftime("%Y-%m-%dT%H:%M:%SZ"),
+    }
+
+    for stream, itags in priority.items():
+        for itag in itags:
+            url = None
+            for afmt in pd["streamingData"]["adaptiveFormats"]:
+                if itag == afmt["itag"]:
+                    url = afmt["url"]
+                    break
+
+            if url:
+                ret[stream] = {itag: url}
+                break
+
+    fn = f"{basepath}/{vid_id}.urls.json"
+    with open(fn, "w", encoding="utf-8", errors="replace") as f:
+        f.write(json.dumps(ret, indent=4))
+
+
+if __name__ == "__main__":
+    try:
+        main()
+    except:
+        # raise
+        pass

bin/partyfuse.py

@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 from __future__ import print_function, unicode_literals
 
-"""copyparty-fuse: remote copyparty as a local filesystem"""
+"""partyfuse: remote copyparty as a local filesystem"""
 __author__ = "ed <copyparty@ocv.me>"
 __copyright__ = 2019
 __license__ = "MIT"
@@ -12,13 +12,20 @@ __url__ = "https://github.com/9001/copyparty/"
 mount a copyparty server (local or remote) as a filesystem
 
 usage:
-  python copyparty-fuse.py ./music http://192.168.1.69:3923/
+  python partyfuse.py http://192.168.1.69:3923/  ./music
 
 dependencies:
   python3 -m pip install --user fusepy
   + on Linux: sudo apk add fuse
   + on Macos: https://osxfuse.github.io/
   + on Windows: https://github.com/billziss-gh/winfsp/releases/latest
+
+note:
+  you probably want to run this on windows clients:
+  https://github.com/9001/copyparty/blob/hovudstraum/contrib/explorer-nothumbs-nofoldertypes.reg
+
+get server cert:
+  awk '/-BEGIN CERTIFICATE-/ {a=1} a; /-END CERTIFICATE-/{exit}' <(openssl s_client -connect 127.0.0.1:3923 </dev/null 2>/dev/null) >cert.pem
 """
 
 
@@ -26,24 +33,35 @@ import re
 import os
 import sys
 import time
+import json
 import stat
 import errno
 import struct
+import codecs
 import builtins
 import platform
+import argparse
 import threading
 import traceback
 import http.client  # py2: httplib
 import urllib.parse
+import calendar
 from datetime import datetime
 from urllib.parse import quote_from_bytes as quote
-
-
-DEBUG = False  # ctrl-f this to configure logging
-
+from urllib.parse import unquote_to_bytes as unquote
 
 WINDOWS = sys.platform == "win32"
 MACOS = platform.system() == "Darwin"
+info = log = dbg = None
+
+
+print(
+    "{} v{} @ {}".format(
+        platform.python_implementation(),
+        ".".join([str(x) for x in sys.version_info]),
+        sys.executable,
+    )
+)
 
 
 try:
@@ -54,14 +72,14 @@ except:
     elif MACOS:
         libfuse = "install https://osxfuse.github.io/"
     else:
-        libfuse = "apt install libfuse\n    modprobe fuse"
+        libfuse = "apt install libfuse3-3\n    modprobe fuse"
 
-    print(
-        "\n  could not import fuse; these may help:"
-        + "\n    python3 -m pip install --user fusepy\n    "
-        + libfuse
-        + "\n"
-    )
+    m = """\033[33m
+  could not import fuse; these may help:
+    {} -m pip install --user fusepy
+    {}
+\033[0m"""
+    print(m.format(sys.executable, libfuse))
     raise
 
 
@@ -97,17 +115,58 @@ def rice_tid():
 
 
 def fancy_log(msg):
-    print("{} {}\n".format(rice_tid(), msg), end="")
+    print("{:10.6f} {} {}\n".format(time.time() % 900, rice_tid(), msg), end="")
 
 
 def null_log(msg):
     pass
 
 
+def hexler(binary):
+    return binary.replace("\r", "\\r").replace("\n", "\\n")
+    return " ".join(["{}\033[36m{:02x}\033[0m".format(b, ord(b)) for b in binary])
+    return " ".join(map(lambda b: format(ord(b), "02x"), binary))
+
+
+def register_wtf8():
+    def wtf8_enc(text):
+        return str(text).encode("utf-8", "surrogateescape"), len(text)
+
+    def wtf8_dec(binary):
+        return bytes(binary).decode("utf-8", "surrogateescape"), len(binary)
+
+    def wtf8_search(encoding_name):
+        return codecs.CodecInfo(wtf8_enc, wtf8_dec, name="wtf-8")
+
+    codecs.register(wtf8_search)
+
+
+bad_good = {}
+good_bad = {}
+
+
+def enwin(txt):
+    return "".join([bad_good.get(x, x) for x in txt])
+
+    for bad, good in bad_good.items():
+        txt = txt.replace(bad, good)
+
+    return txt
+
+
+def dewin(txt):
+    return "".join([good_bad.get(x, x) for x in txt])
+
+    for bad, good in bad_good.items():
+        txt = txt.replace(good, bad)
+
+    return txt
+
+
 class RecentLog(object):
     def __init__(self):
         self.mtx = threading.Lock()
-        self.f = None  # open("copyparty-fuse.log", "wb")
+        self.f = None  # open("partyfuse.log", "wb")
         self.q = []
 
         thr = threading.Thread(target=self.printer)
@@ -115,7 +174,7 @@ class RecentLog(object):
         thr.start()
 
     def put(self, msg):
-        msg = "{} {}\n".format(rice_tid(), msg)
+        msg = "{:10.6f} {} {}\n".format(time.time() % 900, rice_tid(), msg)
         if self.f:
             fmsg = " ".join([datetime.utcnow().strftime("%H%M%S.%f"), str(msg)])
             self.f.write(fmsg.encode("utf-8"))
@@ -138,25 +197,9 @@ class RecentLog(object):
             print("".join(q), end="")
 
 
-if DEBUG:
-    # debug=on,
-    #   windows terminals are slow (cmd.exe, mintty)
-    #   otoh fancy_log beats RecentLog on linux
-    logger = RecentLog().put if WINDOWS else fancy_log
-
-    info = logger
-    log = logger
-    dbg = logger
-else:
-    # debug=off, speed is dontcare
-    info = fancy_log
-    log = null_log
-    dbg = null_log
-
-
-# [windows/cmd/cpy3]  python dev\copyparty\bin\copyparty-fuse.py q: http://192.168.1.159:1234/
-# [windows/cmd/msys2] C:\msys64\mingw64\bin\python3 dev\copyparty\bin\copyparty-fuse.py q: http://192.168.1.159:1234/
-# [windows/mty/msys2] /mingw64/bin/python3 /c/Users/ed/dev/copyparty/bin/copyparty-fuse.py q: http://192.168.1.159:1234/
+# [windows/cmd/cpy3]  python dev\copyparty\bin\partyfuse.py q: http://192.168.1.159:1234/
+# [windows/cmd/msys2] C:\msys64\mingw64\bin\python3 dev\copyparty\bin\partyfuse.py q: http://192.168.1.159:1234/
+# [windows/mty/msys2] /mingw64/bin/python3 /c/Users/ed/dev/copyparty/bin/partyfuse.py q: http://192.168.1.159:1234/
 #
 # [windows] find /q/music/albums/Phant*24bit -printf '%s %p\n' | sort -n | tail -n 8 | sed -r 's/^[0-9]+ //' | while IFS= read -r x; do dd if="$x" of=/dev/null bs=4k count=8192 & done
 # [alpine]  ll t; for x in t/2020_0724_16{2,3}*; do dd if="$x" of=/dev/null bs=4k count=10240 & done
@@ -183,6 +226,8 @@ def html_dec(txt):
         txt.replace("&lt;", "<")
         .replace("&gt;", ">")
         .replace("&quot;", '"')
+        .replace("&#13;", "\r")
+        .replace("&#10;", "\n")
         .replace("&amp;", "&")
     )
 
@@ -195,10 +240,11 @@ class CacheNode(object):
 
 
 class Gateway(object):
-    def __init__(self, base_url):
-        self.base_url = base_url
+    def __init__(self, ar):
+        self.base_url = ar.base_url
+        self.password = ar.a
 
-        ui = urllib.parse.urlparse(base_url)
+        ui = urllib.parse.urlparse(self.base_url)
         self.web_root = ui.path.strip("/")
         try:
             self.web_host, self.web_port = ui.netloc.split(":")
@@ -208,15 +254,25 @@ class Gateway(object):
             if ui.scheme == "http":
                 self.web_port = 80
             elif ui.scheme == "https":
-                raise Exception("todo")
+                self.web_port = 443
             else:
                 raise Exception("bad url?")
 
+        self.ssl_context = None
+        self.use_tls = ui.scheme.lower() == "https"
+        if self.use_tls:
+            import ssl
+
+            if ar.td:
+                self.ssl_context = ssl._create_unverified_context()
+            elif ar.te:
+                self.ssl_context = ssl.SSLContext(ssl.PROTOCOL_TLS)
+                self.ssl_context.load_verify_locations(ar.te)
+
         self.conns = {}
 
     def quotep(self, path):
-        # TODO: mojibake support
-        path = path.encode("utf-8", "ignore")
+        path = path.encode("wtf-8")
         return quote(path, safe="/")
 
     def getconn(self, tid=None):
@@ -226,7 +282,15 @@ class Gateway(object):
         except:
             info("new conn [{}] [{}]".format(self.web_host, self.web_port))
 
-            conn = http.client.HTTPConnection(self.web_host, self.web_port, timeout=260)
+            args = {}
+            if not self.use_tls:
+                C = http.client.HTTPConnection
+            else:
+                C = http.client.HTTPSConnection
+                if self.ssl_context:
+                    args = {"context": self.ssl_context}
+
+            conn = C(self.web_host, self.web_port, timeout=260, **args)
 
             self.conns[tid] = conn
             return conn
@@ -239,45 +303,76 @@ class Gateway(object):
         except:
             pass
 
-    def sendreq(self, *args, **kwargs):
+    def sendreq(self, meth, path, headers, **kwargs):
         tid = get_tid()
+        if self.password:
+            headers["Cookie"] = "=".join(["cppwd", self.password])
+
         try:
             c = self.getconn(tid)
-            c.request(*list(args), **kwargs)
+            c.request(meth, path, headers=headers, **kwargs)
             return c.getresponse()
         except:
-            self.closeconn(tid)
+            dbg("bad conn")
+
+        self.closeconn(tid)
+        try:
             c = self.getconn(tid)
-            c.request(*list(args), **kwargs)
+            c.request(meth, path, headers=headers, **kwargs)
             return c.getresponse()
+        except:
+            info("http connection failed:\n" + traceback.format_exc())
+            if self.use_tls and not self.ssl_context:
+                import ssl
+
+                cert = ssl.get_server_certificate((self.web_host, self.web_port))
+                info("server certificate probably not trusted:\n" + cert)
+
+            raise
 
     def listdir(self, path):
-        web_path = self.quotep("/" + "/".join([self.web_root, path])) + "?dots"
-        r = self.sendreq("GET", web_path)
+        if bad_good:
+            path = dewin(path)
+
+        web_path = self.quotep("/" + "/".join([self.web_root, path])) + "?dots&ls"
+        r = self.sendreq("GET", web_path, {})
         if r.status != 200:
             self.closeconn()
-            raise Exception(
+            log(
                 "http error {} reading dir {} in {}".format(
                     r.status, web_path, rice_tid()
                 )
             )
+            raise FuseOSError(errno.ENOENT)
+
+        ctype = r.getheader("Content-Type", "")
+        if ctype == "application/json":
+            parser = self.parse_jls
+        elif ctype.startswith("text/html"):
+            parser = self.parse_html
+        else:
+            log("listdir on file: {}".format(path))
+            raise FuseOSError(errno.ENOENT)
 
         try:
-            return self.parse_html(r)
+            return parser(r)
         except:
-            traceback.print_exc()
+            info(repr(path) + "\n" + traceback.format_exc())
             raise
 
     def download_file_range(self, path, ofs1, ofs2):
+        if bad_good:
+            path = dewin(path)
+
         web_path = self.quotep("/" + "/".join([self.web_root, path])) + "?raw"
         hdr_range = "bytes={}-{}".format(ofs1, ofs2 - 1)
         info(
             "DL {:4.0f}K\033[36m{:>9}-{:<9}\033[0m{}".format(
-                (ofs2 - ofs1) / 1024.0, ofs1, ofs2 - 1, path
+                (ofs2 - ofs1) / 1024.0, ofs1, ofs2 - 1, hexler(path)
             )
         )
 
-        r = self.sendreq("GET", web_path, headers={"Range": hdr_range})
+        r = self.sendreq("GET", web_path, {"Range": hdr_range})
         if r.status != http.client.PARTIAL_CONTENT:
             self.closeconn()
             raise Exception(
@@ -288,11 +383,35 @@ class Gateway(object):
 
         return r.read()
 
+    def parse_jls(self, datasrc):
+        rsp = b""
+        while True:
+            buf = datasrc.read(1024 * 32)
+            if not buf:
+                break
+
+            rsp += buf
+
+        rsp = json.loads(rsp.decode("utf-8"))
+        ret = []
+        for statfun, nodes in [
+            [self.stat_dir, rsp["dirs"]],
+            [self.stat_file, rsp["files"]],
+        ]:
+            for n in nodes:
+                fname = unquote(n["href"].split("?")[0]).rstrip(b"/").decode("wtf-8")
+                if bad_good:
+                    fname = enwin(fname)
+
+                ret.append([fname, statfun(n["ts"], n["sz"]), 0])
+
+        return ret
+
     def parse_html(self, datasrc):
         ret = []
         remainder = b""
         ptn = re.compile(
-            r"^<tr><td>(-|DIR)</td><td><a [^>]+>([^<]+)</a></td><td>([^<]+)</td><td>([^<]+)</td></tr>$"
+            r'^<tr><td>(-|DIR|<a [^<]+</a>)</td><td><a[^>]* href="([^"]+)"[^>]*>([^<]+)</a></td><td>([^<]+)</td><td>[^<]+</td><td>([^<]+)</td></tr>$'
         )
 
         while True:
@@ -314,18 +433,23 @@ class Gateway(object):
                     # print(line)
                     continue
 
-                ftype, fname, fsize, fdate = m.groups()
-                fname = html_dec(fname)
+                ftype, furl, fname, fsize, fdate = m.groups()
+                fname = furl.rstrip("/").split("/")[-1]
+                fname = unquote(fname)
+                fname = fname.decode("wtf-8")
+                if bad_good:
+                    fname = enwin(fname)
+
                 sz = 1
                 ts = 60 * 60 * 24 * 2
                 try:
                     sz = int(fsize)
-                    ts = datetime.strptime(fdate, "%Y-%m-%d %H:%M:%S").timestamp()
+                    ts = calendar.timegm(time.strptime(fdate, "%Y-%m-%d %H:%M:%S"))
                 except:
                     info("bad HTML or OS [{}] [{}]".format(fdate, fsize))
                     # python cannot strptime(1959-01-01) on windows
 
-                if ftype == "-":
+                if ftype != "DIR":
                     ret.append([fname, self.stat_file(ts, sz), 0])
                 else:
                     ret.append([fname, self.stat_dir(ts, sz), 0])
@@ -358,11 +482,11 @@ class Gateway(object):
 
 
 class CPPF(Operations):
-    def __init__(self, base_url, dircache, filecache):
-        self.gw = Gateway(base_url)
+    def __init__(self, ar):
+        self.gw = Gateway(ar)
         self.junk_fh_ctr = 3
-        self.n_dircache = dircache
-        self.n_filecache = filecache
+        self.n_dircache = ar.cd
+        self.n_filecache = ar.cf
 
         self.dircache = []
         self.dircache_mtx = threading.Lock()
@@ -379,7 +503,11 @@ class CPPF(Operations):
                 cache_path, cache1 = cn.tag
                 cache2 = cache1 + len(cn.data)
                 msg += "\n{:<2} {:>7} {:>10}:{:<9} {}".format(
-                    n, len(cn.data), cache1, cache2, cache_path
+                    n,
+                    len(cn.data),
+                    cache1,
+                    cache2,
+                    cache_path.replace("\r", "\\r").replace("\n", "\\n"),
                 )
         return msg
 
@@ -574,8 +702,18 @@ class CPPF(Operations):
 
         else:
             if get2 - get1 <= 1024 * 1024:
-                h_ofs = get1 - 256 * 1024
-                h_end = get2 + 1024 * 1024
+                # unless the request is for the last n bytes of the file,
+                # grow the start to cache some stuff around the range
+                if get2 < file_sz - 1:
+                    h_ofs = get1 - 1024 * 256
+                else:
+                    h_ofs = get1 - 1024 * 32
+
+                # likewise grow the end unless start is 0
+                if get1 > 0:
+                    h_end = get2 + 1024 * 1024
+                else:
+                    h_end = get2 + 1024 * 64
             else:
                 # big enough, doesn't need pads
                 h_ofs = get1
@@ -610,7 +748,7 @@ class CPPF(Operations):
 
     def _readdir(self, path, fh=None):
         path = path.strip("/")
-        log("readdir [{}] [{}]".format(path, fh))
+        log("readdir [{}] [{}]".format(hexler(path), fh))
 
         ret = self.gw.listdir(path)
         if not self.n_dircache:
@@ -621,6 +759,7 @@ class CPPF(Operations):
             self.dircache.append(cn)
             self.clean_dircache()
 
+        # import pprint; pprint.pprint(ret)
         return ret
 
     def readdir(self, path, fh=None):
@@ -637,7 +776,11 @@ class CPPF(Operations):
         path = path.strip("/")
         ofs2 = offset + length
         file_sz = self.getattr(path)["st_size"]
-        log("read {} |{}| {}:{} max {}".format(path, length, offset, ofs2, file_sz))
+        log(
+            "read {} |{}| {}:{} max {}".format(
+                hexler(path), length, offset, ofs2, file_sz
+            )
+        )
         if ofs2 > file_sz:
             ofs2 = file_sz
             log("truncate to |{}| :{}".format(ofs2 - offset, ofs2))
@@ -676,7 +819,9 @@ class CPPF(Operations):
         return ret
 
     def getattr(self, path, fh=None):
-        log("getattr [{}]".format(path))
+        log("getattr [{}]".format(hexler(path)))
+        if WINDOWS:
+            path = enwin(path)  # windows occasionally decodes f0xx to xx
 
         path = path.strip("/")
         try:
@@ -699,11 +844,24 @@ class CPPF(Operations):
             dents = self._readdir(dirpath)
 
         for cache_name, cache_stat, _ in dents:
+            # if "qw" in cache_name and "qw" in fname:
+            #     info(
+            #         "cmp\n  [{}]\n  [{}]\n\n{}\n".format(
+            #             hexler(cache_name),
+            #             hexler(fname),
+            #             "\n".join(traceback.format_stack()[:-1]),
+            #         )
+            #     )
+
             if cache_name == fname:
                 # dbg("=" + repr(cache_stat))
                 return cache_stat
 
-        info("=ENOENT ({})".format(path))
+        fun = info
+        if MACOS and path.split("/")[-1].startswith("._"):
+            fun = dbg
+
+        fun("=ENOENT ({})".format(hexler(path)))
         raise FuseOSError(errno.ENOENT)
 
     access = None
@@ -773,24 +931,24 @@ class CPPF(Operations):
                 raise FuseOSError(errno.ENOENT)
 
         def open(self, path, flags):
-            dbg("open [{}] [{}]".format(path, flags))
+            dbg("open [{}] [{}]".format(hexler(path), flags))
             return self._open(path)
 
         def opendir(self, path):
-            dbg("opendir [{}]".format(path))
+            dbg("opendir [{}]".format(hexler(path)))
             return self._open(path)
 
         def flush(self, path, fh):
-            dbg("flush [{}] [{}]".format(path, fh))
+            dbg("flush [{}] [{}]".format(hexler(path), fh))
 
         def release(self, ino, fi):
-            dbg("release [{}] [{}]".format(ino, fi))
+            dbg("release [{}] [{}]".format(hexler(ino), fi))
 
         def releasedir(self, ino, fi):
-            dbg("releasedir [{}] [{}]".format(ino, fi))
+            dbg("releasedir [{}] [{}]".format(hexler(ino), fi))
 
         def access(self, path, mode):
-            dbg("access [{}] [{}]".format(path, mode))
+            dbg("access [{}] [{}]".format(hexler(path), mode))
             try:
                 x = self.getattr(path)
                 if x["st_mode"] <= 0:
@@ -799,41 +957,90 @@ class CPPF(Operations):
                 raise FuseOSError(errno.ENOENT)
 
 
+class TheArgparseFormatter(
+    argparse.RawTextHelpFormatter, argparse.ArgumentDefaultsHelpFormatter
+):
+    pass
+
+
 def main():
+    global info, log, dbg
+    time.strptime("19970815", "%Y%m%d")  # python#7980
+
     # filecache helps for reads that are ~64k or smaller;
     #   linux generally does 128k so the cache is a slowdown,
     #   windows likes to use 4k and 64k so cache is required,
     #   value is numChunks (1~3M each) to keep in the cache
-    nf = 24 if WINDOWS or MACOS else 0
+    nf = 24
 
     # dircache is always a boost,
     #   only want to disable it for tests etc,
     #   value is numSec until an entry goes stale
     nd = 1
 
-    try:
-        local, remote = sys.argv[1:3]
-        filecache = nf if len(sys.argv) <= 3 else int(sys.argv[3])
-        dircache = nd if len(sys.argv) <= 4 else float(sys.argv[4])
-    except:
-        where = "local directory"
-        if WINDOWS:
-            where += " or DRIVE:"
+    where = "local directory"
+    if WINDOWS:
+        where += " or DRIVE:"
 
-        print("need arg 1: " + where)
-        print("need arg 2: root url")
-        print("optional 3: num files in filecache ({})".format(nf))
-        print("optional 4: num seconds / dircache ({})".format(nd))
-        print()
-        print("example:")
-        print("  copyparty-fuse.py ./music http://192.168.1.69:3923/music/")
-        if WINDOWS:
-            print("  copyparty-fuse.py M: http://192.168.1.69:3923/music/")
+    ex_pre = "\n  " + os.path.basename(__file__) + "  "
+    examples = ["http://192.168.1.69:3923/music/  ./music"]
+    if WINDOWS:
+        examples.append("http://192.168.1.69:3923/music/  M:")
 
-        return
+    ap = argparse.ArgumentParser(
+        formatter_class=TheArgparseFormatter,
+        epilog="example:" + ex_pre + ex_pre.join(examples),
+    )
+    ap.add_argument(
+        "-cd", metavar="NUM_SECONDS", type=float, default=nd, help="directory cache"
+    )
+    ap.add_argument(
+        "-cf", metavar="NUM_BLOCKS", type=int, default=nf, help="file cache"
+    )
+    ap.add_argument("-a", metavar="PASSWORD", help="password")
+    ap.add_argument("-d", action="store_true", help="enable debug")
+    ap.add_argument("-te", metavar="PEM_FILE", help="certificate to expect/verify")
+    ap.add_argument("-td", action="store_true", help="disable certificate check")
+    ap.add_argument("base_url", type=str, help="remote copyparty URL to mount")
+    ap.add_argument("local_path", type=str, help=where + " to mount it on")
+    ar = ap.parse_args()
+
+    if ar.d:
+        # windows terminals are slow (cmd.exe, mintty)
+        # otoh fancy_log beats RecentLog on linux
+        logger = RecentLog().put if WINDOWS else fancy_log
+
+        info = logger
+        log = logger
+        dbg = logger
+    else:
+        # debug=off, speed is dontcare
+        info = fancy_log
+        log = null_log
+        dbg = null_log
+
+    if ar.a and ar.a.startswith("$"):
+        fn = ar.a[1:]
+        log("reading password from file [{}]".format(fn))
+        with open(fn, "rb") as f:
+            ar.a = f.read().decode("utf-8").strip()
 
     if WINDOWS:
-        os.system("")
+        os.system("rem")
+
+        for ch in '<>:"\\|?*':
+            # microsoft maps illegal characters to f0xx
+            # (e000 to f8ff is basic-plane private-use)
+            bad_good[ch] = chr(ord(ch) + 0xF000)
+
+        for n in range(0, 0x100):
+            # map surrogateescape to another private-use area
+            bad_good[chr(n + 0xDC00)] = chr(n + 0xF100)
+
+        for k, v in bad_good.items():
+            good_bad[v] = k
+
+    register_wtf8()
 
     try:
         with open("/etc/fuse.conf", "rb") as f:
@@ -845,7 +1052,7 @@ def main():
     if not MACOS:
         args["nonempty"] = True
 
-    FUSE(CPPF(remote, dircache, filecache), local, **args)
+    FUSE(CPPF(ar), ar.local_path, encoding="wtf-8", **args)
 
 
 if __name__ == "__main__":

bin/partyfuse2.py

@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 from __future__ import print_function, unicode_literals
 
-"""copyparty-fuseb: remote copyparty as a local filesystem"""
+"""partyfuse2: remote copyparty as a local filesystem"""
 __author__ = "ed <copyparty@ocv.me>"
 __copyright__ = 2020
 __license__ = "MIT"
@@ -11,14 +11,18 @@ import re
 import os
 import sys
 import time
+import json
 import stat
 import errno
 import struct
+import codecs
+import platform
 import threading
 import http.client  # py2: httplib
 import urllib.parse
 from datetime import datetime
 from urllib.parse import quote_from_bytes as quote
+from urllib.parse import unquote_to_bytes as unquote
 
 try:
     import fuse
@@ -28,9 +32,19 @@ try:
     if not hasattr(fuse, "__version__"):
         raise Exception("your fuse-python is way old")
 except:
-    print(
-        "\n  could not import fuse; these may help:\n    python3 -m pip install --user fuse-python\n    apt install libfuse\n    modprobe fuse\n"
-    )
+    if WINDOWS:
+        libfuse = "install https://github.com/billziss-gh/winfsp/releases/latest"
+    elif MACOS:
+        libfuse = "install https://osxfuse.github.io/"
+    else:
+        libfuse = "apt install libfuse\n    modprobe fuse"
+
+    m = """\033[33m
+  could not import fuse; these may help:
+    {} -m pip install --user fuse-python
+    {}
+\033[0m"""
+    print(m.format(sys.executable, libfuse))
     raise
 
 
@@ -38,18 +52,22 @@ except:
 mount a copyparty server (local or remote) as a filesystem
 
 usage:
-  python ./copyparty-fuseb.py -f -o allow_other,auto_unmount,nonempty,url=http://192.168.1.69:3923 /mnt/nas
+  python ./partyfuse2.py -f -o allow_other,auto_unmount,nonempty,pw=wark,url=http://192.168.1.69:3923 /mnt/nas
 
 dependencies:
   sudo apk add fuse-dev python3-dev
   python3 -m pip install --user fuse-python
 
-fork of copyparty-fuse.py based on fuse-python which
+fork of partyfuse.py based on fuse-python which
   appears to be more compliant than fusepy? since this works with samba
     (probably just my garbage code tbh)
 """
 
 
+WINDOWS = sys.platform == "win32"
+MACOS = platform.system() == "Darwin"
+
+
 def threadless_log(msg):
     print(msg + "\n", end="")
 
@@ -93,6 +111,41 @@ def html_dec(txt):
     )
 
 
+def register_wtf8():
+    def wtf8_enc(text):
+        return str(text).encode("utf-8", "surrogateescape"), len(text)
+
+    def wtf8_dec(binary):
+        return bytes(binary).decode("utf-8", "surrogateescape"), len(binary)
+
+    def wtf8_search(encoding_name):
+        return codecs.CodecInfo(wtf8_enc, wtf8_dec, name="wtf-8")
+
+    codecs.register(wtf8_search)
+
+
+bad_good = {}
+good_bad = {}
+
+
+def enwin(txt):
+    return "".join([bad_good.get(x, x) for x in txt])
+
+    for bad, good in bad_good.items():
+        txt = txt.replace(bad, good)
+
+    return txt
+
+
+def dewin(txt):
+    return "".join([good_bad.get(x, x) for x in txt])
+
+    for bad, good in bad_good.items():
+        txt = txt.replace(good, bad)
+
+    return txt
+
+
 class CacheNode(object):
     def __init__(self, tag, data):
         self.tag = tag
@@ -115,8 +168,9 @@ class Stat(fuse.Stat):
 
 
 class Gateway(object):
-    def __init__(self, base_url):
+    def __init__(self, base_url, pw):
         self.base_url = base_url
+        self.pw = pw
 
         ui = urllib.parse.urlparse(base_url)
         self.web_root = ui.path.strip("/")
@@ -135,8 +189,7 @@ class Gateway(object):
         self.conns = {}
 
     def quotep(self, path):
-        # TODO: mojibake support
-        path = path.encode("utf-8", "ignore")
+        path = path.encode("wtf-8")
         return quote(path, safe="/")
 
     def getconn(self, tid=None):
@@ -159,20 +212,29 @@ class Gateway(object):
         except:
             pass
 
-    def sendreq(self, *args, **kwargs):
+    def sendreq(self, *args, **ka):
         tid = get_tid()
+        if self.pw:
+            ck = "cppwd=" + self.pw
+            try:
+                ka["headers"]["Cookie"] = ck
+            except:
+                ka["headers"] = {"Cookie": ck}
         try:
             c = self.getconn(tid)
-            c.request(*list(args), **kwargs)
+            c.request(*list(args), **ka)
             return c.getresponse()
         except:
             self.closeconn(tid)
             c = self.getconn(tid)
-            c.request(*list(args), **kwargs)
+            c.request(*list(args), **ka)
             return c.getresponse()
 
     def listdir(self, path):
-        web_path = self.quotep("/" + "/".join([self.web_root, path])) + "?dots"
+        if bad_good:
+            path = dewin(path)
+
+        web_path = self.quotep("/" + "/".join([self.web_root, path])) + "?dots&ls"
         r = self.sendreq("GET", web_path)
         if r.status != 200:
             self.closeconn()
@@ -182,9 +244,12 @@ class Gateway(object):
                 )
             )
 
-        return self.parse_html(r)
+        return self.parse_jls(r)
 
     def download_file_range(self, path, ofs1, ofs2):
+        if bad_good:
+            path = dewin(path)
+
         web_path = self.quotep("/" + "/".join([self.web_root, path])) + "?raw"
         hdr_range = "bytes={}-{}".format(ofs1, ofs2 - 1)
         log("downloading {}".format(hdr_range))
@@ -200,40 +265,27 @@ class Gateway(object):
 
         return r.read()
 
-    def parse_html(self, datasrc):
-        ret = []
-        remainder = b""
-        ptn = re.compile(
-            r"^<tr><td>(-|DIR)</td><td><a [^>]+>([^<]+)</a></td><td>([^<]+)</td><td>([^<]+)</td></tr>$"
-        )
-
+    def parse_jls(self, datasrc):
+        rsp = b""
         while True:
-            buf = remainder + datasrc.read(4096)
-            # print('[{}]'.format(buf.decode('utf-8')))
+            buf = datasrc.read(1024 * 32)
             if not buf:
                 break
 
-            remainder = b""
-            endpos = buf.rfind(b"\n")
-            if endpos >= 0:
-                remainder = buf[endpos + 1 :]
-                buf = buf[:endpos]
+            rsp += buf
 
-            lines = buf.decode("utf-8").split("\n")
-            for line in lines:
-                m = ptn.match(line)
-                if not m:
-                    # print(line)
-                    continue
+        rsp = json.loads(rsp.decode("utf-8"))
+        ret = []
+        for statfun, nodes in [
+            [self.stat_dir, rsp["dirs"]],
+            [self.stat_file, rsp["files"]],
+        ]:
+            for n in nodes:
+                fname = unquote(n["href"].split("?")[0]).rstrip(b"/").decode("wtf-8")
+                if bad_good:
+                    fname = enwin(fname)
 
-                ftype, fname, fsize, fdate = m.groups()
-                fname = html_dec(fname)
-                ts = datetime.strptime(fdate, "%Y-%m-%d %H:%M:%S").timestamp()
-                sz = int(fsize)
-                if ftype == "-":
-                    ret.append([fname, self.stat_file(ts, sz), 0])
-                else:
-                    ret.append([fname, self.stat_dir(ts, sz), 0])
+                ret.append([fname, statfun(n["ts"], n["sz"]), 0])
 
         return ret
 
@@ -262,6 +314,7 @@ class CPPF(Fuse):
         Fuse.__init__(self, *args, **kwargs)
 
         self.url = None
+        self.pw = None
 
         self.dircache = []
         self.dircache_mtx = threading.Lock()
@@ -271,7 +324,7 @@ class CPPF(Fuse):
 
     def init2(self):
         # TODO figure out how python-fuse wanted this to go
-        self.gw = Gateway(self.url)  # .decode('utf-8'))
+        self.gw = Gateway(self.url, self.pw)  # .decode('utf-8'))
         info("up")
 
     def clean_dircache(self):
@@ -536,6 +589,8 @@ class CPPF(Fuse):
 
     def getattr(self, path):
         log("getattr [{}]".format(path))
+        if WINDOWS:
+            path = enwin(path)  # windows occasionally decodes f0xx to xx
 
         path = path.strip("/")
         try:
@@ -567,8 +622,26 @@ class CPPF(Fuse):
 
 
 def main():
+    time.strptime("19970815", "%Y%m%d")  # python#7980
+    register_wtf8()
+    if WINDOWS:
+        os.system("rem")
+
+        for ch in '<>:"\\|?*':
+            # microsoft maps illegal characters to f0xx
+            # (e000 to f8ff is basic-plane private-use)
+            bad_good[ch] = chr(ord(ch) + 0xF000)
+
+        for n in range(0, 0x100):
+            # map surrogateescape to another private-use area
+            bad_good[chr(n + 0xDC00)] = chr(n + 0xF100)
+
+        for k, v in bad_good.items():
+            good_bad[v] = k
+
     server = CPPF()
     server.parser.add_option(mountopt="url", metavar="BASE_URL", default=None)
+    server.parser.add_option(mountopt="pw", metavar="PASSWORD", default=None)
     server.parse(values=server, errex=1)
     if not server.url or not str(server.url).startswith("http"):
         print("\nerror:")
@@ -576,7 +649,7 @@ def main():
         print("  need argument: mount-path")
         print("example:")
         print(
-            "  ./copyparty-fuseb.py -f -o allow_other,auto_unmount,nonempty,url=http://192.168.1.69:3923 /mnt/nas"
+            "  ./partyfuse2.py -f -o allow_other,auto_unmount,nonempty,pw=wark,url=http://192.168.1.69:3923 /mnt/nas"
         )
         sys.exit(1)
 

bin/partyjournal.py

@@ -0,0 +1,177 @@
+#!/usr/bin/env python3
+
+"""
+partyjournal.py: chronological history of uploads
+2021-12-31, v0.1, ed <irc.rizon.net>, MIT-Licensed
+https://github.com/9001/copyparty/blob/hovudstraum/bin/partyjournal.py
+
+produces a chronological list of all uploads,
+by collecting info from up2k databases and the filesystem
+
+specify subnet `192.168.1.*` with argument `.=192.168.1.`,
+affecting all successive mappings
+
+usage:
+  ./partyjournal.py > partyjournal.html .=192.168.1. cart=125 steen=114 steen=131 sleepy=121 fscarlet=144 ed=101 ed=123
+
+"""
+
+import sys
+import base64
+import sqlite3
+import argparse
+from datetime import datetime
+from urllib.parse import quote_from_bytes as quote
+from urllib.parse import unquote_to_bytes as unquote
+
+
+FS_ENCODING = sys.getfilesystemencoding()
+
+
+class APF(argparse.ArgumentDefaultsHelpFormatter, argparse.RawDescriptionHelpFormatter):
+    pass
+
+
+##
+## snibbed from copyparty
+
+
+def s3dec(v):
+    if not v.startswith("//"):
+        return v
+
+    v = base64.urlsafe_b64decode(v.encode("ascii")[2:])
+    return v.decode(FS_ENCODING, "replace")
+
+
+def quotep(txt):
+    btxt = txt.encode("utf-8", "replace")
+    quot1 = quote(btxt, safe=b"/")
+    quot1 = quot1.encode("ascii")
+    quot2 = quot1.replace(b" ", b"+")
+    return quot2.decode("utf-8", "replace")
+
+
+def html_escape(s, quote=False, crlf=False):
+    """html.escape but also newlines"""
+    s = s.replace("&", "&amp;").replace("<", "&lt;").replace(">", "&gt;")
+    if quote:
+        s = s.replace('"', "&quot;").replace("'", "&#x27;")
+    if crlf:
+        s = s.replace("\r", "&#13;").replace("\n", "&#10;")
+
+    return s
+
+
+## end snibs
+##
+
+
+def main():
+    ap = argparse.ArgumentParser(formatter_class=APF)
+    ap.add_argument("who", nargs="*")
+    ar = ap.parse_args()
+
+    imap = {}
+    subnet = ""
+    for v in ar.who:
+        if "=" not in v:
+            raise Exception("bad who: " + v)
+
+        k, v = v.split("=")
+        if k == ".":
+            subnet = v
+            continue
+
+        imap["{}{}".format(subnet, v)] = k
+
+    print(repr(imap), file=sys.stderr)
+
+    print(
+        """\
+<!DOCTYPE html>
+<html lang="en">
+<head><meta charset="utf-8"><style>
+
+html, body {
+    color: #ccc;
+    background: #222;
+    font-family: sans-serif;
+}
+a {
+    color: #fc5;
+}
+td, th {
+    padding: .2em .5em;
+    border: 1px solid #999;
+    border-width: 0 1px 1px 0;
+    white-space: nowrap;
+}
+td:nth-child(1),
+td:nth-child(2),
+td:nth-child(3) {
+    font-family: monospace, monospace;
+    text-align: right;
+}
+tr:first-child {
+    position: sticky;
+    top: -1px;
+}
+th {
+    background: #222;
+    text-align: left;
+}
+
+</style></head><body><table><tr>
+    <th>wark</th>
+    <th>time</th>
+    <th>size</th>
+    <th>who</th>
+    <th>link</th>
+</tr>"""
+    )
+
+    db_path = ".hist/up2k.db"
+    conn = sqlite3.connect(db_path)
+    q = r"pragma table_info(up)"
+    inf = conn.execute(q).fetchall()
+    cols = [x[1] for x in inf]
+    print("<!-- " + str(cols) + " -->")
+    # ['w', 'mt', 'sz', 'rd', 'fn', 'ip', 'at']
+
+    q = r"select * from up order by case when at > 0 then at else mt end"
+    for w, mt, sz, rd, fn, ip, at in conn.execute(q):
+        link = "/".join([s3dec(x) for x in [rd, fn] if x])
+        if fn.startswith("put-") and sz < 4096:
+            try:
+                with open(link, "rb") as f:
+                    txt = f.read().decode("utf-8", "replace")
+            except:
+                continue
+
+            if txt.startswith("msg="):
+                txt = txt.encode("utf-8", "replace")
+                txt = unquote(txt.replace(b"+", b" "))
+                link = txt.decode("utf-8")[4:]
+
+        sz = "{:,}".format(sz)
+        v = [
+            w[:16],
+            datetime.utcfromtimestamp(at if at > 0 else mt).strftime(
+                "%Y-%m-%d %H:%M:%S"
+            ),
+            sz,
+            imap.get(ip, ip),
+        ]
+
+        row = "<tr>\n  "
+        row += "\n  ".join(["<td>{}</th>".format(x) for x in v])
+        row += '\n  <td><a href="{}">{}</a></td>'.format(link, html_escape(link))
+        row += "\n</tr>"
+        print(row)
+
+    print("</table></body></html>")
+
+
+if __name__ == "__main__":
+    main()

bin/prisonparty.sh

@@ -0,0 +1,128 @@
+#!/bin/bash
+set -e
+
+# runs copyparty (or any other program really) in a chroot
+#
+# assumption: these directories, and everything within, are owned by root
+sysdirs=( /bin /lib /lib32 /lib64 /sbin /usr )
+
+
+# error-handler
+help() { cat <<'EOF'
+
+usage:
+  ./prisonparty.sh <ROOTDIR> <UID> <GID> [VOLDIR [VOLDIR...]] -- python3 copyparty-sfx.py [...]
+
+example:
+  ./prisonparty.sh /var/lib/copyparty-jail 1000 1000 /mnt/nas/music -- python3 copyparty-sfx.py -v /mnt/nas/music::rwmd
+
+example for running straight from source (instead of using an sfx):
+  PYTHONPATH=$PWD ./prisonparty.sh /var/lib/copyparty-jail 1000 1000 /mnt/nas/music -- python3 -um copyparty -v /mnt/nas/music::rwmd
+
+note that if you have python modules installed as --user (such as bpm/key detectors),
+  you should add /home/foo/.local as a VOLDIR
+
+EOF
+exit 1
+}
+
+
+# read arguments
+trap help EXIT
+jail="$(realpath "$1")"; shift
+uid="$1"; shift
+gid="$1"; shift
+
+vols=()
+while true; do
+	v="$1"; shift
+	[ "$v" = -- ] && break  # end of volumes
+	[ "$#" -eq 0 ] && break  # invalid usage
+	vols+=( "$(realpath "$v")" )
+done
+pybin="$1"; shift
+pybin="$(command -v "$pybin")"
+pyarg=
+while true; do
+	v="$1"
+	[ "${v:0:1}" = - ] || break
+	pyarg="$pyarg $v"
+	shift
+done
+cpp="$1"; shift
+[ -d "$cpp" ] && cppdir="$PWD" || {
+	# sfx, not module
+	cpp="$(realpath "$cpp")"
+	cppdir="$(dirname "$cpp")"
+}
+trap - EXIT
+
+
+# debug/vis
+echo
+echo "chroot-dir = $jail"
+echo "user:group = $uid:$gid"
+echo " copyparty = $cpp"
+echo
+printf '\033[33m%s\033[0m\n' "copyparty can access these folders and all their subdirectories:"
+for v in "${vols[@]}"; do
+	printf '\033[36m ├─\033[0m %s \033[36m ── added by (You)\033[0m\n' "$v"
+done
+printf '\033[36m ├─\033[0m %s \033[36m ── where the copyparty binary is\033[0m\n' "$cppdir"
+printf '\033[36m ╰─\033[0m %s \033[36m ── the folder you are currently in\033[0m\n' "$PWD"
+vols+=("$cppdir" "$PWD")
+echo
+
+
+# remove any trailing slashes
+jail="${jail%/}"
+
+
+# bind-mount system directories and volumes
+printf '%s\n' "${sysdirs[@]}" "${vols[@]}" | sed -r 's`/$``' | LC_ALL=C sort | uniq |
+while IFS= read -r v; do
+	[ -e "$v" ] || {
+		# printf '\033[1;31mfolder does not exist:\033[0m %s\n' "/$v"
+		continue
+	}
+	i1=$(stat -c%D.%i "$v"      2>/dev/null || echo a)
+	i2=$(stat -c%D.%i "$jail$v" 2>/dev/null || echo b)
+	# echo "v [$v] i1 [$i1] i2 [$i2]"
+	[ $i1 = $i2 ] && continue
+	
+	mkdir -p "$jail$v"
+	mount --bind "$v" "$jail$v"
+done
+
+
+cln() {
+	rv=$?
+	# cleanup if not in use
+	lsof "$jail" | grep -qF "$jail" &&
+		echo "chroot is in use, will not cleanup" ||
+	{
+		mount | grep -F " on $jail" |
+		awk '{sub(/ type .*/,"");sub(/.* on /,"");print}' |
+		LC_ALL=C sort -r  | tee /dev/stderr | tr '\n' '\0' | xargs -r0 umount
+	}
+	exit $rv
+}
+trap cln EXIT
+
+
+# create a tmp
+mkdir -p "$jail/tmp"
+chmod 777 "$jail/tmp"
+
+
+# run copyparty
+export HOME=$(getent passwd $uid | cut -d: -f6)
+export USER=$(getent passwd $uid | cut -d: -f1)
+export LOGNAME="$USER"
+#echo "pybin [$pybin]"
+#echo "pyarg [$pyarg]"
+#echo "cpp [$cpp]"
+chroot --userspec=$uid:$gid "$jail" "$pybin" $pyarg "$cpp" "$@" &
+p=$!
+trap 'kill $p' INT TERM
+wait

bin/unforget.py

@@ -0,0 +1,99 @@
+#!/usr/bin/env python3
+
+"""
+unforget.py: rebuild db from logfiles
+2022-09-07, v0.1, ed <irc.rizon.net>, MIT-Licensed
+https://github.com/9001/copyparty/blob/hovudstraum/bin/unforget.py
+
+only makes sense if running copyparty with --no-forget
+(e.g. immediately shifting uploads to other storage)
+
+usage:
+  xz -d < log | ./unforget.py .hist/up2k.db
+
+"""
+
+import re
+import sys
+import json
+import base64
+import sqlite3
+import argparse
+
+
+FS_ENCODING = sys.getfilesystemencoding()
+
+
+class APF(argparse.ArgumentDefaultsHelpFormatter, argparse.RawDescriptionHelpFormatter):
+    pass
+
+
+mem_cur = sqlite3.connect(":memory:").cursor()
+mem_cur.execute(r"create table a (b text)")
+
+
+def s3enc(rd: str, fn: str) -> tuple[str, str]:
+    ret: list[str] = []
+    for v in [rd, fn]:
+        try:
+            mem_cur.execute("select * from a where b = ?", (v,))
+            ret.append(v)
+        except:
+            wtf8 = v.encode(FS_ENCODING, "surrogateescape")
+            ret.append("//" + base64.urlsafe_b64encode(wtf8).decode("ascii"))
+
+    return ret[0], ret[1]
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("db")
+    ar = ap.parse_args()
+
+    db = sqlite3.connect(ar.db).cursor()
+    ptn_times = re.compile(r"no more chunks, setting times \(([0-9]+)")
+    at = 0
+    ctr = 0
+
+    for ln in [x.decode("utf-8", "replace").rstrip() for x in sys.stdin.buffer]:
+        if "no more chunks, setting times (" in ln:
+            m = ptn_times.search(ln)
+            if m:
+                at = int(m.group(1))
+
+        if '"hash": []' in ln:
+            try:
+                ofs = ln.find("{")
+                j = json.loads(ln[ofs:])
+            except:
+                pass
+
+            w = j["wark"]
+            if db.execute("select w from up where w = ?", (w,)).fetchone():
+                continue
+
+            # PYTHONPATH=/home/ed/dev/copyparty/ python3 -m copyparty -e2dsa  -v foo:foo:rwmd,ed -aed:wark --no-forget
+            # 05:34:43.845 127.0.0.1 42496       no more chunks, setting times (1662528883, 1658001882)
+            # 05:34:43.863 127.0.0.1 42496       {"name": "f\"2", "purl": "/foo/bar/baz/", "size": 1674, "lmod": 1658001882, "sprs": true, "hash": [], "wark": "LKIWpp2jEAh9dH3fu-DobuURFGEKlODXDGTpZ1otMhUg"}
+            # |                      w                       |     mt     |  sz  |   rd    | fn  |    ip     |     at     |
+            # | LKIWpp2jEAh9dH3fu-DobuURFGEKlODXDGTpZ1otMhUg | 1658001882 | 1674 | bar/baz | f"2 | 127.0.0.1 | 1662528883 |
+
+            rd, fn = s3enc(j["purl"].strip("/"), j["name"])
+            ip = ln.split(" ")[1].split("m")[-1]
+
+            q = "insert into up values (?,?,?,?,?,?,?)"
+            v = (w, int(j["lmod"]), int(j["size"]), rd, fn, ip, at)
+            db.execute(q, v)
+            ctr += 1
+            if ctr % 1024 == 1023:
+                print(f"{ctr} commit...")
+                db.connection.commit()
+
+    if ctr:
+        db.connection.commit()
+
+    print(f"unforgot {ctr} files")
+
+
+if __name__ == "__main__":
+    main()

bin/up2k.py

@@ -0,0 +1,978 @@
+#!/usr/bin/env python3
+from __future__ import print_function, unicode_literals
+
+"""
+up2k.py: upload to copyparty
+2022-11-29, v0.22, ed <irc.rizon.net>, MIT-Licensed
+https://github.com/9001/copyparty/blob/hovudstraum/bin/up2k.py
+
+- dependencies: requests
+- supports python 2.6, 2.7, and 3.3 through 3.11
+
+- almost zero error-handling
+- but if something breaks just try again and it'll autoresume
+"""
+
+import os
+import sys
+import stat
+import math
+import time
+import atexit
+import signal
+import base64
+import hashlib
+import platform
+import threading
+import datetime
+
+try:
+    import argparse
+except:
+    m = "\n  ERROR: need 'argparse'; download it here:\n   https://github.com/ThomasWaldmann/argparse/raw/master/argparse.py\n"
+    print(m)
+    raise
+
+try:
+    import requests
+except ImportError:
+    if sys.version_info > (2, 7):
+        m = "\nERROR: need 'requests'; please run this command:\n {0} -m pip install --user requests\n"
+    else:
+        m = "requests/2.18.4 urllib3/1.23 chardet/3.0.4 certifi/2020.4.5.1 idna/2.7"
+        m = ["   https://pypi.org/project/" + x + "/#files" for x in m.split()]
+        m = "\n  ERROR: need these:\n" + "\n".join(m) + "\n"
+
+    print(m.format(sys.executable))
+    sys.exit(1)
+
+
+# from copyparty/__init__.py
+PY2 = sys.version_info < (3,)
+if PY2:
+    from Queue import Queue
+    from urllib import quote, unquote
+
+    sys.dont_write_bytecode = True
+    bytes = str
+else:
+    from queue import Queue
+    from urllib.parse import unquote_to_bytes as unquote
+    from urllib.parse import quote_from_bytes as quote
+
+    unicode = str
+
+VT100 = platform.system() != "Windows"
+
+
+req_ses = requests.Session()
+
+
+class Daemon(threading.Thread):
+    def __init__(self, target, name=None, a=None):
+        # type: (Any, Any, Any) -> None
+        threading.Thread.__init__(self, target=target, args=a or (), name=name)
+        self.daemon = True
+        self.start()
+
+
+class File(object):
+    """an up2k upload task; represents a single file"""
+
+    def __init__(self, top, rel, size, lmod):
+        self.top = top  # type: bytes
+        self.rel = rel.replace(b"\\", b"/")  # type: bytes
+        self.size = size  # type: int
+        self.lmod = lmod  # type: float
+
+        self.abs = os.path.join(top, rel)  # type: bytes
+        self.name = self.rel.split(b"/")[-1].decode("utf-8", "replace")  # type: str
+
+        # set by get_hashlist
+        self.cids = []  # type: list[tuple[str, int, int]]  # [ hash, ofs, sz ]
+        self.kchunks = {}  # type: dict[str, tuple[int, int]]  # hash: [ ofs, sz ]
+
+        # set by handshake
+        self.recheck = False  # duplicate; redo handshake after all files done
+        self.ucids = []  # type: list[str]  # chunks which need to be uploaded
+        self.wark = None  # type: str
+        self.url = None  # type: str
+
+        # set by upload
+        self.up_b = 0  # type: int
+        self.up_c = 0  # type: int
+
+        # t = "size({}) lmod({}) top({}) rel({}) abs({}) name({})\n"
+        # eprint(t.format(self.size, self.lmod, self.top, self.rel, self.abs, self.name))
+
+
+class FileSlice(object):
+    """file-like object providing a fixed window into a file"""
+
+    def __init__(self, file, cid):
+        # type: (File, str) -> None
+
+        self.car, self.len = file.kchunks[cid]
+        self.cdr = self.car + self.len
+        self.ofs = 0  # type: int
+        self.f = open(file.abs, "rb", 512 * 1024)
+        self.f.seek(self.car)
+
+        # https://stackoverflow.com/questions/4359495/what-is-exactly-a-file-like-object-in-python
+        # IOBase, RawIOBase, BufferedIOBase
+        funs = "close closed __enter__ __exit__ __iter__ isatty __next__ readable seekable writable"
+        try:
+            for fun in funs.split():
+                setattr(self, fun, getattr(self.f, fun))
+        except:
+            pass  # py27 probably
+
+    def tell(self):
+        return self.ofs
+
+    def seek(self, ofs, wh=0):
+        if wh == 1:
+            ofs = self.ofs + ofs
+        elif wh == 2:
+            ofs = self.len + ofs  # provided ofs is negative
+
+        if ofs < 0:
+            ofs = 0
+        elif ofs >= self.len:
+            ofs = self.len - 1
+
+        self.ofs = ofs
+        self.f.seek(self.car + ofs)
+
+    def read(self, sz):
+        sz = min(sz, self.len - self.ofs)
+        ret = self.f.read(sz)
+        self.ofs += len(ret)
+        return ret
+
+
+class MTHash(object):
+    def __init__(self, cores):
+        self.f = None
+        self.sz = 0
+        self.csz = 0
+        self.omutex = threading.Lock()
+        self.imutex = threading.Lock()
+        self.work_q = Queue()
+        self.done_q = Queue()
+        self.thrs = []
+        for _ in range(cores):
+            self.thrs.append(Daemon(self.worker))
+
+    def hash(self, f, fsz, chunksz, pcb=None, pcb_opaque=None):
+        with self.omutex:
+            self.f = f
+            self.sz = fsz
+            self.csz = chunksz
+
+            chunks = {}
+            nchunks = int(math.ceil(fsz / chunksz))
+            for nch in range(nchunks):
+                self.work_q.put(nch)
+
+            ex = ""
+            for nch in range(nchunks):
+                qe = self.done_q.get()
+                try:
+                    nch, dig, ofs, csz = qe
+                    chunks[nch] = [dig, ofs, csz]
+                except:
+                    ex = ex or qe
+
+                if pcb:
+                    pcb(pcb_opaque, chunksz * nch)
+
+            if ex:
+                raise Exception(ex)
+
+            ret = []
+            for n in range(nchunks):
+                ret.append(chunks[n])
+
+            self.f = None
+            self.csz = 0
+            self.sz = 0
+            return ret
+
+    def worker(self):
+        while True:
+            ofs = self.work_q.get()
+            try:
+                v = self.hash_at(ofs)
+            except Exception as ex:
+                v = str(ex)
+
+            self.done_q.put(v)
+
+    def hash_at(self, nch):
+        f = self.f
+        ofs = ofs0 = nch * self.csz
+        hashobj = hashlib.sha512()
+        chunk_sz = chunk_rem = min(self.csz, self.sz - ofs)
+        while chunk_rem > 0:
+            with self.imutex:
+                f.seek(ofs)
+                buf = f.read(min(chunk_rem, 1024 * 1024 * 12))
+
+            if not buf:
+                raise Exception("EOF at " + str(ofs))
+
+            hashobj.update(buf)
+            chunk_rem -= len(buf)
+            ofs += len(buf)
+
+        digest = hashobj.digest()[:33]
+        digest = base64.urlsafe_b64encode(digest).decode("utf-8")
+        return nch, digest, ofs0, chunk_sz
+
+
+_print = print
+
+
+def eprint(*a, **ka):
+    ka["file"] = sys.stderr
+    ka["end"] = ""
+    if not PY2:
+        ka["flush"] = True
+
+    _print(*a, **ka)
+    if PY2 or not VT100:
+        sys.stderr.flush()
+
+
+def flushing_print(*a, **ka):
+    _print(*a, **ka)
+    if "flush" not in ka:
+        sys.stdout.flush()
+
+
+if not VT100:
+    print = flushing_print
+
+
+def termsize():
+    env = os.environ
+
+    def ioctl_GWINSZ(fd):
+        try:
+            import fcntl, termios, struct
+
+            cr = struct.unpack("hh", fcntl.ioctl(fd, termios.TIOCGWINSZ, "1234"))
+        except:
+            return
+        return cr
+
+    cr = ioctl_GWINSZ(0) or ioctl_GWINSZ(1) or ioctl_GWINSZ(2)
+    if not cr:
+        try:
+            fd = os.open(os.ctermid(), os.O_RDONLY)
+            cr = ioctl_GWINSZ(fd)
+            os.close(fd)
+        except:
+            pass
+    if not cr:
+        try:
+            cr = (env["LINES"], env["COLUMNS"])
+        except:
+            cr = (25, 80)
+    return int(cr[1]), int(cr[0])
+
+
+class CTermsize(object):
+    def __init__(self):
+        self.ev = False
+        self.margin = None
+        self.g = None
+        self.w, self.h = termsize()
+
+        try:
+            signal.signal(signal.SIGWINCH, self.ev_sig)
+        except:
+            return
+
+        Daemon(self.worker)
+
+    def worker(self):
+        while True:
+            time.sleep(0.5)
+            if not self.ev:
+                continue
+
+            self.ev = False
+            self.w, self.h = termsize()
+
+            if self.margin is not None:
+                self.scroll_region(self.margin)
+
+    def ev_sig(self, *a, **ka):
+        self.ev = True
+
+    def scroll_region(self, margin):
+        self.margin = margin
+        if margin is None:
+            self.g = None
+            eprint("\033[s\033[r\033[u")
+        else:
+            self.g = 1 + self.h - margin
+            t = "{0}\033[{1}A".format("\n" * margin, margin)
+            eprint("{0}\033[s\033[1;{1}r\033[u".format(t, self.g - 1))
+
+
+ss = CTermsize()
+
+
+def _scd(err, top):
+    """non-recursive listing of directory contents, along with stat() info"""
+    with os.scandir(top) as dh:
+        for fh in dh:
+            abspath = os.path.join(top, fh.name)
+            try:
+                yield [abspath, fh.stat()]
+            except Exception as ex:
+                err.append((abspath, str(ex)))
+
+
+def _lsd(err, top):
+    """non-recursive listing of directory contents, along with stat() info"""
+    for name in os.listdir(top):
+        abspath = os.path.join(top, name)
+        try:
+            yield [abspath, os.stat(abspath)]
+        except Exception as ex:
+            err.append((abspath, str(ex)))
+
+
+if hasattr(os, "scandir") and sys.version_info > (3, 6):
+    statdir = _scd
+else:
+    statdir = _lsd
+
+
+def walkdir(err, top, seen):
+    """recursive statdir"""
+    atop = os.path.abspath(os.path.realpath(top))
+    if atop in seen:
+        err.append((top, "recursive-symlink"))
+        return
+
+    seen = seen[:] + [atop]
+    for ap, inf in sorted(statdir(err, top)):
+        if stat.S_ISDIR(inf.st_mode):
+            try:
+                for x in walkdir(err, ap, seen):
+                    yield x
+            except Exception as ex:
+                err.append((ap, str(ex)))
+        else:
+            yield ap, inf
+
+
+def walkdirs(err, tops):
+    """recursive statdir for a list of tops, yields [top, relpath, stat]"""
+    sep = "{0}".format(os.sep).encode("ascii")
+    for top in tops:
+        if top[-1:] == sep:
+            stop = top.rstrip(sep)
+        else:
+            stop = os.path.dirname(top)
+
+        if os.path.isdir(top):
+            for ap, inf in walkdir(err, top, []):
+                yield stop, ap[len(stop) :].lstrip(sep), inf
+        else:
+            d, n = top.rsplit(sep, 1)
+            yield d, n, os.stat(top)
+
+
+# mostly from copyparty/util.py
+def quotep(btxt):
+    quot1 = quote(btxt, safe=b"/")
+    if not PY2:
+        quot1 = quot1.encode("ascii")
+
+    return quot1.replace(b" ", b"+")
+
+
+# from copyparty/util.py
+def humansize(sz, terse=False):
+    """picks a sensible unit for the given extent"""
+    for unit in ["B", "KiB", "MiB", "GiB", "TiB"]:
+        if sz < 1024:
+            break
+
+        sz /= 1024.0
+
+    ret = " ".join([str(sz)[:4].rstrip("."), unit])
+
+    if not terse:
+        return ret
+
+    return ret.replace("iB", "").replace(" ", "")
+
+
+# from copyparty/up2k.py
+def up2k_chunksize(filesize):
+    """gives The correct chunksize for up2k hashing"""
+    chunksize = 1024 * 1024
+    stepsize = 512 * 1024
+    while True:
+        for mul in [1, 2]:
+            nchunks = math.ceil(filesize * 1.0 / chunksize)
+            if nchunks <= 256 or (chunksize >= 32 * 1024 * 1024 and nchunks < 4096):
+                return chunksize
+
+            chunksize += stepsize
+            stepsize *= mul
+
+
+# mostly from copyparty/up2k.py
+def get_hashlist(file, pcb, mth):
+    # type: (File, any, any) -> None
+    """generates the up2k hashlist from file contents, inserts it into `file`"""
+
+    chunk_sz = up2k_chunksize(file.size)
+    file_rem = file.size
+    file_ofs = 0
+    ret = []
+    with open(file.abs, "rb", 512 * 1024) as f:
+        if mth and file.size >= 1024 * 512:
+            ret = mth.hash(f, file.size, chunk_sz, pcb, file)
+            file_rem = 0
+
+        while file_rem > 0:
+            # same as `hash_at` except for `imutex` / bufsz
+            hashobj = hashlib.sha512()
+            chunk_sz = chunk_rem = min(chunk_sz, file_rem)
+            while chunk_rem > 0:
+                buf = f.read(min(chunk_rem, 64 * 1024))
+                if not buf:
+                    raise Exception("EOF at " + str(f.tell()))
+
+                hashobj.update(buf)
+                chunk_rem -= len(buf)
+
+            digest = hashobj.digest()[:33]
+            digest = base64.urlsafe_b64encode(digest).decode("utf-8")
+
+            ret.append([digest, file_ofs, chunk_sz])
+            file_ofs += chunk_sz
+            file_rem -= chunk_sz
+
+            if pcb:
+                pcb(file, file_ofs)
+
+    file.cids = ret
+    file.kchunks = {}
+    for k, v1, v2 in ret:
+        file.kchunks[k] = [v1, v2]
+
+
+def handshake(url, file, pw, search):
+    # type: (str, File, Any, bool) -> tuple[list[str], bool]
+    """
+    performs a handshake with the server; reply is:
+      if search, a list of search results
+      otherwise, a list of chunks to upload
+    """
+
+    req = {
+        "hash": [x[0] for x in file.cids],
+        "name": file.name,
+        "lmod": file.lmod,
+        "size": file.size,
+    }
+    if search:
+        req["srch"] = 1
+
+    headers = {"Content-Type": "text/plain"}  # wtf ed
+    if pw:
+        headers["Cookie"] = "=".join(["cppwd", pw])
+
+    if file.url:
+        url = file.url
+    elif b"/" in file.rel:
+        url += quotep(file.rel.rsplit(b"/", 1)[0]).decode("utf-8", "replace")
+
+    while True:
+        try:
+            r = req_ses.post(url, headers=headers, json=req)
+            break
+        except Exception as ex:
+            em = str(ex).split("SSLError(")[-1]
+            eprint("handshake failed, retrying: {0}\n  {1}\n\n".format(file.name, em))
+            time.sleep(1)
+
+    sc = r.status_code
+    if sc >= 400:
+        txt = r.text
+        if sc == 422 or "<pre>partial upload exists at a different" in txt:
+            file.recheck = True
+            return [], False
+        elif sc == 409 or "<pre>upload rejected, file already exists" in txt:
+            return [], False
+
+        raise Exception("http {0}: {1}".format(sc, txt))
+
+    try:
+        r = r.json()
+    except:
+        raise Exception(r.text)
+
+    if search:
+        return r["hits"], False
+
+    try:
+        pre, url = url.split("://")
+        pre += "://"
+    except:
+        pre = ""
+
+    file.url = pre + url.split("/")[0] + r["purl"]
+    file.name = r["name"]
+    file.wark = r["wark"]
+
+    return r["hash"], r["sprs"]
+
+
+def upload(file, cid, pw):
+    # type: (File, str, Any) -> None
+    """upload one specific chunk, `cid` (a chunk-hash)"""
+
+    headers = {
+        "X-Up2k-Hash": cid,
+        "X-Up2k-Wark": file.wark,
+        "Content-Type": "application/octet-stream",
+    }
+    if pw:
+        headers["Cookie"] = "=".join(["cppwd", pw])
+
+    f = FileSlice(file, cid)
+    try:
+        r = req_ses.post(file.url, headers=headers, data=f)
+        if not r:
+            raise Exception(repr(r))
+
+        _ = r.content
+    finally:
+        f.f.close()
+
+
+class Ctl(object):
+    """
+    this will be the coordinator which runs everything in parallel
+    (hashing, handshakes, uploads)  but right now it's p dumb
+    """
+
+    def __init__(self, ar):
+        self.ar = ar
+        ar.files = [
+            os.path.abspath(os.path.realpath(x.encode("utf-8")))
+            + (x[-1:] if x[-1:] == os.sep else "").encode("utf-8")
+            for x in ar.files
+        ]
+        ar.url = ar.url.rstrip("/") + "/"
+        if "://" not in ar.url:
+            ar.url = "http://" + ar.url
+
+        eprint("\nscanning {0} locations\n".format(len(ar.files)))
+
+        nfiles = 0
+        nbytes = 0
+        err = []
+        for _, _, inf in walkdirs(err, ar.files):
+            nfiles += 1
+            nbytes += inf.st_size
+
+        if err:
+            eprint("\n# failed to access {0} paths:\n".format(len(err)))
+            for ap, msg in err:
+                if ar.v:
+                    eprint("{0}\n `-{1}\n\n".format(ap.decode("utf-8", "replace"), msg))
+                else:
+                    eprint(ap.decode("utf-8", "replace") + "\n")
+
+            eprint("^ failed to access those {0} paths ^\n\n".format(len(err)))
+
+            if not ar.v:
+                eprint("hint: set -v for detailed error messages\n")
+
+            if not ar.ok:
+                eprint("hint: aborting because --ok is not set\n")
+                return
+
+        eprint("found {0} files, {1}\n\n".format(nfiles, humansize(nbytes)))
+        self.nfiles = nfiles
+        self.nbytes = nbytes
+
+        if ar.td:
+            requests.packages.urllib3.disable_warnings()
+            req_ses.verify = False
+        if ar.te:
+            req_ses.verify = ar.te
+
+        self.filegen = walkdirs([], ar.files)
+        if ar.safe:
+            self._safe()
+        else:
+            self.hash_f = 0
+            self.hash_c = 0
+            self.hash_b = 0
+            self.up_f = 0
+            self.up_c = 0
+            self.up_b = 0
+            self.up_br = 0
+            self.hasher_busy = 1
+            self.handshaker_busy = 0
+            self.uploader_busy = 0
+            self.serialized = False
+
+            self.t0 = time.time()
+            self.t0_up = None
+            self.spd = None
+
+            self.mutex = threading.Lock()
+            self.q_handshake = Queue()  # type: Queue[File]
+            self.q_upload = Queue()  # type: Queue[tuple[File, str]]
+            self.recheck = []  # type: list[File]
+
+            self.st_hash = [None, "(idle, starting...)"]  # type: tuple[File, int]
+            self.st_up = [None, "(idle, starting...)"]  # type: tuple[File, int]
+
+            self.mth = MTHash(ar.J) if ar.J > 1 else None
+
+            self._fancy()
+
+    def _safe(self):
+        """minimal basic slow boring fallback codepath"""
+        search = self.ar.s
+        for nf, (top, rel, inf) in enumerate(self.filegen):
+            file = File(top, rel, inf.st_size, inf.st_mtime)
+            upath = file.abs.decode("utf-8", "replace")
+
+            print("{0} {1}\n  hash...".format(self.nfiles - nf, upath))
+            get_hashlist(file, None, None)
+
+            burl = self.ar.url[:12] + self.ar.url[8:].split("/")[0] + "/"
+            while True:
+                print("  hs...")
+                hs, _ = handshake(self.ar.url, file, self.ar.a, search)
+                if search:
+                    if hs:
+                        for hit in hs:
+                            print("  found: {0}{1}".format(burl, hit["rp"]))
+                    else:
+                        print("  NOT found")
+                    break
+
+                file.ucids = hs
+                if not hs:
+                    break
+
+                print("{0} {1}".format(self.nfiles - nf, upath))
+                ncs = len(hs)
+                for nc, cid in enumerate(hs):
+                    print("  {0} up {1}".format(ncs - nc, cid))
+                    upload(file, cid, self.ar.a)
+
+            print("  ok!")
+            if file.recheck:
+                self.recheck.append(file)
+
+        if not self.recheck:
+            return
+
+        eprint("finalizing {0} duplicate files".format(len(self.recheck)))
+        for file in self.recheck:
+            handshake(self.ar.url, file, self.ar.a, search)
+
+    def _fancy(self):
+        if VT100:
+            atexit.register(self.cleanup_vt100)
+            ss.scroll_region(3)
+
+        Daemon(self.hasher)
+        for _ in range(self.ar.j):
+            Daemon(self.handshaker)
+            Daemon(self.uploader)
+
+        idles = 0
+        while idles < 3:
+            time.sleep(0.07)
+            with self.mutex:
+                if (
+                    self.q_handshake.empty()
+                    and self.q_upload.empty()
+                    and not self.hasher_busy
+                    and not self.handshaker_busy
+                    and not self.uploader_busy
+                ):
+                    idles += 1
+                else:
+                    idles = 0
+
+            if VT100:
+                maxlen = ss.w - len(str(self.nfiles)) - 14
+                txt = "\033[s\033[{0}H".format(ss.g)
+                for y, k, st, f in [
+                    [0, "hash", self.st_hash, self.hash_f],
+                    [1, "send", self.st_up, self.up_f],
+                ]:
+                    txt += "\033[{0}H{1}:".format(ss.g + y, k)
+                    file, arg = st
+                    if not file:
+                        txt += " {0}\033[K".format(arg)
+                    else:
+                        if y:
+                            p = 100 * file.up_b / file.size
+                        else:
+                            p = 100 * arg / file.size
+
+                        name = file.abs.decode("utf-8", "replace")[-maxlen:]
+                        if "/" in name:
+                            name = "\033[36m{0}\033[0m/{1}".format(*name.rsplit("/", 1))
+
+                        t = "{0:6.1f}% {1} {2}\033[K"
+                        txt += t.format(p, self.nfiles - f, name)
+
+                txt += "\033[{0}H ".format(ss.g + 2)
+            else:
+                txt = " "
+
+            if not self.up_br:
+                spd = self.hash_b / (time.time() - self.t0)
+                eta = (self.nbytes - self.hash_b) / (spd + 1)
+            else:
+                spd = self.up_br / (time.time() - self.t0_up)
+                spd = self.spd = (self.spd or spd) * 0.9 + spd * 0.1
+                eta = (self.nbytes - self.up_b) / (spd + 1)
+
+            spd = humansize(spd)
+            eta = str(datetime.timedelta(seconds=int(eta)))
+            sleft = humansize(self.nbytes - self.up_b)
+            nleft = self.nfiles - self.up_f
+            tail = "\033[K\033[u" if VT100 else "\r"
+
+            t = "{0} eta @ {1}/s, {2}, {3}# left".format(eta, spd, sleft, nleft)
+            eprint(txt + "\033]0;{0}\033\\\r{0}{1}".format(t, tail))
+
+        if not self.recheck:
+            return
+
+        eprint("finalizing {0} duplicate files".format(len(self.recheck)))
+        for file in self.recheck:
+            handshake(self.ar.url, file, self.ar.a, False)
+
+    def cleanup_vt100(self):
+        ss.scroll_region(None)
+        eprint("\033[J\033]0;\033\\")
+
+    def cb_hasher(self, file, ofs):
+        self.st_hash = [file, ofs]
+
+    def hasher(self):
+        prd = None
+        ls = {}
+        for top, rel, inf in self.filegen:
+            if self.ar.z:
+                rd = os.path.dirname(rel)
+                if prd != rd:
+                    prd = rd
+                    headers = {}
+                    if self.ar.a:
+                        headers["Cookie"] = "=".join(["cppwd", self.ar.a])
+
+                    ls = {}
+                    try:
+                        print("      ls ~{0}".format(rd.decode("utf-8", "replace")))
+                        r = req_ses.get(
+                            self.ar.url.encode("utf-8") + quotep(rd) + b"?ls",
+                            headers=headers,
+                        )
+                        for f in r.json()["files"]:
+                            rfn = f["href"].split("?")[0].encode("utf-8", "replace")
+                            ls[unquote(rfn)] = f
+                    except:
+                        print("   mkdir ~{0}".format(rd.decode("utf-8", "replace")))
+
+                rf = ls.get(os.path.basename(rel), None)
+                if rf and rf["sz"] == inf.st_size and abs(rf["ts"] - inf.st_mtime) <= 1:
+                    self.nfiles -= 1
+                    self.nbytes -= inf.st_size
+                    continue
+
+            file = File(top, rel, inf.st_size, inf.st_mtime)
+            while True:
+                with self.mutex:
+                    if (
+                        self.hash_f - self.up_f == 1
+                        or (
+                            self.hash_b - self.up_b < 1024 * 1024 * 1024
+                            and self.hash_c - self.up_c < 512
+                        )
+                    ) and (
+                        not self.ar.nh
+                        or (
+                            self.q_upload.empty()
+                            and self.q_handshake.empty()
+                            and not self.uploader_busy
+                        )
+                    ):
+                        break
+
+                time.sleep(0.05)
+
+            get_hashlist(file, self.cb_hasher, self.mth)
+            with self.mutex:
+                self.hash_f += 1
+                self.hash_c += len(file.cids)
+                self.hash_b += file.size
+
+            self.q_handshake.put(file)
+
+        self.hasher_busy = 0
+        self.st_hash = [None, "(finished)"]
+
+    def handshaker(self):
+        search = self.ar.s
+        burl = self.ar.url[:8] + self.ar.url[8:].split("/")[0] + "/"
+        while True:
+            file = self.q_handshake.get()
+            if not file:
+                self.q_upload.put(None)
+                break
+
+            with self.mutex:
+                self.handshaker_busy += 1
+
+            upath = file.abs.decode("utf-8", "replace")
+            hs, sprs = handshake(self.ar.url, file, self.ar.a, search)
+            if search:
+                if hs:
+                    for hit in hs:
+                        t = "found: {0}\n  {1}{2}\n"
+                        print(t.format(upath, burl, hit["rp"]), end="")
+                else:
+                    print("NOT found: {0}\n".format(upath), end="")
+
+                with self.mutex:
+                    self.up_f += 1
+                    self.up_c += len(file.cids)
+                    self.up_b += file.size
+                    self.handshaker_busy -= 1
+
+                continue
+
+            if file.recheck:
+                self.recheck.append(file)
+
+            with self.mutex:
+                if hs and not sprs and not self.serialized:
+                    t = "server filesystem does not support sparse files; serializing uploads\n"
+                    eprint(t)
+                    self.serialized = True
+                    for _ in range(self.ar.j - 1):
+                        self.q_upload.put(None)
+                if not hs:
+                    # all chunks done
+                    self.up_f += 1
+                    self.up_c += len(file.cids) - file.up_c
+                    self.up_b += file.size - file.up_b
+
+                if hs and file.up_c:
+                    # some chunks failed
+                    self.up_c -= len(hs)
+                    file.up_c -= len(hs)
+                    for cid in hs:
+                        sz = file.kchunks[cid][1]
+                        self.up_b -= sz
+                        file.up_b -= sz
+
+                file.ucids = hs
+                self.handshaker_busy -= 1
+
+            if not hs:
+                kw = "uploaded" if file.up_b else "   found"
+                print("{0} {1}".format(kw, upath))
+            for cid in hs:
+                self.q_upload.put([file, cid])
+
+    def uploader(self):
+        while True:
+            task = self.q_upload.get()
+            if not task:
+                self.st_up = [None, "(finished)"]
+                break
+
+            with self.mutex:
+                self.uploader_busy += 1
+                self.t0_up = self.t0_up or time.time()
+
+            file, cid = task
+            try:
+                upload(file, cid, self.ar.a)
+            except:
+                eprint("upload failed, retrying: {0} #{1}\n".format(file.name, cid[:8]))
+                pass  # handshake will fix it
+
+            with self.mutex:
+                sz = file.kchunks[cid][1]
+                file.ucids = [x for x in file.ucids if x != cid]
+                if not file.ucids:
+                    self.q_handshake.put(file)
+
+                self.st_up = [file, cid]
+                file.up_b += sz
+                self.up_b += sz
+                self.up_br += sz
+                file.up_c += 1
+                self.up_c += 1
+                self.uploader_busy -= 1
+
+
+class APF(argparse.ArgumentDefaultsHelpFormatter, argparse.RawDescriptionHelpFormatter):
+    pass
+
+
+def main():
+    time.strptime("19970815", "%Y%m%d")  # python#7980
+    if not VT100:
+        os.system("rem")  # enables colors
+
+    cores = (os.cpu_count() if hasattr(os, "cpu_count") else 0) or 2
+    hcores = min(cores, 3)  # 4% faster than 4+ on py3.9 @ r5-4500U
+
+    # fmt: off
+    ap = app = argparse.ArgumentParser(formatter_class=APF, epilog="""
+NOTE:
+source file/folder selection uses rsync syntax, meaning that:
+  "foo" uploads the entire folder to URL/foo/
+  "foo/" uploads the CONTENTS of the folder into URL/
+""")
+
+    ap.add_argument("url", type=unicode, help="server url, including destination folder")
+    ap.add_argument("files", type=unicode, nargs="+", help="files and/or folders to process")
+    ap.add_argument("-v", action="store_true", help="verbose")
+    ap.add_argument("-a", metavar="PASSWORD", help="password")
+    ap.add_argument("-s", action="store_true", help="file-search (disables upload)")
+    ap.add_argument("--ok", action="store_true", help="continue even if some local files are inaccessible")
+    ap = app.add_argument_group("performance tweaks")
+    ap.add_argument("-j", type=int, metavar="THREADS", default=4, help="parallel connections")
+    ap.add_argument("-J", type=int, metavar="THREADS", default=hcores, help="num cpu-cores to use for hashing; set 0 or 1 for single-core hashing")
+    ap.add_argument("-nh", action="store_true", help="disable hashing while uploading")
+    ap.add_argument("--safe", action="store_true", help="use simple fallback approach")
+    ap.add_argument("-z", action="store_true", help="ZOOMIN' (skip uploading files if they exist at the destination with the ~same last-modified timestamp, so same as yolo / turbo with date-chk but even faster)")
+    ap = app.add_argument_group("tls")
+    ap.add_argument("-te", metavar="PEM_FILE", help="certificate to expect/verify")
+    ap.add_argument("-td", action="store_true", help="disable certificate check")
+    # fmt: on
+
+    Ctl(app.parse_args())
+
+
+if __name__ == "__main__":
+    main()

bin/up2k.sh

@@ -8,7 +8,7 @@ set -e
 ##
 ## config
 
-datalen=$((2*1024*1024*1024))
+datalen=$((128*1024*1024))
 target=127.0.0.1
 posturl=/inc
 passwd=wark
@@ -37,10 +37,10 @@ gendata() {
 # pipe a chunk, get the base64 checksum
 gethash() {
     printf $(
-        sha512sum | cut -c-64 |
+        sha512sum | cut -c-66 |
         sed -r 's/ .*//;s/(..)/\\x\1/g'
     ) |
-    base64 -w0 | cut -c-43 |
+    base64 -w0 | cut -c-44 |
     tr '+/' '-_'
 }
 
@@ -123,7 +123,7 @@ printf '\033[36m'
 {
     {
         cat <<EOF
-POST $posturl/handshake.php HTTP/1.1
+POST $posturl/ HTTP/1.1
 Connection: Close
 Cookie: cppwd=$passwd
 Content-Type: text/plain;charset=UTF-8
@@ -145,14 +145,16 @@ printf '\033[0m\nwark: %s\n' $wark
 ##
 ## wait for signal to continue
 
-w8=/dev/shm/$salt.w8
-touch $w8
+true || {
+    w8=/dev/shm/$salt.w8
+    touch $w8
 
-echo "ready;  rm -f $w8"
+    echo "ready;  rm -f $w8"
 
-while [ -e $w8 ]; do
-    sleep 0.2
-done
+    while [ -e $w8 ]; do
+        sleep 0.2
+    done
+}
 
 
 ##
@@ -175,7 +177,7 @@ while [ $remains -gt 0 ]; do
     
     {
         cat <<EOF
-POST $posturl/chunkpit.php HTTP/1.1
+POST $posturl/ HTTP/1.1
 Connection: Keep-Alive
 Cookie: cppwd=$passwd
 Content-Type: application/octet-stream

contrib/README.md

@@ -0,0 +1,52 @@
+### [`plugins/`](plugins/)
+* example extensions
+
+### [`copyparty.bat`](copyparty.bat)
+* launches copyparty with no arguments (anon read+write within same folder)
+* intended for windows machines with no python.exe in PATH
+* works on windows, linux and macos
+* assumes `copyparty-sfx.py` was renamed to `copyparty.py` in the same folder as `copyparty.bat`
+
+### [`index.html`](index.html)
+* drop-in redirect from an httpd to copyparty
+* assumes the webserver and copyparty is running on the same server/IP
+* modify `10.13.1.1` as necessary if you wish to support browsers without javascript
+
+### [`sharex.sxcu`](sharex.sxcu)
+* sharex config file to upload screenshots and grab the URL
+* `RequestURL`: full URL to the target folder
+* `pw`: password (remove the `pw` line if anon-write)
+
+however if your copyparty is behind a reverse-proxy, you may want to use [`sharex-html.sxcu`](sharex-html.sxcu) instead:
+* `RequestURL`: full URL to the target folder
+* `URL`: full URL to the root folder (with trailing slash) followed by `$regex:1|1$`
+* `pw`: password (remove `Parameters` if anon-write)
+
+### [`media-osd-bgone.ps1`](media-osd-bgone.ps1)
+* disables the [windows OSD popup](https://user-images.githubusercontent.com/241032/122821375-0e08df80-d2dd-11eb-9fd9-184e8aacf1d0.png) (the thing on the left) which appears every time you hit media hotkeys to adjust volume or change song while playing music with the copyparty web-ui, or most other audio players really
+
+### [`explorer-nothumbs-nofoldertypes.reg`](explorer-nothumbs-nofoldertypes.reg)
+* disables thumbnails and folder-type detection in windows explorer
+* makes it way faster (especially for slow/networked locations (such as partyfuse))
+
+### [`webdav-basicauth.reg`](webdav-basicauth.reg)
+* enables webdav basic-auth over plaintext http; takes effect after a reboot OR after running `webdav-unlimit.bat`
+
+### [`webdav-unlimit.bat`](webdav-unlimit.bat)
+* removes the 47.6 MiB filesize limit when downloading from webdav
+
+### [`cfssl.sh`](cfssl.sh)
+* creates CA and server certificates using cfssl
+* give a 3rd argument to install it to your copyparty config
+* systemd service at [`systemd/cfssl.service`](systemd/cfssl.service)
+
+# OS integration
+init-scripts to start copyparty as a service
+* [`systemd/copyparty.service`](systemd/copyparty.service) runs the sfx normally
+* [`rc/copyparty`](rc/copyparty) runs sfx normally on freebsd, create a `copyparty` user
+* [`systemd/prisonparty.service`](systemd/prisonparty.service) runs the sfx in a chroot
+* [`openrc/copyparty`](openrc/copyparty)
+
+# Reverse-proxy
+copyparty has basic support for running behind another webserver
+* [`nginx/copyparty.conf`](nginx/copyparty.conf)

contrib/cfssl.sh

@@ -0,0 +1,73 @@
+#!/bin/bash
+set -e
+
+# ca-name and server-fqdn
+ca_name="$1"
+srv_fqdn="$2"
+
+[ -z "$srv_fqdn" ] && {
+	echo "need arg 1: ca name"
+	echo "need arg 2: server fqdn and/or IPs, comma-separated"
+	echo "optional arg 3: if set, write cert into copyparty cfg"
+	exit 1
+}
+
+
+gen_ca() {
+	(tee /dev/stderr <<EOF
+{"CN": "$ca_name ca",
+"CA": {"expiry":"87600h", "pathlen":0},
+"key": {"algo":"rsa", "size":4096},
+"names": [{"O":"$ca_name ca"}]}
+EOF
+	)|
+	cfssl gencert -initca - |
+	cfssljson -bare ca
+	
+	mv ca-key.pem ca.key
+	rm ca.csr
+}
+
+
+gen_srv() {
+	(tee /dev/stderr <<EOF
+{"key": {"algo":"rsa", "size":4096},
+"names": [{"O":"$ca_name - $srv_fqdn"}]}
+EOF
+	)|
+	cfssl gencert -ca ca.pem -ca-key ca.key \
+		-profile=www -hostname="$srv_fqdn" - |
+	cfssljson -bare "$srv_fqdn"
+
+	mv "$srv_fqdn-key.pem" "$srv_fqdn.key"
+	rm "$srv_fqdn.csr"
+}
+
+
+# create ca if not exist
+[ -e ca.key ] ||
+	gen_ca
+
+# always create server cert
+gen_srv
+
+
+# dump cert info
+show() {
+	openssl x509 -text -noout -in $1 |
+	awk '!o; {o=0} /[0-9a-f:]{16}/{o=1}'
+}
+show ca.pem
+show "$srv_fqdn.pem"
+
+
+# write cert into copyparty config
+[ -z "$3" ] || {
+	mkdir -p ~/.config/copyparty
+	cat "$srv_fqdn".{key,pem} ca.pem >~/.config/copyparty/cert.pem 
+}
+
+
+# rm *.key *.pem
+# cfssl print-defaults config
+# cfssl print-defaults csr

contrib/copyparty.bat

@@ -0,0 +1,33 @@
+exec python "$(dirname "$0")"/copyparty.py
+
+@rem on linux, the above will execute and the script will terminate
+@rem on windows, the rest of this script will run
+
+@echo off
+cls
+
+set py=
+for /f %%i in ('where python 2^>nul') do (
+    set "py=%%i"
+    goto c1
+)
+:c1
+
+if [%py%] == [] (
+    for /f %%i in ('where /r "%localappdata%\programs\python" python 2^>nul') do (
+        set "py=%%i"
+        goto c2
+    )
+)
+:c2
+
+if [%py%] == [] set "py=c:\python27\python.exe"
+
+if not exist "%py%" (
+    echo could not find python
+    echo(
+    pause
+    exit /b
+)
+
+start cmd /c %py% "%~dp0\copyparty.py"

contrib/explorer-nothumbs-nofoldertypes.reg

@@ -0,0 +1,31 @@
+Windows Registry Editor Version 5.00
+
+; this will do 3 things, all optional:
+;  1) disable thumbnails
+;  2) delete all existing folder type settings/detections
+;  3) disable folder type detection (force default columns)
+;
+; this makes the file explorer way faster,
+; especially on slow/networked locations
+
+
+; =====================================================================
+; 1) disable thumbnails
+
+[HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Explorer\Advanced]
+"IconsOnly"=dword:00000001
+
+
+; =====================================================================
+; 2) delete all existing folder type settings/detections
+
+[-HKEY_CURRENT_USER\Software\Classes\Local Settings\Software\Microsoft\Windows\Shell\Bags]
+
+[-HKEY_CURRENT_USER\Software\Classes\Local Settings\Software\Microsoft\Windows\Shell\BagMRU]
+
+
+; =====================================================================
+; 3) disable folder type detection
+
+[HKEY_CURRENT_USER\Software\Classes\Local Settings\Software\Microsoft\Windows\Shell\Bags\AllFolders\Shell]
+"FolderType"="NotSpecified"

contrib/index.html

@@ -0,0 +1,43 @@
+<!DOCTYPE html>
+<html lang="en">
+
+<head>
+	<meta charset="utf-8">
+	<title>⇆🎉 redirect</title>
+	<meta http-equiv="X-UA-Compatible" content="IE=edge">
+	<style>
+
+html, body {
+	font-family: sans-serif;
+}
+body {
+	padding: 1em 2em;
+	font-size: 1.5em;
+}
+a {
+	font-size: 1.2em;
+	padding: .1em;
+}
+
+</style>
+</head>
+<body>
+	<span id="desc">you probably want</span> <a id="redir" href="//10.13.1.1:3923/">copyparty</a>
+	<script>
+
+var a = document.getElementById('redir'),
+	proto = window.location.protocol.indexOf('https') === 0 ? 'https' : 'http',
+	loc = window.location.hostname || '127.0.0.1',
+	port = a.getAttribute('href').split(':').pop().split('/')[0],
+	url = proto + '://' + loc + ':' + port + '/';
+
+a.setAttribute('href', url);
+document.getElementById('desc').innerHTML = 'redirecting to';
+
+setTimeout(function() {
+	window.location.href = url;
+}, 500);
+
+</script>
+</body>
+</html>

contrib/media-osd-bgone.ps1

@@ -0,0 +1,104 @@
+# media-osd-bgone.ps1: disable media-control OSD on win10do
+# v1.1, 2021-06-25, ed <irc.rizon.net>, MIT-licensed
+# https://github.com/9001/copyparty/blob/hovudstraum/contrib/media-osd-bgone.ps1
+#
+# locates the first window that looks like the media OSD and minimizes it;
+# doing this once after each reboot should do the trick
+# (adjust the width/height filter if it doesn't work)
+#
+# ---------------------------------------------------------------------
+#
+# tip: save the following as "media-osd-bgone.bat" next to this script:
+#   start cmd /c "powershell -command ""set-executionpolicy -scope process bypass; .\media-osd-bgone.ps1"" & ping -n 2 127.1 >nul"
+#
+# then create a shortcut to that bat-file and move the shortcut here:
+#   %appdata%\Microsoft\Windows\Start Menu\Programs\Startup
+#
+# and now this will autorun on bootup
+
+
+Add-Type -TypeDefinition @"
+using System;
+using System.IO;
+using System.Threading;
+using System.Diagnostics;
+using System.Runtime.InteropServices;
+using System.Windows.Forms;
+
+namespace A {
+  public class B : Control {
+
+    [DllImport("user32.dll")]
+    static extern void keybd_event(byte bVk, byte bScan, uint dwFlags, int dwExtraInfo);
+
+    [DllImport("user32.dll", SetLastError = true)]
+    static extern IntPtr FindWindowEx(IntPtr hwndParent, IntPtr hwndChildAfter, string lpszClass, string lpszWindow);
+
+    [DllImport("user32.dll", SetLastError=true)]
+    static extern bool GetWindowRect(IntPtr hwnd, out RECT lpRect);
+
+    [DllImport("user32.dll")]
+    static extern bool ShowWindow(IntPtr hWnd, int nCmdShow);
+
+    [StructLayout(LayoutKind.Sequential)]
+    public struct RECT {
+      public int x;
+      public int y;
+      public int x2;
+      public int y2;
+    }
+    
+    bool fa() {
+      RECT r;
+      IntPtr it = IntPtr.Zero;
+      while ((it = FindWindowEx(IntPtr.Zero, it, "NativeHWNDHost", "")) != IntPtr.Zero) {
+        if (FindWindowEx(it, IntPtr.Zero, "DirectUIHWND", "") == IntPtr.Zero)
+          continue;
+        
+        if (!GetWindowRect(it, out r))
+          continue;
+
+        int w = r.x2 - r.x + 1;
+        int h = r.y2 - r.y + 1;
+
+        Console.WriteLine("[*] hwnd {0:x} @ {1}x{2} sz {3}x{4}", it, r.x, r.y, w, h);
+        if (h != 141)
+          continue;
+        
+        ShowWindow(it, 6);
+        Console.WriteLine("[+] poof");
+        return true;
+      }
+      return false;
+    }
+
+    void fb() {
+      keybd_event((byte)Keys.VolumeMute, 0, 0, 0);
+      keybd_event((byte)Keys.VolumeMute, 0, 2, 0);
+      Thread.Sleep(500);
+      keybd_event((byte)Keys.VolumeMute, 0, 0, 0);
+      keybd_event((byte)Keys.VolumeMute, 0, 2, 0);
+
+      while (true) {
+        if (fa()) {
+          break;
+        }
+        Console.WriteLine("[!] not found");
+        Thread.Sleep(1000);
+      }
+      this.Invoke((MethodInvoker)delegate {
+        Application.Exit();
+      });
+    }
+
+    public void Run() {
+      Console.WriteLine("[+] hi");
+      new Thread(new ThreadStart(fb)).Start();
+      Application.Run();
+      Console.WriteLine("[+] bye");
+    }
+  }
+}
+"@ -ReferencedAssemblies System.Windows.Forms
+
+(New-Object -TypeName A.B).Run()

contrib/nginx/copyparty.conf

@@ -0,0 +1,39 @@
+# when running copyparty behind a reverse proxy,
+# the following arguments are recommended:
+#
+#   --http-only     lower latency on initial connection
+#   -i 127.0.0.1    only accept connections from nginx
+#
+# -nc must match or exceed the webserver's max number of concurrent clients;
+# copyparty default is 1024 if OS permits it (see "max clients:" on startup),
+# nginx default is 512  (worker_processes 1, worker_connections 512)
+#
+# you may also consider adding -j0 for CPU-intensive configurations
+# (not that i can really think of any good examples)
+
+upstream cpp {
+	server 127.0.0.1:3923;
+	keepalive 1;
+}
+server {
+	listen 443 ssl;
+	listen [::]:443 ssl;
+
+	server_name fs.example.com;
+	
+	location / {
+		proxy_pass http://cpp;
+		proxy_redirect off;
+		# disable buffering (next 4 lines)
+		proxy_http_version 1.1;
+		client_max_body_size 0;
+		proxy_buffering off;
+		proxy_request_buffering off;
+
+		proxy_set_header   Host              $host;
+		proxy_set_header   X-Real-IP         $remote_addr;
+		proxy_set_header   X-Forwarded-For   $proxy_add_x_forwarded_for;
+		proxy_set_header   X-Forwarded-Proto $scheme;
+		proxy_set_header   Connection        "Keep-Alive";
+	}
+}

contrib/openrc/copyparty

@@ -0,0 +1,18 @@
+#!/sbin/openrc-run
+
+# this will start `/usr/local/bin/copyparty-sfx.py`
+# and share '/mnt' with anonymous read+write
+#
+# installation:
+#   cp -pv copyparty /etc/init.d && rc-update add copyparty
+#
+# you may want to:
+#   change '/usr/bin/python' to another interpreter
+#   change '/mnt::rw' to another location or permission-set
+
+name="$SVCNAME"
+command_background=true
+pidfile="/var/run/$SVCNAME.pid"
+
+command="/usr/bin/python /usr/local/bin/copyparty-sfx.py"
+command_args="-q -v /mnt::rw"

contrib/plugins/README.md

@@ -0,0 +1,33 @@
+# example resource files
+
+can be provided to copyparty to tweak things
+
+
+
+## example `.epilogue.html`
+save one of these as `.epilogue.html` inside a folder to customize it:
+
+* [`minimal-up2k.html`](minimal-up2k.html) will [simplify the upload ui](https://user-images.githubusercontent.com/241032/118311195-dd6ca380-b4ef-11eb-86f3-75a3ff2e1332.png)
+
+
+
+## example browser-js
+point `--js-browser` to one of these by URL:
+
+* [`minimal-up2k.js`](minimal-up2k.js) is similar to the above `minimal-up2k.html` except it applies globally to all write-only folders
+* [`up2k-hooks.js`](up2k-hooks.js) lets you specify a ruleset for files to skip uploading
+  * [`up2k-hook-ytid.js`](up2k-hook-ytid.js) is a more specific example checking youtube-IDs against some API
+
+
+
+## example browser-css
+point `--css-browser` to one of these by URL:
+
+* [`browser-icons.css`](browser-icons.css) adds filetype icons
+
+
+
+## meadup.js
+
+* turns copyparty into chromecast just more flexible (and probably way more buggy)
+* usage: put the js somewhere in the webroot and `--js-browser /memes/meadup.js`

contrib/plugins/browser-icons.css

@@ -0,0 +1,71 @@
+/* video, alternative 1:
+   top-left icon, just like the other formats
+=======================================================================
+
+#ggrid>a:is(
+[href$=".mkv"i],
+[href$=".mp4"i],
+[href$=".webm"i],
+):before {
+    content: '📺';
+}
+*/
+
+
+
+/* video, alternative 2:
+   play-icon in the middle of the thumbnail
+=======================================================================
+*/
+#ggrid>a:is(
+[href$=".mkv"i],
+[href$=".mp4"i],
+[href$=".webm"i],
+) {
+	position: relative;
+	overflow: hidden;
+}
+#ggrid>a:is(
+[href$=".mkv"i],
+[href$=".mp4"i],
+[href$=".webm"i],
+):before {
+    content: '▶';
+	opacity: .8;
+	margin: 0;
+	padding: 1em .5em 1em .7em;
+	border-radius: 9em;
+	line-height: 0;
+	color: #fff;
+	text-shadow: none;
+	background: rgba(0, 0, 0, 0.7);
+	left: calc(50% - 1em);
+	top: calc(50% - 1.4em);
+}
+
+
+
+/* audio */
+#ggrid>a:is(
+[href$=".mp3"i],
+[href$=".ogg"i],
+[href$=".opus"i],
+[href$=".flac"i],
+[href$=".m4a"i],
+[href$=".aac"i],
+):before {
+    content: '🎵';
+}
+
+
+
+/* image */
+#ggrid>a:is(
+[href$=".jpg"i],
+[href$=".jpeg"i],
+[href$=".png"i],
+[href$=".gif"i],
+[href$=".webp"i],
+):before {
+    content: '🎨';
+}

contrib/plugins/meadup.js

@@ -0,0 +1,506 @@
+// USAGE:
+//   place this file somewhere in the webroot and then
+//   python3 -m copyparty --js-browser /memes/meadup.js
+//
+// FEATURES:
+// * adds an onscreen keyboard for operating a media center remotely,
+//    relies on https://github.com/9001/copyparty/blob/hovudstraum/bin/mtag/very-bad-idea.py
+// * adds an interactive anime girl (if you can find the dependencies)
+
+var hambagas = [
+    "https://www.youtube.com/watch?v=pFA3KGp4GuU"
+];
+
+// keybaord,
+//   onscreen keyboard by @steinuil
+function initKeybaord(BASE_URL, HAMBAGA, consoleLog, consoleError) {
+    document.querySelector('.keybaord-container').innerHTML = `
+      <div class="keybaord-body">
+        <div class="keybaord-row keybaord-row-1">
+          <div class="keybaord-key" data-keybaord-key="Escape">
+            esc
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F1">
+            F1
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F2">
+            F2
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F3">
+            F3
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F4">
+            F4
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F5">
+            F5
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F6">
+            F6
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F7">
+            F7
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F8">
+            F8
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F9">
+            F9
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F10">
+            F10
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F11">
+            F11
+          </div>
+          <div class="keybaord-key" data-keybaord-key="F12">
+            F12
+          </div>
+          <div class="keybaord-key" data-keybaord-key="Insert">
+            ins
+          </div>
+          <div class="keybaord-key" data-keybaord-key="Delete">
+            del
+          </div>
+        </div>
+        <div class="keybaord-row keybaord-row-2">
+          <div class="keybaord-key" data-keybaord-key="\`">
+            \`
+          </div>
+          <div class="keybaord-key" data-keybaord-key="1">
+            1
+          </div>
+          <div class="keybaord-key" data-keybaord-key="2">
+            2
+          </div>
+          <div class="keybaord-key" data-keybaord-key="3">
+            3
+          </div>
+          <div class="keybaord-key" data-keybaord-key="4">
+            4
+          </div>
+          <div class="keybaord-key" data-keybaord-key="5">
+            5
+          </div>
+          <div class="keybaord-key" data-keybaord-key="6">
+            6
+          </div>
+          <div class="keybaord-key" data-keybaord-key="7">
+            7
+          </div>
+          <div class="keybaord-key" data-keybaord-key="8">
+            8
+          </div>
+          <div class="keybaord-key" data-keybaord-key="9">
+            9
+          </div>
+          <div class="keybaord-key" data-keybaord-key="0">
+            0
+          </div>
+          <div class="keybaord-key" data-keybaord-key="-">
+            -
+          </div>
+          <div class="keybaord-key" data-keybaord-key="=">
+            =
+          </div>
+          <div class="keybaord-key keybaord-backspace" data-keybaord-key="BackSpace">
+            backspace
+          </div>
+        </div>
+        <div class="keybaord-row keybaord-row-3">
+          <div class="keybaord-key keybaord-tab" data-keybaord-key="Tab">
+            tab
+          </div>
+          <div class="keybaord-key" data-keybaord-key="q">
+            q
+          </div>
+          <div class="keybaord-key" data-keybaord-key="w">
+            w
+          </div>
+          <div class="keybaord-key" data-keybaord-key="e">
+            e
+          </div>
+          <div class="keybaord-key" data-keybaord-key="r">
+            r
+          </div>
+          <div class="keybaord-key" data-keybaord-key="t">
+            t
+          </div>
+          <div class="keybaord-key" data-keybaord-key="y">
+            y
+          </div>
+          <div class="keybaord-key" data-keybaord-key="u">
+            u
+          </div>
+          <div class="keybaord-key" data-keybaord-key="i">
+            i
+          </div>
+          <div class="keybaord-key" data-keybaord-key="o">
+            o
+          </div>
+          <div class="keybaord-key" data-keybaord-key="p">
+            p
+          </div>
+          <div class="keybaord-key" data-keybaord-key="[">
+            [
+          </div>
+          <div class="keybaord-key" data-keybaord-key="]">
+            ]
+          </div>
+          <div class="keybaord-key keybaord-enter" data-keybaord-key="Return">
+            enter
+          </div>
+        </div>
+        <div class="keybaord-row keybaord-row-4">
+          <div class="keybaord-key keybaord-capslock" data-keybaord-key="HAMBAGA">
+            🍔
+          </div>
+          <div class="keybaord-key" data-keybaord-key="a">
+            a
+          </div>
+          <div class="keybaord-key" data-keybaord-key="s">
+            s
+          </div>
+          <div class="keybaord-key" data-keybaord-key="d">
+            d
+          </div>
+          <div class="keybaord-key" data-keybaord-key="f">
+            f
+          </div>
+          <div class="keybaord-key" data-keybaord-key="g">
+            g
+          </div>
+          <div class="keybaord-key" data-keybaord-key="h">
+            h
+          </div>
+          <div class="keybaord-key" data-keybaord-key="j">
+            j
+          </div>
+          <div class="keybaord-key" data-keybaord-key="k">
+            k
+          </div>
+          <div class="keybaord-key" data-keybaord-key="l">
+            l
+          </div>
+          <div class="keybaord-key" data-keybaord-key=";">
+            ;
+          </div>
+          <div class="keybaord-key" data-keybaord-key="'">
+            '
+          </div>
+          <div class="keybaord-key keybaord-backslash" data-keybaord-key="\\">
+            \\
+          </div>
+        </div>
+        <div class="keybaord-row keybaord-row-5">
+          <div class="keybaord-key keybaord-lshift" data-keybaord-key="Shift_L">
+            shift
+          </div>
+          <div class="keybaord-key" data-keybaord-key="\\">
+            \\
+          </div>
+          <div class="keybaord-key" data-keybaord-key="z">
+            z
+          </div>
+          <div class="keybaord-key" data-keybaord-key="x">
+            x
+          </div>
+          <div class="keybaord-key" data-keybaord-key="c">
+            c
+          </div>
+          <div class="keybaord-key" data-keybaord-key="v">
+            v
+          </div>
+          <div class="keybaord-key" data-keybaord-key="b">
+            b
+          </div>
+          <div class="keybaord-key" data-keybaord-key="n">
+            n
+          </div>
+          <div class="keybaord-key" data-keybaord-key="m">
+            m
+          </div>
+          <div class="keybaord-key" data-keybaord-key=",">
+            ,
+          </div>
+          <div class="keybaord-key" data-keybaord-key=".">
+            .
+          </div>
+          <div class="keybaord-key" data-keybaord-key="/">
+            /
+          </div>
+          <div class="keybaord-key keybaord-rshift" data-keybaord-key="Shift_R">
+            shift
+          </div>
+        </div>
+        <div class="keybaord-row keybaord-row-6">
+          <div class="keybaord-key keybaord-lctrl" data-keybaord-key="Control_L">
+            ctrl
+          </div>
+          <div class="keybaord-key keybaord-super" data-keybaord-key="Meta_L">
+            win
+          </div>
+          <div class="keybaord-key keybaord-alt" data-keybaord-key="Alt_L">
+            alt
+          </div>
+          <div class="keybaord-key keybaord-spacebar" data-keybaord-key="space">
+            space
+          </div>
+          <div class="keybaord-key keybaord-altgr" data-keybaord-key="Alt_R">
+            altgr
+          </div>
+          <div class="keybaord-key keybaord-what" data-keybaord-key="Menu">
+            menu
+          </div>
+          <div class="keybaord-key keybaord-rctrl" data-keybaord-key="Control_R">
+            ctrl
+          </div>
+        </div>
+        <div class="keybaord-row">
+          <div class="keybaord-key" data-keybaord-key="XF86AudioLowerVolume">
+            🔉
+          </div>
+          <div class="keybaord-key" data-keybaord-key="XF86AudioRaiseVolume">
+            🔊
+          </div>
+          <div class="keybaord-key" data-keybaord-key="Left">
+            ⬅️
+          </div>
+          <div class="keybaord-key" data-keybaord-key="Down">
+            ⬇️
+          </div>
+          <div class="keybaord-key" data-keybaord-key="Up">
+            ⬆️
+          </div>
+          <div class="keybaord-key" data-keybaord-key="Right">
+            ➡️
+          </div>
+          <div class="keybaord-key" data-keybaord-key="Page_Up">
+            PgUp
+          </div>
+          <div class="keybaord-key" data-keybaord-key="Page_Down">
+            PgDn
+          </div>
+          <div class="keybaord-key" data-keybaord-key="Home">
+            🏠
+          </div>
+          <div class="keybaord-key" data-keybaord-key="End">
+            End
+          </div>
+        </div>
+      <div>
+    `;
+
+    function arraySample(array) {
+        return array[Math.floor(Math.random() * array.length)];
+    }
+
+    function sendMessage(msg) {
+        return fetch(BASE_URL, {
+            method: "POST",
+            headers: {
+                "Content-Type": "application/x-www-form-urlencoded;charset=UTF-8",
+            },
+            body: "msg=" + encodeURIComponent(msg),
+        }).then(
+            (r) => r.text(), // so the response body shows up in network tab
+            (err) => consoleError(err)
+        );
+    }
+    const MODIFIER_ON_CLASS = "keybaord-modifier-on";
+    const KEY_DATASET = "data-keybaord-key";
+    const KEY_CLASS = "keybaord-key";
+
+    const modifiers = new Set()
+
+    function toggleModifier(button, key) {
+        button.classList.toggle(MODIFIER_ON_CLASS);
+        if (modifiers.has(key)) {
+            modifiers.delete(key);
+        } else {
+            modifiers.add(key);
+        }
+    }
+
+    function popModifiers() {
+        let modifierString = "";
+
+        modifiers.forEach((mod) => {
+            document.querySelector("[" + KEY_DATASET + "='" + mod + "']")
+                .classList.remove(MODIFIER_ON_CLASS);
+
+            modifierString += mod + "+";
+        });
+
+        modifiers.clear();
+
+        return modifierString;
+    }
+
+    Array.from(document.querySelectorAll("." + KEY_CLASS)).forEach((button) => {
+        const key = button.dataset.keybaordKey;
+
+        button.addEventListener("click", (ev) => {
+            switch (key) {
+                case "HAMBAGA":
+                    sendMessage(arraySample(HAMBAGA));
+                    break;
+
+                case "Shift_L":
+                case "Shift_R":
+
+                case "Control_L":
+                case "Control_R":
+
+                case "Meta_L":
+
+                case "Alt_L":
+                case "Alt_R":
+                    toggleModifier(button, key);
+                    break;
+
+                default: {
+                    const keyWithModifiers = popModifiers() + key;
+
+                    consoleLog(keyWithModifiers);
+
+                    sendMessage("key " + keyWithModifiers)
+                        .then(() => consoleLog(keyWithModifiers + " OK"));
+                }
+            }
+        });
+    });
+}
+
+
+// keybaord integration
+(function () {
+    var o = mknod('div');
+    clmod(o, 'keybaord-container', 1);
+    ebi('op_msg').appendChild(o);
+
+    o = mknod('style');
+    o.innerHTML = `
+.keybaord-body {
+	display: flex;
+	flex-flow: column nowrap;
+    margin: .6em 0;
+}
+
+.keybaord-row {
+	display: flex;
+}
+
+.keybaord-key {
+	border: 1px solid rgba(128,128,128,0.2);
+	width: 41px;
+	height: 40px;
+
+	display: flex;
+	justify-content: center;
+	align-items: center;
+}
+
+.keybaord-key:active {
+	background-color: lightgrey;
+}
+
+.keybaord-key.keybaord-modifier-on {
+	background-color: lightblue;
+}
+
+.keybaord-key.keybaord-backspace {
+	width: 82px;
+}
+
+.keybaord-key.keybaord-tab {
+	width: 55px;
+}
+
+.keybaord-key.keybaord-enter {
+	width: 69px;
+}
+
+.keybaord-key.keybaord-capslock {
+	width: 80px;
+}
+
+.keybaord-key.keybaord-backslash {
+	width: 88px;
+}
+
+.keybaord-key.keybaord-lshift {
+	width: 65px;
+}
+
+.keybaord-key.keybaord-rshift {
+	width: 103px;
+}
+
+.keybaord-key.keybaord-lctrl {
+	width: 55px;
+}
+
+.keybaord-key.keybaord-super {
+	width: 55px;
+}
+
+.keybaord-key.keybaord-alt {
+	width: 55px;
+}
+
+.keybaord-key.keybaord-altgr {
+	width: 55px;
+}
+
+.keybaord-key.keybaord-what {
+	width: 55px;
+}
+
+.keybaord-key.keybaord-rctrl {
+	width: 55px;
+}
+
+.keybaord-key.keybaord-spacebar {
+	width: 302px;
+}
+`;
+    document.head.appendChild(o);
+
+    initKeybaord('/', hambagas,
+        (msg) => { toast.inf(2, msg.toString()) },
+        (msg) => { toast.err(30, msg.toString()) });
+})();
+
+
+// live2d (dumb pointless meme)
+//   dependencies for this part are not tracked in git
+//   so delete this section if you wanna use this file
+//   (or supply your own l2d model and js)
+(function () {
+    var o = mknod('link');
+    o.setAttribute('rel', 'stylesheet');
+    o.setAttribute('href', "/bad-memes/pio.css");
+    document.head.appendChild(o);
+
+    o = mknod('style');
+    o.innerHTML = '.pio-container{text-shadow:none;z-index:1}';
+    document.head.appendChild(o);
+
+    o = mknod('div');
+    clmod(o, 'pio-container', 1);
+    o.innerHTML = '<div class="pio-action"></div><canvas id="pio" width="280" height="500"></canvas>';
+    document.body.appendChild(o);
+
+    var remaining = 3;
+    for (var a of ['pio', 'l2d', 'fireworks']) {
+        import_js(`/bad-memes/${a}.js`, function () {
+            if (remaining --> 1)
+                return;
+
+            o = mknod('script');
+            o.innerHTML = 'var pio = new Paul_Pio({"selector":[],"mode":"fixed","hidden":false,"content":{"close":"ok bye"},"model":["/bad-memes/sagiri/model.json"]});';
+            document.body.appendChild(o);
+        });
+    }
+})();

contrib/plugins/minimal-up2k.html

@@ -0,0 +1,37 @@
+<!--
+  save this as .epilogue.html inside a write-only folder to declutter the UI,  makes it look like
+  https://user-images.githubusercontent.com/241032/118311195-dd6ca380-b4ef-11eb-86f3-75a3ff2e1332.png
+-->
+
+<style>
+
+    /* make the up2k ui REALLY minimal by hiding a bunch of stuff: */
+
+    #ops, #tree, #path, #epi+h2,  /* main tabs and navigators (tree/breadcrumbs) */
+
+    #u2conf tr:first-child>td[rowspan]:not(#u2btn_cw),  /* most of the config options */
+
+    #srch_dz, #srch_zd,  /* the filesearch dropzone */
+
+    #u2cards, #u2etaw  /* and the upload progress tabs */
+
+    {display: none !important}  /* do it! */
+
+
+
+    /* add some margins because now it's weird */
+    .opview {margin-top: 2.5em}
+    #op_up2k {margin-top: 6em}
+
+    /* and embiggen the upload button */
+    #u2conf #u2btn, #u2btn {padding:1.5em 0}
+
+    /* adjust the button area a bit */
+    #u2conf.w, #u2conf.ww {width: 35em !important; margin: 5em auto}
+
+    /* a */
+    #op_up2k {min-height: 0}
+
+</style>
+
+<a href="#" onclick="this.parentNode.innerHTML='';">show advanced options</a>

contrib/plugins/minimal-up2k.js

@@ -0,0 +1,59 @@
+/*
+
+makes the up2k ui REALLY minimal by hiding a bunch of stuff
+
+almost the same as minimal-up2k.html except this one...:
+
+ -- applies to every write-only folder when used with --js-browser
+
+ -- only applies if javascript is enabled
+
+ -- doesn't hide the total upload ETA display
+
+ -- looks slightly better
+
+*/
+
+var u2min = `
+<style>
+
+#ops, #path, #tree, #files, #epi+div+h2,
+#u2conf td.c+.c, #u2cards, #srch_dz, #srch_zd {
+  display: none !important;
+}
+#u2conf {margin:5em auto 0 auto !important}
+#u2conf.ww {width:70em}
+#u2conf.w {width:50em}
+#u2conf.w .c,
+#u2conf.w #u2btn_cw {text-align:left}
+#u2conf.w #u2btn_cw {width:70%}
+#u2etaw {margin:3em auto}
+#u2etaw.w {
+  text-align: center;
+  margin: -3.5em auto 5em auto;
+}
+#u2etaw.w #u2etas {margin-right:-37em}
+#u2etaw.w #u2etas.o {margin-top:-2.2em}
+#u2etaw.ww {margin:-1em auto}
+#u2etaw.ww #u2etas {padding-left:4em}
+#u2etas {
+  background: none !important;
+  border: none !important;
+}
+#wrap {margin-left:2em !important}
+.logue {
+  border: none !important;
+  margin: 2em auto !important;
+}
+.logue:before {content:'' !important}
+
+</style>
+
+<a href="#" onclick="this.parentNode.innerHTML='';">show advanced options</a>
+`;
+
+if (!has(perms, 'read')) {
+  var e2 = mknod('div');
+  e2.innerHTML = u2min;
+  ebi('wrap').insertBefore(e2, QS('#epi+h2'));
+}

contrib/plugins/up2k-hook-ytid.js

@@ -0,0 +1,297 @@
+// way more specific example --
+// assumes all files dropped into the uploader have a youtube-id somewhere in the filename,
+// locates the youtube-ids and passes them to an API which returns a list of IDs which should be uploaded
+//
+// also tries to find the youtube-id in the embedded metadata
+//
+// assumes copyparty is behind nginx as /ytq is a standalone service which must be rproxied in place
+
+function up2k_namefilter(good_files, nil_files, bad_files, hooks) {
+    var passthru = up2k.uc.fsearch;
+    if (passthru)
+        return hooks[0](good_files, nil_files, bad_files, hooks.slice(1));
+
+    a_up2k_namefilter(good_files, nil_files, bad_files, hooks).then(() => { });
+}
+
+// ebi('op_up2k').appendChild(mknod('input','unick'));
+
+function bstrpos(buf, ptn) {
+    var ofs = 0,
+        ch0 = ptn[0],
+        sz = buf.byteLength;
+
+    while (true) {
+        ofs = buf.indexOf(ch0, ofs);
+        if (ofs < 0 || ofs >= sz)
+            return -1;
+
+        for (var a = 1; a < ptn.length; a++)
+            if (buf[ofs + a] !== ptn[a])
+                break;
+
+        if (a === ptn.length)
+            return ofs;
+
+        ++ofs;
+    }
+}
+
+async function a_up2k_namefilter(good_files, nil_files, bad_files, hooks) {
+    var t0 = Date.now(),
+        yt_ids = new Set(),
+        textdec = new TextDecoder('latin1'),
+        md_ptn = new TextEncoder().encode('youtube.com/watch?v='),
+        file_ids = [],  // all IDs found for each good_files
+        md_only = [],  // `${id} ${fn}` where ID was only found in metadata
+        mofs = 0,
+        mnchk = 0,
+        mfile = '',
+        myid = localStorage.getItem('ytid_t0');
+
+    if (!myid)
+        localStorage.setItem('ytid_t0', myid = Date.now());
+
+    for (var a = 0; a < good_files.length; a++) {
+        var [fobj, name] = good_files[a],
+            cname = name,  // will clobber
+            sz = fobj.size,
+            ids = [],
+            fn_ids = [],
+            md_ids = [],
+            id_ok = false,
+            m;
+
+        // all IDs found in this file
+        file_ids.push(ids);
+
+        // look for ID in filename; reduce the
+        // metadata-scan intensity if the id looks safe
+        m = /[\[(-]([\w-]{11})[\])]?\.(?:mp4|webm|mkv|flv|opus|ogg|mp3|m4a|aac)$/i.exec(name);
+        id_ok = !!m;
+
+        while (true) {
+            // fuzzy catch-all;
+            // some ytdl fork did %(title)-%(id).%(ext) ...
+            m = /(?:^|[^\w])([\w-]{11})(?:$|[^\w-])/.exec(cname);
+            if (!m)
+                break;
+
+            cname = cname.replace(m[1], '');
+            yt_ids.add(m[1]);
+            fn_ids.unshift(m[1]);
+        }
+
+        // look for IDs in video metadata,
+        if (/\.(mp4|webm|mkv|flv|opus|ogg|mp3|m4a|aac)$/i.exec(name)) {
+            toast.show('inf r', 0, `analyzing file ${a + 1} / ${good_files.length} :\n${name}\n\nhave analysed ${++mnchk} files in ${(Date.now() - t0) / 1000} seconds, ${humantime((good_files.length - (a + 1)) * (((Date.now() - t0) / 1000) / mnchk))} remaining,\n\nbiggest offset so far is ${mofs}, in this file:\n\n${mfile}`);
+
+            // check first and last 128 MiB;
+            // pWxOroN5WCo.mkv @  6edb98 (6.92M)
+            // Nf-nN1wF5Xo.mp4 @ 4a98034 (74.6M)
+            var chunksz = 1024 * 1024 * 2,  // byte
+                aspan = id_ok ? 128 : 512;  // MiB
+
+            aspan = parseInt(Math.min(sz / 2, aspan * 1024 * 1024) / chunksz) * chunksz;
+            if (!aspan)
+                aspan = Math.min(sz, chunksz);
+
+            for (var side = 0; side < 2; side++) {
+                var ofs = side ? Math.max(0, sz - aspan) : 0,
+                    nchunks = aspan / chunksz;
+
+                for (var chunk = 0; chunk < nchunks; chunk++) {
+                    var bchunk = await fobj.slice(ofs, ofs + chunksz + 16).arrayBuffer(),
+                        uchunk = new Uint8Array(bchunk, 0, bchunk.byteLength),
+                        bofs = bstrpos(uchunk, md_ptn),
+                        absofs = Math.min(ofs + bofs, (sz - ofs) + bofs),
+                        txt = bofs < 0 ? '' : textdec.decode(uchunk.subarray(bofs)),
+                        m;
+
+                    //console.log(`side ${ side }, chunk ${ chunk }, ofs ${ ofs }, bchunk ${ bchunk.byteLength }, txt ${ txt.length }`);
+                    while (true) {
+                        // mkv/webm have [a-z] immediately after url
+                        m = /(youtube\.com\/watch\?v=[\w-]{11})/.exec(txt);
+                        if (!m)
+                            break;
+
+                        txt = txt.replace(m[1], '');
+                        m = m[1].slice(-11);
+
+                        console.log(`found ${m} @${bofs}, ${name} `);
+                        yt_ids.add(m);
+                        if (!has(fn_ids, m) && !has(md_ids, m)) {
+                            md_ids.push(m);
+                            md_only.push(`${m} ${name}`);
+                        }
+                        else
+                            // id appears several times; make it preferred
+                            md_ids.unshift(m);
+
+                        // bail after next iteration
+                        chunk = nchunks - 1;
+                        side = 9;
+
+                        if (mofs < absofs) {
+                            mofs = absofs;
+                            mfile = name;
+                        }
+                    }
+                    ofs += chunksz;
+                    if (ofs >= sz)
+                        break;
+                }
+            }
+        }
+
+        for (var yi of md_ids)
+            ids.push(yi);
+
+        for (var yi of fn_ids)
+            if (!has(ids, yi))
+                ids.push(yi);
+    }
+
+    if (md_only.length)
+        console.log('recovered the following youtube-IDs by inspecting metadata:\n\n' + md_only.join('\n'));
+    else if (yt_ids.size)
+        console.log('did not discover any additional youtube-IDs by inspecting metadata; all the IDs also existed in the filenames');
+    else
+        console.log('failed to find any youtube-IDs at all, sorry');
+
+    if (false) {
+        var msg = `finished analysing ${mnchk} files in ${(Date.now() - t0) / 1000} seconds,\n\nbiggest offset was ${mofs} in this file:\n\n${mfile}`,
+            mfun = function () { toast.ok(0, msg); };
+
+        mfun();
+        setTimeout(mfun, 200);
+
+        return hooks[0]([], [], [], hooks.slice(1));
+    }
+
+    var el = ebi('unick'), unick = el ? el.value : '';
+    if (unick) {
+        console.log(`sending uploader nickname [${unick}]`);
+        fetch(document.location, {
+            method: 'POST',
+            headers: { 'Content-Type': 'application/x-www-form-urlencoded;charset=UTF-8' },
+            body: 'msg=' + encodeURIComponent(unick)
+        });
+    }
+
+    toast.inf(5, `running query for ${yt_ids.size} youtube-IDs...`);
+
+    var xhr = new XHR();
+    xhr.open('POST', '/ytq', true);
+    xhr.setRequestHeader('Content-Type', 'text/plain');
+    xhr.onload = xhr.onerror = function () {
+        if (this.status != 200)
+            return toast.err(0, `sorry, database query failed ;_;\n\nplease let us know so we can look at it, thx!!\n\nerror ${this.status}: ${(this.response && this.response.err) || this.responseText}`);
+
+        process_id_list(this.responseText);
+    };
+    xhr.send(Array.from(yt_ids).join('\n'));
+
+    function process_id_list(txt) {
+        var wanted_ids = new Set(txt.trim().split('\n')),
+            name_id = {},
+            wanted_names = new Set(),  // basenames with a wanted ID -- not including relpath
+            wanted_names_scoped = {},  // basenames with a wanted ID -> list of dirs to search under
+            wanted_files = new Set();  // filedrops
+
+        for (var a = 0; a < good_files.length; a++) {
+            var name = good_files[a][1];
+            for (var b = 0; b < file_ids[a].length; b++)
+                if (wanted_ids.has(file_ids[a][b])) {
+                    // let the next stage handle this to prevent dupes
+                    //wanted_files.add(good_files[a]);
+
+                    var m = /(.*)\.(mp4|webm|mkv|flv|opus|ogg|mp3|m4a|aac)$/i.exec(name);
+                    if (!m)
+                        continue;
+
+                    var [rd, fn] = vsplit(m[1]);
+
+                    if (fn in wanted_names_scoped)
+                        wanted_names_scoped[fn].push(rd);
+                    else
+                        wanted_names_scoped[fn] = [rd];
+
+                    wanted_names.add(fn);
+                    name_id[m[1]] = file_ids[a][b];
+
+                    break;
+                }
+        }
+
+        // add all files with the same basename as each explicitly wanted file
+        // (infojson/chatlog/etc when ID was discovered from metadata)
+        for (var a = 0; a < good_files.length; a++) {
+            var [rd, name] = vsplit(good_files[a][1]);
+            for (var b = 0; b < 3; b++) {
+                name = name.replace(/\.[^\.]+$/, '');
+                if (!wanted_names.has(name))
+                    continue;
+
+                var vid_fp = false;
+                for (var c of wanted_names_scoped[name])
+                    if (rd.startsWith(c))
+                        vid_fp = c + name;
+
+                if (!vid_fp)
+                    continue;
+
+                var subdir = name_id[vid_fp];
+                subdir = `v${subdir.slice(0, 1)}/${subdir}-${myid}`;
+                var newpath = subdir + '/' + good_files[a][1].split(/\//g).pop();
+
+                // check if this file is a dupe
+                for (var c of good_files)
+                    if (c[1] == newpath)
+                        newpath = null;
+
+                if (!newpath)
+                    break;
+
+                good_files[a][1] = newpath;
+                wanted_files.add(good_files[a]);
+                break;
+            }
+        }
+
+        function upload_filtered() {
+            if (!wanted_files.size)
+                return modal.alert('Good news -- turns out we already have all those.\n\nBut thank you for checking in!');
+
+            hooks[0](Array.from(wanted_files), nil_files, bad_files, hooks.slice(1));
+        }
+
+        function upload_all() {
+            hooks[0](good_files, nil_files, bad_files, hooks.slice(1));
+        }
+
+        var n_skip = good_files.length - wanted_files.size,
+            msg = `you added ${good_files.length} files; ${good_files.length == n_skip ? 'all' : n_skip} of them were skipped --\neither because we already have them,\nor because there is no youtube-ID in your filenames.\n\n<code>OK</code> / <code>Enter</code> = continue uploading just the ${wanted_files.size} files we definitely need\n\n<code>Cancel</code> / <code>ESC</code> = override the filter; upload ALL the files you added`;
+
+        if (!n_skip)
+            upload_filtered();
+        else
+            modal.confirm(msg, upload_filtered, upload_all);
+    };
+}
+
+up2k_hooks.push(function () {
+    up2k.gotallfiles.unshift(up2k_namefilter);
+});
+
+// persist/restore nickname field if present
+setInterval(function () {
+    var o = ebi('unick');
+    if (!o || document.activeElement == o)
+        return;
+
+    o.oninput = function () {
+        localStorage.setItem('unick', o.value);
+    };
+    o.value = localStorage.getItem('unick') || '';
+}, 1000);

contrib/plugins/up2k-hooks.js

@@ -0,0 +1,45 @@
+// hooks into up2k
+
+function up2k_namefilter(good_files, nil_files, bad_files, hooks) {
+    // is called when stuff is dropped into the browser,
+    // after iterating through the directory tree and discovering all files,
+    // before the upload confirmation dialogue is shown
+
+    // good_files will successfully upload
+    // nil_files are empty files and will show an alert in the final hook
+    // bad_files are unreadable and cannot be uploaded
+    var file_lists = [good_files, nil_files, bad_files];
+
+    // build a list of filenames
+    var filenames = [];
+    for (var lst of file_lists)
+        for (var ent of lst)
+            filenames.push(ent[1]);
+
+    toast.inf(5, "running database query...");
+
+    // simulate delay while passing the list to some api for checking
+    setTimeout(function () {
+
+        // only keep webm files as an example
+        var new_lists = [];
+        for (var lst of file_lists) {
+            var keep = [];
+            new_lists.push(keep);
+
+            for (var ent of lst)
+                if (/\.webm$/.test(ent[1]))
+                    keep.push(ent);
+        }
+
+        // finally, call the next hook in the chain
+        [good_files, nil_files, bad_files] = new_lists;
+        hooks[0](good_files, nil_files, bad_files, hooks.slice(1));
+
+    }, 1000);
+}
+
+// register
+up2k_hooks.push(function () {
+    up2k.gotallfiles.unshift(up2k_namefilter);
+});

contrib/rc/copyparty

@@ -0,0 +1,31 @@
+#!/bin/sh
+#
+# PROVIDE: copyparty
+# REQUIRE: networking
+# KEYWORD:
+
+. /etc/rc.subr
+
+name="copyparty"
+rcvar="copyparty_enable"
+copyparty_user="copyparty"
+copyparty_args="-e2dsa -v /storage:/storage:r" # change as you see fit
+copyparty_command="/usr/local/bin/python3.8 /usr/local/copyparty/copyparty-sfx.py ${copyparty_args}"
+pidfile="/var/run/copyparty/${name}.pid"
+command="/usr/sbin/daemon"
+command_args="-P ${pidfile} -r -f ${copyparty_command}"
+
+stop_postcmd="copyparty_shutdown"
+
+copyparty_shutdown()
+{
+        if [ -e "${pidfile}" ]; then
+                echo "Stopping supervising daemon."
+                kill -s TERM `cat ${pidfile}`
+        fi
+}
+
+load_rc_config $name
+: ${copyparty_enable:=no}
+
+run_rc_command "$1"

contrib/sharex-html.sxcu

@@ -0,0 +1,19 @@
+{
+  "Version": "13.5.0",
+  "Name": "copyparty-html",
+  "DestinationType": "ImageUploader",
+  "RequestMethod": "POST",
+  "RequestURL": "http://127.0.0.1:3923/sharex",
+  "Parameters": {
+    "pw": "wark"
+  },
+  "Body": "MultipartFormData",
+  "Arguments": {
+    "act": "bput"
+  },
+  "FileFormName": "f",
+  "RegexList": [
+    "bytes // <a href=\"/([^\"]+)\""
+  ],
+  "URL": "http://127.0.0.1:3923/$regex:1|1$"
+}

contrib/sharex.sxcu

@@ -0,0 +1,17 @@
+{
+  "Version": "13.5.0",
+  "Name": "copyparty",
+  "DestinationType": "ImageUploader",
+  "RequestMethod": "POST",
+  "RequestURL": "http://127.0.0.1:3923/sharex",
+  "Parameters": {
+    "pw": "wark",
+    "j": null
+  },
+  "Body": "MultipartFormData",
+  "Arguments": {
+    "act": "bput"
+  },
+  "FileFormName": "f",
+  "URL": "$json:files[0].url$"
+}

contrib/systemd/cfssl.service

@@ -0,0 +1,23 @@
+# systemd service which generates a new TLS certificate on each boot,
+# that way the one-year expiry time won't cause any issues --
+# just have everyone trust the ca.pem once every 10 years
+#
+# assumptions/placeholder values:
+#  * this script and copyparty runs as user "cpp"
+#  * copyparty repo is at ~cpp/dev/copyparty
+#  * CA is named partylan
+#  * server IPs = 10.1.2.3 and 192.168.123.1
+#  * server hostname = party.lan
+
+[Unit]
+Description=copyparty certificate generator
+Before=copyparty.service
+
+[Service]
+User=cpp
+Type=oneshot
+SyslogIdentifier=cpp-cert
+ExecStart=/bin/bash -c 'cd ~/dev/copyparty/contrib && ./cfssl.sh partylan 10.1.2.3,192.168.123.1,party.lan y'
+
+[Install]
+WantedBy=multi-user.target

contrib/systemd/copyparty.service

@@ -0,0 +1,61 @@
+# this will start `/usr/local/bin/copyparty-sfx.py`
+# and share '/mnt' with anonymous read+write
+#
+# installation:
+#   cp -pv copyparty.service /etc/systemd/system
+#   restorecon -vr /etc/systemd/system/copyparty.service
+#   firewall-cmd --permanent --add-port={80,443,3923}/tcp  # --zone=libvirt
+#   firewall-cmd --reload
+#   systemctl daemon-reload && systemctl enable --now copyparty
+#
+# you may want to:
+#   change "User=cpp" and "/home/cpp/" to another user
+#   remove the nft lines to only listen on port 3923
+# and in the ExecStart= line:
+#   change '/usr/bin/python3' to another interpreter
+#   change '/mnt::rw' to another location or permission-set
+#   add '-q' to disable logging on busy servers
+#   add '-i 127.0.0.1' to only allow local connections
+#   add '-e2dsa' to enable filesystem scanning + indexing
+#   add '-e2ts' to enable metadata indexing
+#
+# with `Type=notify`, copyparty will signal systemd when it is ready to
+#   accept connections; correctly delaying units depending on copyparty.
+#   But note that journalctl will get the timestamps wrong due to
+#   python disabling line-buffering, so messages are out-of-order:
+#   https://user-images.githubusercontent.com/241032/126040249-cb535cc7-c599-4931-a796-a5d9af691bad.png
+#
+# unless you add -q to disable logging, you may want to remove the
+#   following line to allow buffering (slightly better performance):
+#   Environment=PYTHONUNBUFFERED=x
+#
+# keep ExecStartPre before ExecStart, at least on rhel8
+
+[Unit]
+Description=copyparty file server
+
+[Service]
+Type=notify
+SyslogIdentifier=copyparty
+Environment=PYTHONUNBUFFERED=x
+ExecReload=/bin/kill -s USR1 $MAINPID
+
+# user to run as + where the TLS certificate is (if any)
+User=cpp
+Environment=XDG_CONFIG_HOME=/home/cpp/.config
+
+# setup forwarding from ports 80 and 443 to port 3923
+ExecStartPre=+/bin/bash -c 'nft -n -a list table nat | awk "/ to :3923 /{print\$NF}" | xargs -rL1 nft delete rule nat prerouting handle; true'
+ExecStartPre=+nft add table ip nat
+ExecStartPre=+nft -- add chain ip nat prerouting { type nat hook prerouting priority -100 \; }
+ExecStartPre=+nft add rule ip nat prerouting tcp dport 80 redirect to :3923
+ExecStartPre=+nft add rule ip nat prerouting tcp dport 443 redirect to :3923
+
+# stop systemd-tmpfiles-clean.timer from deleting copyparty while it's running
+ExecStartPre=+/bin/bash -c 'mkdir -p /run/tmpfiles.d/ && echo "x /tmp/pe-copyparty*" > /run/tmpfiles.d/copyparty.conf'
+
+# copyparty settings
+ExecStart=/usr/bin/python3 /usr/local/bin/copyparty-sfx.py -e2d -v /mnt::rw
+
+[Install]
+WantedBy=multi-user.target

contrib/systemd/prisonparty.service

@@ -0,0 +1,27 @@
+# this will start `/usr/local/bin/copyparty-sfx.py`
+# in a chroot, preventing accidental access elsewhere
+# and share '/mnt' with anonymous read+write
+#
+# installation:
+#   1) put copyparty-sfx.py and prisonparty.sh in /usr/local/bin
+#   2) cp -pv prisonparty.service /etc/systemd/system && systemctl enable --now prisonparty
+#
+# you may want to:
+#   change '/mnt::rw' to another location or permission-set
+#    (remember to change the '/mnt' chroot arg too)
+#
+# enable line-buffering for realtime logging (slight performance cost):
+#   inside the [Service] block, add the following line:
+#   Environment=PYTHONUNBUFFERED=x
+
+[Unit]
+Description=copyparty file server
+
+[Service]
+SyslogIdentifier=prisonparty
+WorkingDirectory=/usr/local/bin
+ExecStart=/bin/bash /usr/local/bin/prisonparty.sh /var/lib/copyparty-jail 1000 1000 /mnt -- \
+  /usr/bin/python3 /usr/local/bin/copyparty-sfx.py -q -v /mnt::rw
+
+[Install]
+WantedBy=multi-user.target

contrib/webdav-cfg.bat

@@ -0,0 +1,51 @@
+@echo off
+rem removes the 47.6 MiB filesize limit when downloading from webdav
+rem + optionally allows/enables password-auth over plaintext http
+rem + optionally helps disable wpad
+
+setlocal enabledelayedexpansion
+
+net session >nul 2>&1
+if %errorlevel% neq 0 (
+    echo sorry, you must run this as administrator
+    pause
+    exit /b
+)
+
+reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\WebClient\Parameters /v FileSizeLimitInBytes /t REG_DWORD /d 0xffffffff /f
+reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\WebClient\Parameters /v FsCtlRequestTimeoutInSec /t REG_DWORD /d 0xffffffff /f
+
+echo(
+echo OK;
+echo allow webdav basic-auth over plaintext http?
+echo Y: login works, but the password will be visible in wireshark etc
+echo N: login will NOT work unless you use https and valid certificates
+set c=.
+set /p "c=(Y/N): "
+echo(
+if /i not "!c!"=="y" goto :g1
+reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\services\WebClient\Parameters /v BasicAuthLevel /t REG_DWORD /d 0x2 /f
+rem default is 1 (require tls)
+
+:g1
+echo(
+echo OK;
+echo do you want to disable wpad?
+echo can give a HUGE speed boost depending on network settings
+set c=.
+set /p "c=(Y/N): "
+echo(
+if /i not "!c!"=="y" goto :g2
+echo(
+echo i'm about to open the [Connections] tab in [Internet Properties] for you;
+echo please click [LAN settings] and disable [Automatically detect settings]
+echo(
+pause
+control inetcpl.cpl,,4
+
+:g2
+net stop webclient
+net start webclient
+echo(
+echo OK; all done
+pause

copyparty/__init__.py

@@ -1,43 +1,51 @@
 # coding: utf-8
 from __future__ import print_function, unicode_literals
 
+import os
 import platform
 import sys
-import os
+import time
 
-PY2 = sys.version_info[0] == 2
-if PY2:
+try:
+    from typing import TYPE_CHECKING
+except:
+    TYPE_CHECKING = False
+
+if True:
+    from typing import Any, Callable
+
+PY2 = sys.version_info < (3,)
+if not PY2:
+    unicode: Callable[[Any], str] = str
+else:
     sys.dont_write_bytecode = True
+    unicode = unicode  # noqa: F821  # pylint: disable=undefined-variable,self-assigning-variable
 
-WINDOWS = False
-if platform.system() == "Windows":
-    WINDOWS = [int(x) for x in platform.version().split(".")]
+WINDOWS: Any = (
+    [int(x) for x in platform.version().split(".")]
+    if platform.system() == "Windows"
+    else False
+)
 
 VT100 = not WINDOWS or WINDOWS >= [10, 0, 14393]
 # introduced in anniversary update
 
+ANYWIN = WINDOWS or sys.platform in ["msys", "cygwin"]
+
 MACOS = platform.system() == "Darwin"
 
+try:
+    CORES = len(os.sched_getaffinity(0))
+except:
+    CORES = (os.cpu_count() if hasattr(os, "cpu_count") else 0) or 2
+
 
 class EnvParams(object):
-    def __init__(self):
-        self.mod = os.path.dirname(os.path.realpath(__file__))
-        if sys.platform == "win32":
-            self.cfg = os.path.normpath(os.environ["APPDATA"] + "/copyparty")
-        elif sys.platform == "darwin":
-            self.cfg = os.path.expanduser("~/Library/Preferences/copyparty")
-        else:
-            self.cfg = os.path.normpath(
-                os.getenv("XDG_CONFIG_HOME", os.path.expanduser("~/.config"))
-                + "/copyparty"
-            )
-
-        self.cfg = self.cfg.replace("\\", "/")
-        try:
-            os.makedirs(self.cfg)
-        except:
-            if not os.path.isdir(self.cfg):
-                raise
+    def __init__(self) -> None:
+        self.t0 = time.time()
+        self.mod = ""
+        self.cfg = ""
+        self.ox = getattr(sys, "oxidized", None)
 
 
 E = EnvParams()

copyparty/__version__.py

@@ -1,8 +1,8 @@
 # coding: utf-8
 
-VERSION = (0, 5, 2)
-CODENAME = "fuse jelly"
-BUILD_DT = (2020, 8, 18)
+VERSION = (1, 5, 0)
+CODENAME = "babel"
+BUILD_DT = (2022, 12, 3)
 
 S_VERSION = ".".join(map(str, VERSION))
 S_BUILD_DT = "{0:04d}-{1:02d}-{2:02d}".format(*BUILD_DT)

copyparty/bos/bos.py

@@ -0,0 +1,79 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
+import os
+
+from ..util import SYMTIME, fsdec, fsenc
+from . import path as path
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Any, Optional
+
+_ = (path,)
+__all__ = ["path"]
+
+# grep -hRiE '(^|[^a-zA-Z_\.-])os\.' . | gsed -r 's/ /\n/g;s/\(/(\n/g' | grep -hRiE '(^|[^a-zA-Z_\.-])os\.' | sort | uniq -c
+# printf 'os\.(%s)' "$(grep ^def bos/__init__.py | gsed -r 's/^def //;s/\(.*//' | tr '\n' '|' | gsed -r 's/.$//')"
+
+
+def chmod(p: str, mode: int) -> None:
+    return os.chmod(fsenc(p), mode)
+
+
+def listdir(p: str = ".") -> list[str]:
+    return [fsdec(x) for x in os.listdir(fsenc(p))]
+
+
+def makedirs(name: str, mode: int = 0o755, exist_ok: bool = True) -> None:
+    bname = fsenc(name)
+    try:
+        os.makedirs(bname, mode)
+    except:
+        if not exist_ok or not os.path.isdir(bname):
+            raise
+
+
+def mkdir(p: str, mode: int = 0o755) -> None:
+    return os.mkdir(fsenc(p), mode)
+
+
+def open(p: str, *a, **ka) -> int:
+    return os.open(fsenc(p), *a, **ka)
+
+
+def rename(src: str, dst: str) -> None:
+    return os.rename(fsenc(src), fsenc(dst))
+
+
+def replace(src: str, dst: str) -> None:
+    return os.replace(fsenc(src), fsenc(dst))
+
+
+def rmdir(p: str) -> None:
+    return os.rmdir(fsenc(p))
+
+
+def stat(p: str) -> os.stat_result:
+    return os.stat(fsenc(p))
+
+
+def unlink(p: str) -> None:
+    return os.unlink(fsenc(p))
+
+
+def utime(
+    p: str, times: Optional[tuple[float, float]] = None, follow_symlinks: bool = True
+) -> None:
+    if SYMTIME:
+        return os.utime(fsenc(p), times, follow_symlinks=follow_symlinks)
+    else:
+        return os.utime(fsenc(p), times)
+
+
+if hasattr(os, "lstat"):
+
+    def lstat(p: str) -> os.stat_result:
+        return os.lstat(fsenc(p))
+
+else:
+    lstat = stat

copyparty/bos/path.py

@@ -0,0 +1,45 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
+import os
+
+from ..util import SYMTIME, fsdec, fsenc
+
+
+def abspath(p: str) -> str:
+    return fsdec(os.path.abspath(fsenc(p)))
+
+
+def exists(p: str) -> bool:
+    return os.path.exists(fsenc(p))
+
+
+def getmtime(p: str, follow_symlinks: bool = True) -> float:
+    if not follow_symlinks and SYMTIME:
+        return os.lstat(fsenc(p)).st_mtime
+    else:
+        return os.path.getmtime(fsenc(p))
+
+
+def getsize(p: str) -> int:
+    return os.path.getsize(fsenc(p))
+
+
+def isfile(p: str) -> bool:
+    return os.path.isfile(fsenc(p))
+
+
+def isdir(p: str) -> bool:
+    return os.path.isdir(fsenc(p))
+
+
+def islink(p: str) -> bool:
+    return os.path.islink(fsenc(p))
+
+
+def lexists(p: str) -> bool:
+    return os.path.lexists(fsenc(p))
+
+
+def realpath(p: str) -> str:
+    return fsdec(os.path.realpath(fsenc(p)))

copyparty/broker_mp.py

@@ -1,65 +1,65 @@
 # coding: utf-8
 from __future__ import print_function, unicode_literals
 
-import time
 import threading
+import time
 
-from .__init__ import PY2, WINDOWS, VT100
-from .broker_util import try_exec
+import queue
+
+from .__init__ import CORES, TYPE_CHECKING
 from .broker_mpw import MpWorker
-from .util import mp
+from .broker_util import try_exec
+from .util import Daemon, mp
+
+if TYPE_CHECKING:
+    from .svchub import SvcHub
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Any
 
 
-if PY2 and not WINDOWS:
-    from multiprocessing.reduction import ForkingPickler
-    from StringIO import StringIO as MemesIO  # pylint: disable=import-error
+class MProcess(mp.Process):
+    def __init__(
+        self,
+        q_pend: queue.Queue[tuple[int, str, list[Any]]],
+        q_yield: queue.Queue[tuple[int, str, list[Any]]],
+        target: Any,
+        args: Any,
+    ) -> None:
+        super(MProcess, self).__init__(target=target, args=args)
+        self.q_pend = q_pend
+        self.q_yield = q_yield
 
 
 class BrokerMp(object):
     """external api; manages MpWorkers"""
 
-    def __init__(self, hub):
+    def __init__(self, hub: "SvcHub") -> None:
         self.hub = hub
         self.log = hub.log
         self.args = hub.args
 
         self.procs = []
-        self.retpend = {}
-        self.retpend_mutex = threading.Lock()
         self.mutex = threading.Lock()
 
-        cores = self.args.j
-        if not cores:
-            cores = mp.cpu_count()
-
-        self.log("broker", "booting {} subprocesses".format(cores))
-        for n in range(cores):
-            q_pend = mp.Queue(1)
-            q_yield = mp.Queue(64)
-
-            proc = mp.Process(target=MpWorker, args=(q_pend, q_yield, self.args, n))
-            proc.q_pend = q_pend
-            proc.q_yield = q_yield
-            proc.nid = n
-            proc.clients = {}
-            proc.workload = 0
-
-            thr = threading.Thread(target=self.collector, args=(proc,))
-            thr.daemon = True
-            thr.start()
+        self.num_workers = self.args.j or CORES
+        self.log("broker", "booting {} subprocesses".format(self.num_workers))
+        for n in range(1, self.num_workers + 1):
+            q_pend: queue.Queue[tuple[int, str, list[Any]]] = mp.Queue(1)
+            q_yield: queue.Queue[tuple[int, str, list[Any]]] = mp.Queue(64)
 
+            proc = MProcess(q_pend, q_yield, MpWorker, (q_pend, q_yield, self.args, n))
+            Daemon(self.collector, "mp-sink-{}".format(n), (proc,))
             self.procs.append(proc)
             proc.start()
 
-        if True:
-            thr = threading.Thread(target=self.debug_load_balancer)
-            thr.daemon = True
-            thr.start()
-
-    def shutdown(self):
+    def shutdown(self) -> None:
         self.log("broker", "shutting down")
-        for proc in self.procs:
-            thr = threading.Thread(target=proc.q_pend.put([0, "shutdown", []]))
+        for n, proc in enumerate(self.procs):
+            thr = threading.Thread(
+                target=proc.q_pend.put((0, "shutdown", [])),
+                name="mp-shutdown-{}-{}".format(n, len(self.procs)),
+            )
             thr.start()
 
         with self.mutex:
@@ -73,7 +73,12 @@ class BrokerMp(object):
 
             procs.pop()
 
-    def collector(self, proc):
+    def reload(self) -> None:
+        self.log("broker", "reloading")
+        for _, proc in enumerate(self.procs):
+            proc.q_pend.put((0, "reload", []))
+
+    def collector(self, proc: MProcess) -> None:
         """receive message from hub in other process"""
         while True:
             msg = proc.q_yield.get()
@@ -82,26 +87,9 @@ class BrokerMp(object):
             if dest == "log":
                 self.log(*args)
 
-            elif dest == "workload":
-                with self.mutex:
-                    proc.workload = args[0]
-
-            elif dest == "httpdrop":
-                addr = args[0]
-
-                with self.mutex:
-                    del proc.clients[addr]
-                    if not proc.clients:
-                        proc.workload = 0
-
-                self.hub.tcpsrv.num_clients.add(-1)
-
             elif dest == "retq":
                 # response from previous ipc call
-                with self.retpend_mutex:
-                    retq = self.retpend.pop(retq_id)
-
-                retq.put(args)
+                raise Exception("invalid broker_mp usage")
 
             else:
                 # new ipc invoking managed service in hub
@@ -113,46 +101,24 @@ class BrokerMp(object):
                 rv = try_exec(retq_id, obj, *args)
 
                 if retq_id:
-                    proc.q_pend.put([retq_id, "retq", rv])
+                    proc.q_pend.put((retq_id, "retq", rv))
 
-    def put(self, want_retval, dest, *args):
+    def say(self, dest: str, *args: Any) -> None:
         """
         send message to non-hub component in other process,
         returns a Queue object which eventually contains the response if want_retval
         (not-impl here since nothing uses it yet)
         """
-        if dest == "httpconn":
-            sck, addr = args
-            sck2 = sck
-            if PY2:
-                buf = MemesIO()
-                ForkingPickler(buf).dump(sck)
-                sck2 = buf.getvalue()
+        if dest == "listen":
+            for p in self.procs:
+                p.q_pend.put((0, dest, [args[0], len(self.procs)]))
 
-            proc = sorted(self.procs, key=lambda x: x.workload)[0]
-            proc.q_pend.put([0, dest, [sck2, addr]])
+        elif dest == "set_netdevs":
+            for p in self.procs:
+                p.q_pend.put((0, dest, list(args)))
 
-            with self.mutex:
-                proc.clients[addr] = 50
-                proc.workload += 50
+        elif dest == "cb_httpsrv_up":
+            self.hub.cb_httpsrv_up()
 
         else:
             raise Exception("what is " + str(dest))
-
-    def debug_load_balancer(self):
-        fmt = "\033[1m{}\033[0;36m{:4}\033[0m "
-        if not VT100:
-            fmt = "({}{:4})"
-
-        last = ""
-        while self.procs:
-            msg = ""
-            for proc in self.procs:
-                msg += fmt.format(len(proc.clients), proc.workload)
-
-            if msg != last:
-                last = msg
-                with self.hub.log_mutex:
-                    print(msg)
-
-            time.sleep(0.1)

copyparty/broker_mpw.py

@@ -1,87 +1,104 @@
 # coding: utf-8
 from __future__ import print_function, unicode_literals
 
-import sys
-import time
+import argparse
+import os
 import signal
+import sys
 import threading
 
-from .__init__ import PY2, WINDOWS
-from .broker_util import ExceptionalQueue
+import queue
+
+from .__init__ import ANYWIN
+from .authsrv import AuthSrv
+from .broker_util import BrokerCli, ExceptionalQueue
 from .httpsrv import HttpSrv
-from .util import FAKE_MP
+from .util import FAKE_MP, Daemon, HMaccas
 
-if PY2 and not WINDOWS:
-    import pickle  # nosec
+if True:  # pylint: disable=using-constant-test
+    from types import FrameType
+
+    from typing import Any, Optional, Union
 
 
-class MpWorker(object):
+class MpWorker(BrokerCli):
     """one single mp instance"""
 
-    def __init__(self, q_pend, q_yield, args, n):
+    def __init__(
+        self,
+        q_pend: queue.Queue[tuple[int, str, list[Any]]],
+        q_yield: queue.Queue[tuple[int, str, list[Any]]],
+        args: argparse.Namespace,
+        n: int,
+    ) -> None:
+        super(MpWorker, self).__init__()
+
         self.q_pend = q_pend
         self.q_yield = q_yield
         self.args = args
         self.n = n
 
-        self.retpend = {}
+        self.log = self._log_disabled if args.q and not args.lo else self._log_enabled
+
+        self.retpend: dict[int, Any] = {}
         self.retpend_mutex = threading.Lock()
         self.mutex = threading.Lock()
-        self.workload_thr_active = False
 
         # we inherited signal_handler from parent,
         # replace it with something harmless
         if not FAKE_MP:
-            signal.signal(signal.SIGINT, self.signal_handler)
+            sigs = [signal.SIGINT, signal.SIGTERM]
+            if not ANYWIN:
+                sigs.append(signal.SIGUSR1)
+
+            for sig in sigs:
+                signal.signal(sig, self.signal_handler)
+
+        # starting to look like a good idea
+        self.asrv = AuthSrv(args, None, False)
 
         # instantiate all services here (TODO: inheritance?)
-        self.httpsrv = HttpSrv(self)
-        self.httpsrv.disconnect_func = self.httpdrop
+        self.iphash = HMaccas(os.path.join(self.args.E.cfg, "iphash"), 8)
+        self.httpsrv = HttpSrv(self, n)
 
         # on winxp and some other platforms,
         # use thr.join() to block all signals
-        thr = threading.Thread(target=self.main)
-        thr.daemon = True
-        thr.start()
-        thr.join()
+        Daemon(self.main, "mpw-main").join()
 
-    def signal_handler(self, signal, frame):
+    def signal_handler(self, sig: Optional[int], frame: Optional[FrameType]) -> None:
         # print('k')
         pass
 
-    def log(self, src, msg):
-        self.q_yield.put([0, "log", [src, msg]])
+    def _log_enabled(self, src: str, msg: str, c: Union[int, str] = 0) -> None:
+        self.q_yield.put((0, "log", [src, msg, c]))
 
-    def logw(self, msg):
-        self.log("mp{}".format(self.n), msg)
+    def _log_disabled(self, src: str, msg: str, c: Union[int, str] = 0) -> None:
+        pass
 
-    def httpdrop(self, addr):
-        self.q_yield.put([0, "httpdrop", [addr]])
+    def logw(self, msg: str, c: Union[int, str] = 0) -> None:
+        self.log("mp{}".format(self.n), msg, c)
 
-    def main(self):
+    def main(self) -> None:
         while True:
             retq_id, dest, args = self.q_pend.get()
 
             # self.logw("work: [{}]".format(d[0]))
             if dest == "shutdown":
+                self.httpsrv.shutdown()
                 self.logw("ok bye")
                 sys.exit(0)
                 return
 
-            elif dest == "httpconn":
-                sck, addr = args
-                if PY2:
-                    sck = pickle.loads(sck)  # nosec
+            elif dest == "reload":
+                self.logw("mpw.asrv reloading")
+                self.asrv.reload()
+                self.logw("mpw.asrv reloaded")
 
-                self.log("%s %s" % addr, "-" * 4 + "C-qpop")
-                self.httpsrv.accept(sck, addr)
+            elif dest == "listen":
+                self.httpsrv.listen(args[0], args[1])
 
-                with self.mutex:
-                    if not self.workload_thr_active:
-                        self.workload_thr_alive = True
-                        thr = threading.Thread(target=self.thr_workload)
-                        thr.daemon = True
-                        thr.start()
+            elif dest == "set_netdevs":
+                self.httpsrv.set_netdevs(args[0])
 
             elif dest == "retq":
                 # response from previous ipc call
@@ -93,28 +110,14 @@ class MpWorker(object):
             else:
                 raise Exception("what is " + str(dest))
 
-    def put(self, want_retval, dest, *args):
-        if want_retval:
-            retq = ExceptionalQueue(1)
-            retq_id = id(retq)
-            with self.retpend_mutex:
-                self.retpend[retq_id] = retq
-        else:
-            retq = None
-            retq_id = 0
+    def ask(self, dest: str, *args: Any) -> ExceptionalQueue:
+        retq = ExceptionalQueue(1)
+        retq_id = id(retq)
+        with self.retpend_mutex:
+            self.retpend[retq_id] = retq
 
-        self.q_yield.put([retq_id, dest, args])
+        self.q_yield.put((retq_id, dest, list(args)))
         return retq
 
-    def thr_workload(self):
-        """announce workloads to MpSrv (the mp controller / loadbalancer)"""
-        # avoid locking in extract_filedata by tracking difference here
-        while True:
-            time.sleep(0.2)
-            with self.mutex:
-                if self.httpsrv.num_clients() == 0:
-                    # no clients rn, termiante thread
-                    self.workload_thr_alive = False
-                    return
-
-            self.q_yield.put([0, "workload", [self.httpsrv.workload]])
+    def say(self, dest: str, *args: Any) -> None:
+        self.q_yield.put((0, dest, list(args)))

copyparty/broker_thr.py

@@ -1,51 +1,73 @@
 # coding: utf-8
 from __future__ import print_function, unicode_literals
 
+import os
 import threading
 
+from .__init__ import TYPE_CHECKING
+from .broker_util import BrokerCli, ExceptionalQueue, try_exec
 from .httpsrv import HttpSrv
-from .broker_util import ExceptionalQueue, try_exec
+from .util import HMaccas
+
+if TYPE_CHECKING:
+    from .svchub import SvcHub
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Any
 
 
-class BrokerThr(object):
+class BrokerThr(BrokerCli):
     """external api; behaves like BrokerMP but using plain threads"""
 
-    def __init__(self, hub):
+    def __init__(self, hub: "SvcHub") -> None:
+        super(BrokerThr, self).__init__()
+
         self.hub = hub
         self.log = hub.log
         self.args = hub.args
+        self.asrv = hub.asrv
 
         self.mutex = threading.Lock()
+        self.num_workers = 1
 
         # instantiate all services here (TODO: inheritance?)
-        self.httpsrv = HttpSrv(self)
-        self.httpsrv.disconnect_func = self.httpdrop
+        self.iphash = HMaccas(os.path.join(self.args.E.cfg, "iphash"), 8)
+        self.httpsrv = HttpSrv(self, None)
+        self.reload = self.noop
 
-    def shutdown(self):
+    def shutdown(self) -> None:
         # self.log("broker", "shutting down")
+        self.httpsrv.shutdown()
+
+    def noop(self) -> None:
         pass
 
-    def put(self, want_retval, dest, *args):
-        if dest == "httpconn":
-            sck, addr = args
-            self.log("%s %s" % addr, "-" * 4 + "C-qpop")
-            self.httpsrv.accept(sck, addr)
+    def ask(self, dest: str, *args: Any) -> ExceptionalQueue:
 
-        else:
-            # new ipc invoking managed service in hub
-            obj = self.hub
-            for node in dest.split("."):
-                obj = getattr(obj, node)
+        # new ipc invoking managed service in hub
+        obj = self.hub
+        for node in dest.split("."):
+            obj = getattr(obj, node)
 
-            # TODO will deadlock if dest performs another ipc
-            rv = try_exec(want_retval, obj, *args)
-            if not want_retval:
-                return
+        rv = try_exec(True, obj, *args)
 
-            # pretend we're broker_mp
-            retq = ExceptionalQueue(1)
-            retq.put(rv)
-            return retq
+        # pretend we're broker_mp
+        retq = ExceptionalQueue(1)
+        retq.put(rv)
+        return retq
 
-    def httpdrop(self, addr):
-        self.hub.tcpsrv.num_clients.add(-1)
+    def say(self, dest: str, *args: Any) -> None:
+        if dest == "listen":
+            self.httpsrv.listen(args[0], 1)
+            return
+
+        if dest == "set_netdevs":
+            self.httpsrv.set_netdevs(args[0])
+            return
+
+        # new ipc invoking managed service in hub
+        obj = self.hub
+        for node in dest.split("."):
+            obj = getattr(obj, node)
+
+        try_exec(False, obj, *args)

copyparty/broker_util.py

@@ -1,17 +1,28 @@
 # coding: utf-8
 from __future__ import print_function, unicode_literals
 
-
+import argparse
 import traceback
 
-from .util import Pebkac, Queue
+from queue import Queue
+
+from .__init__ import TYPE_CHECKING
+from .authsrv import AuthSrv
+from .util import HMaccas, Pebkac
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Any, Optional, Union
+
+    from .util import RootLogger
+
+if TYPE_CHECKING:
+    from .httpsrv import HttpSrv
 
 
 class ExceptionalQueue(Queue, object):
-    def get(self, block=True, timeout=None):
+    def get(self, block: bool = True, timeout: Optional[float] = None) -> Any:
         rv = super(ExceptionalQueue, self).get(block, timeout)
 
-        # TODO: how expensive is this?
         if isinstance(rv, list):
             if rv[0] == "exception":
                 if rv[1] == "pebkac":
@@ -22,7 +33,29 @@ class ExceptionalQueue(Queue, object):
         return rv
 
 
-def try_exec(want_retval, func, *args):
+class BrokerCli(object):
+    """
+    helps mypy understand httpsrv.broker but still fails a few levels deeper,
+    for example resolving httpconn.* in httpcli -- see lines tagged #mypy404
+    """
+
+    log: "RootLogger"
+    args: argparse.Namespace
+    asrv: AuthSrv
+    httpsrv: "HttpSrv"
+    iphash: HMaccas
+
+    def __init__(self) -> None:
+        pass
+
+    def ask(self, dest: str, *args: Any) -> ExceptionalQueue:
+        return ExceptionalQueue(1)
+
+    def say(self, dest: str, *args: Any) -> None:
+        pass
+
+
+def try_exec(want_retval: Union[bool, int], func: Any, *args: list[Any]) -> Any:
     try:
         return func(*args)
 

copyparty/dxml.py

@@ -0,0 +1,72 @@
+import importlib
+import sys
+import xml.etree.ElementTree as ET
+
+from .__init__ import PY2
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Any, Optional
+
+
+def get_ET() -> ET.XMLParser:
+    pn = "xml.etree.ElementTree"
+    cn = "_elementtree"
+
+    cmod = sys.modules.pop(cn, None)
+    if not cmod:
+        return ET.XMLParser  # type: ignore
+
+    pmod = sys.modules.pop(pn)
+    sys.modules[cn] = None  # type: ignore
+
+    ret = importlib.import_module(pn)
+    for name, mod in ((pn, pmod), (cn, cmod)):
+        if mod:
+            sys.modules[name] = mod
+        else:
+            sys.modules.pop(name, None)
+
+    sys.modules["xml.etree"].ElementTree = pmod  # type: ignore
+    ret.ParseError = ET.ParseError  # type: ignore
+    return ret.XMLParser  # type: ignore
+
+
+XMLParser: ET.XMLParser = get_ET()
+
+
+class DXMLParser(XMLParser):  # type: ignore
+    def __init__(self) -> None:
+        tb = ET.TreeBuilder()
+        super(DXMLParser, self).__init__(target=tb)
+
+        p = self._parser if PY2 else self.parser
+        p.StartDoctypeDeclHandler = self.nope
+        p.EntityDeclHandler = self.nope
+        p.UnparsedEntityDeclHandler = self.nope
+        p.ExternalEntityRefHandler = self.nope
+
+    def nope(self, *a: Any, **ka: Any) -> None:
+        raise BadXML("{}, {}".format(a, ka))
+
+
+class BadXML(Exception):
+    pass
+
+
+def parse_xml(txt: str) -> ET.Element:
+    parser = DXMLParser()
+    parser.feed(txt)
+    return parser.close()  # type: ignore
+
+
+def mktnod(name: str, text: str) -> ET.Element:
+    el = ET.Element(name)
+    el.text = text
+    return el
+
+
+def mkenod(name: str, sub_el: Optional[ET.Element] = None) -> ET.Element:
+    el = ET.Element(name)
+    if sub_el is not None:
+        el.append(sub_el)
+    return el

copyparty/fsutil.py

@@ -0,0 +1,152 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
+import os
+import re
+import time
+
+from .__init__ import ANYWIN, MACOS
+from .authsrv import AXS, VFS
+from .bos import bos
+from .util import chkcmd, min_ex
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Optional, Union
+
+    from .util import RootLogger
+
+
+class Fstab(object):
+    def __init__(self, log: "RootLogger"):
+        self.log_func = log
+
+        self.trusted = False
+        self.tab: Optional[VFS] = None
+        self.cache: dict[str, str] = {}
+        self.age = 0.0
+
+    def log(self, msg: str, c: Union[int, str] = 0) -> None:
+        self.log_func("fstab", msg, c)
+
+    def get(self, path: str) -> str:
+        if len(self.cache) > 9000:
+            self.age = time.time()
+            self.tab = None
+            self.cache = {}
+
+        fs = "ext4"
+        msg = "failed to determine filesystem at [{}]; assuming {}\n{}"
+
+        if ANYWIN:
+            fs = "vfat"
+            try:
+                path = self._winpath(path)
+            except:
+                self.log(msg.format(path, fs, min_ex()), 3)
+                return fs
+
+        path = path.lstrip("/")
+        try:
+            return self.cache[path]
+        except:
+            pass
+
+        try:
+            fs = self.get_w32(path) if ANYWIN else self.get_unix(path)
+        except:
+            self.log(msg.format(path, fs, min_ex()), 3)
+
+        fs = fs.lower()
+        self.cache[path] = fs
+        self.log("found {} at {}".format(fs, path))
+        return fs
+
+    def _winpath(self, path: str) -> str:
+        # try to combine volume-label + st_dev (vsn)
+        path = path.replace("/", "\\")
+        vid = path.split(":", 1)[0].strip("\\").split("\\", 1)[0]
+        try:
+            return "{}*{}".format(vid, bos.stat(path).st_dev)
+        except:
+            return vid
+
+    def build_fallback(self) -> None:
+        self.tab = VFS(self.log_func, "idk", "/", AXS(), {})
+        self.trusted = False
+
+    def build_tab(self) -> None:
+        self.log("building tab")
+
+        sptn = r"^.*? on (.*) type ([^ ]+) \(.*"
+        if MACOS:
+            sptn = r"^.*? on (.*) \(([^ ]+), .*"
+
+        ptn = re.compile(sptn)
+        so, _ = chkcmd(["mount"])
+        tab1: list[tuple[str, str]] = []
+        for ln in so.split("\n"):
+            m = ptn.match(ln)
+            if not m:
+                continue
+
+            zs1, zs2 = m.groups()
+            tab1.append((str(zs1), str(zs2)))
+
+        tab1.sort(key=lambda x: (len(x[0]), x[0]))
+        path1, fs1 = tab1[0]
+        tab = VFS(self.log_func, fs1, path1, AXS(), {})
+        for path, fs in tab1[1:]:
+            tab.add(fs, path.lstrip("/"))
+
+        self.tab = tab
+
+    def relabel(self, path: str, nval: str) -> None:
+        assert self.tab
+        self.cache = {}
+        if ANYWIN:
+            path = self._winpath(path)
+
+        path = path.lstrip("/")
+        ptn = re.compile(r"^[^\\/]*")
+        vn, rem = self.tab._find(path)
+        if not self.trusted:
+            # no mtab access; have to build as we go
+            if "/" in rem:
+                self.tab.add("idk", os.path.join(vn.vpath, rem.split("/")[0]))
+            if rem:
+                self.tab.add(nval, path)
+            else:
+                vn.realpath = nval
+
+            return
+
+        visit = [vn]
+        while visit:
+            vn = visit.pop()
+            vn.realpath = ptn.sub(nval, vn.realpath)
+            visit.extend(list(vn.nodes.values()))
+
+    def get_unix(self, path: str) -> str:
+        if not self.tab:
+            try:
+                self.build_tab()
+                self.trusted = True
+            except:
+                # prisonparty or other restrictive environment
+                self.log("failed to build tab:\n{}".format(min_ex()), 3)
+                self.build_fallback()
+
+        assert self.tab
+        ret = self.tab._find(path)[0]
+        if self.trusted or path == ret.vpath:
+            return ret.realpath.split("/")[0]
+        else:
+            return "idk"
+
+    def get_w32(self, path: str) -> str:
+        if not self.tab:
+            self.build_fallback()
+
+        assert self.tab
+        ret = self.tab._find(path)[0]
+        return ret.realpath

copyparty/ftpd.py

@@ -0,0 +1,448 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
+import argparse
+import logging
+import os
+import stat
+import sys
+import time
+
+from pyftpdlib.authorizers import AuthenticationFailed, DummyAuthorizer
+from pyftpdlib.filesystems import AbstractedFS, FilesystemError
+from pyftpdlib.handlers import FTPHandler
+from pyftpdlib.servers import FTPServer
+
+from .__init__ import PY2, TYPE_CHECKING, E
+from .bos import bos
+from .util import Daemon, Pebkac, exclude_dotfiles, fsenc, ipnorm
+
+try:
+    from pyftpdlib.ioloop import IOLoop
+except ImportError:
+    p = os.path.join(E.mod, "vend")
+    print("loading asynchat from " + p)
+    sys.path.append(p)
+    from pyftpdlib.ioloop import IOLoop
+
+
+if TYPE_CHECKING:
+    from .svchub import SvcHub
+
+if True:  # pylint: disable=using-constant-test
+    import typing
+    from typing import Any, Optional
+
+
+class FtpAuth(DummyAuthorizer):
+    def __init__(self, hub: "SvcHub") -> None:
+        super(FtpAuth, self).__init__()
+        self.hub = hub
+
+    def validate_authentication(
+        self, username: str, password: str, handler: Any
+    ) -> None:
+        handler.username = "{}:{}".format(username, password)
+
+        ip = handler.addr[0]
+        if ip.startswith("::ffff:"):
+            ip = ip[7:]
+
+        ip = ipnorm(ip)
+        bans = self.hub.bans
+        if ip in bans:
+            rt = bans[ip] - time.time()
+            if rt < 0:
+                logging.info("client unbanned")
+                del bans[ip]
+            else:
+                raise AuthenticationFailed("banned")
+
+        asrv = self.hub.asrv
+        if username == "anonymous":
+            uname = "*"
+        else:
+            uname = asrv.iacct.get(password, "") or asrv.iacct.get(username, "") or "*"
+
+        if not uname or not (asrv.vfs.aread.get(uname) or asrv.vfs.awrite.get(uname)):
+            g = self.hub.gpwd
+            if g.lim:
+                bonk, ip = g.bonk(ip, handler.username)
+                if bonk:
+                    logging.warning("client banned: invalid passwords")
+                    bans[ip] = bonk
+
+            raise AuthenticationFailed("Authentication failed.")
+
+        handler.username = uname
+
+    def get_home_dir(self, username: str) -> str:
+        return "/"
+
+    def has_user(self, username: str) -> bool:
+        asrv = self.hub.asrv
+        return username in asrv.acct
+
+    def has_perm(self, username: str, perm: int, path: Optional[str] = None) -> bool:
+        return True  # handled at filesystem layer
+
+    def get_perms(self, username: str) -> str:
+        return "elradfmwMT"
+
+    def get_msg_login(self, username: str) -> str:
+        return "sup {}".format(username)
+
+    def get_msg_quit(self, username: str) -> str:
+        return "cya"
+
+
+class FtpFs(AbstractedFS):
+    def __init__(
+        self, root: str, cmd_channel: Any
+    ) -> None:  # pylint: disable=super-init-not-called
+        self.h = self.cmd_channel = cmd_channel  # type: FTPHandler
+        self.hub: "SvcHub" = cmd_channel.hub
+        self.args = cmd_channel.args
+
+        self.uname = self.hub.asrv.iacct.get(cmd_channel.password, "*")
+
+        self.cwd = "/"  # pyftpdlib convention of leading slash
+        self.root = "/var/lib/empty"
+
+        self.can_read = self.can_write = self.can_move = False
+        self.can_delete = self.can_get = self.can_upget = False
+
+        self.listdirinfo = self.listdir
+        self.chdir(".")
+
+    def v2a(
+        self,
+        vpath: str,
+        r: bool = False,
+        w: bool = False,
+        m: bool = False,
+        d: bool = False,
+    ) -> str:
+        try:
+            vpath = vpath.replace("\\", "/").lstrip("/")
+            vfs, rem = self.hub.asrv.vfs.get(vpath, self.uname, r, w, m, d)
+            if not vfs.realpath:
+                raise FilesystemError("no filesystem mounted at this path")
+
+            return os.path.join(vfs.realpath, rem)
+        except Pebkac as ex:
+            raise FilesystemError(str(ex))
+
+    def rv2a(
+        self,
+        vpath: str,
+        r: bool = False,
+        w: bool = False,
+        m: bool = False,
+        d: bool = False,
+    ) -> str:
+        return self.v2a(os.path.join(self.cwd, vpath), r, w, m, d)
+
+    def ftp2fs(self, ftppath: str) -> str:
+        # return self.v2a(ftppath)
+        return ftppath  # self.cwd must be vpath
+
+    def fs2ftp(self, fspath: str) -> str:
+        # raise NotImplementedError()
+        return fspath
+
+    def validpath(self, path: str) -> bool:
+        if "/.hist/" in path:
+            if "/up2k." in path or path.endswith("/dir.txt"):
+                raise FilesystemError("access to this file is forbidden")
+
+        return True
+
+    def open(self, filename: str, mode: str) -> typing.IO[Any]:
+        r = "r" in mode
+        w = "w" in mode or "a" in mode or "+" in mode
+
+        ap = self.rv2a(filename, r, w)
+        if w:
+            try:
+                st = bos.stat(ap)
+                td = time.time() - st.st_mtime
+            except:
+                td = 0
+
+            if td < -1 or td > self.args.ftp_wt:
+                raise FilesystemError("cannot open existing file for writing")
+
+        self.validpath(ap)
+        return open(fsenc(ap), mode)
+
+    def chdir(self, path: str) -> None:
+        nwd = join(self.cwd, path)
+        vfs, rem = self.hub.asrv.vfs.get(nwd, self.uname, False, False)
+        ap = vfs.canonical(rem)
+        if not bos.path.isdir(ap):
+            # returning 550 is library-default and suitable
+            raise FilesystemError("Failed to change directory")
+
+        self.cwd = nwd
+        (
+            self.can_read,
+            self.can_write,
+            self.can_move,
+            self.can_delete,
+            self.can_get,
+            self.can_upget,
+        ) = self.hub.asrv.vfs.can_access(self.cwd.lstrip("/"), self.h.username)
+
+    def mkdir(self, path: str) -> None:
+        ap = self.rv2a(path, w=True)
+        bos.mkdir(ap)
+
+    def listdir(self, path: str) -> list[str]:
+        vpath = join(self.cwd, path).lstrip("/")
+        try:
+            vfs, rem = self.hub.asrv.vfs.get(vpath, self.uname, True, False)
+
+            fsroot, vfs_ls1, vfs_virt = vfs.ls(
+                rem,
+                self.uname,
+                not self.args.no_scandir,
+                [[True, False], [False, True]],
+            )
+            vfs_ls = [x[0] for x in vfs_ls1]
+            vfs_ls.extend(vfs_virt.keys())
+
+            if not self.args.ed:
+                vfs_ls = exclude_dotfiles(vfs_ls)
+
+            vfs_ls.sort()
+            return vfs_ls
+        except:
+            if vpath:
+                # display write-only folders as empty
+                return []
+
+            # return list of volumes
+            r = {x.split("/")[0]: 1 for x in self.hub.asrv.vfs.all_vols.keys()}
+            return list(sorted(list(r.keys())))
+
+    def rmdir(self, path: str) -> None:
+        ap = self.rv2a(path, d=True)
+        bos.rmdir(ap)
+
+    def remove(self, path: str) -> None:
+        if self.args.no_del:
+            raise FilesystemError("the delete feature is disabled in server config")
+
+        vp = join(self.cwd, path).lstrip("/")
+        try:
+            self.hub.up2k.handle_rm(self.uname, self.h.remote_ip, [vp], [])
+        except Exception as ex:
+            raise FilesystemError(str(ex))
+
+    def rename(self, src: str, dst: str) -> None:
+        if not self.can_move:
+            raise FilesystemError("not allowed for user " + self.h.username)
+
+        if self.args.no_mv:
+            t = "the rename/move feature is disabled in server config"
+            raise FilesystemError(t)
+
+        svp = join(self.cwd, src).lstrip("/")
+        dvp = join(self.cwd, dst).lstrip("/")
+        try:
+            self.hub.up2k.handle_mv(self.uname, svp, dvp)
+        except Exception as ex:
+            raise FilesystemError(str(ex))
+
+    def chmod(self, path: str, mode: str) -> None:
+        pass
+
+    def stat(self, path: str) -> os.stat_result:
+        try:
+            ap = self.rv2a(path, r=True)
+            return bos.stat(ap)
+        except:
+            ap = self.rv2a(path)
+            st = bos.stat(ap)
+            if not stat.S_ISDIR(st.st_mode):
+                raise
+
+            return st
+
+    def utime(self, path: str, timeval: float) -> None:
+        ap = self.rv2a(path, w=True)
+        return bos.utime(ap, (timeval, timeval))
+
+    def lstat(self, path: str) -> os.stat_result:
+        ap = self.rv2a(path)
+        return bos.stat(ap)
+
+    def isfile(self, path: str) -> bool:
+        try:
+            st = self.stat(path)
+            return stat.S_ISREG(st.st_mode)
+        except:
+            return False  # expected for mojibake in ftp_SIZE()
+
+    def islink(self, path: str) -> bool:
+        ap = self.rv2a(path)
+        return bos.path.islink(ap)
+
+    def isdir(self, path: str) -> bool:
+        try:
+            st = self.stat(path)
+            return stat.S_ISDIR(st.st_mode)
+        except:
+            return True
+
+    def getsize(self, path: str) -> int:
+        ap = self.rv2a(path)
+        return bos.path.getsize(ap)
+
+    def getmtime(self, path: str) -> float:
+        ap = self.rv2a(path)
+        return bos.path.getmtime(ap)
+
+    def realpath(self, path: str) -> str:
+        return path
+
+    def lexists(self, path: str) -> bool:
+        ap = self.rv2a(path)
+        return bos.path.lexists(ap)
+
+    def get_user_by_uid(self, uid: int) -> str:
+        return "root"
+
+    def get_group_by_uid(self, gid: int) -> str:
+        return "root"
+
+
+class FtpHandler(FTPHandler):
+    abstracted_fs = FtpFs
+    hub: "SvcHub"
+    args: argparse.Namespace
+
+    def __init__(self, conn: Any, server: Any, ioloop: Any = None) -> None:
+        self.hub: "SvcHub" = FtpHandler.hub
+        self.args: argparse.Namespace = FtpHandler.args
+
+        if PY2:
+            FTPHandler.__init__(self, conn, server, ioloop)
+        else:
+            super(FtpHandler, self).__init__(conn, server, ioloop)
+
+        # abspath->vpath mapping to resolve log_transfer paths
+        self.vfs_map: dict[str, str] = {}
+
+        # reduce non-debug logging
+        self.log_cmds_list = [x for x in self.log_cmds_list if x not in ("CWD", "XCWD")]
+
+    def ftp_STOR(self, file: str, mode: str = "w") -> Any:
+        # Optional[str]
+        vp = join(self.fs.cwd, file).lstrip("/")
+        ap = self.fs.v2a(vp)
+        self.vfs_map[ap] = vp
+        # print("ftp_STOR: {} {} => {}".format(vp, mode, ap))
+        ret = FTPHandler.ftp_STOR(self, file, mode)
+        # print("ftp_STOR: {} {} OK".format(vp, mode))
+        return ret
+
+    def log_transfer(
+        self,
+        cmd: str,
+        filename: bytes,
+        receive: bool,
+        completed: bool,
+        elapsed: float,
+        bytes: int,
+    ) -> Any:
+        # None
+        ap = filename.decode("utf-8", "replace")
+        vp = self.vfs_map.pop(ap, None)
+        # print("xfer_end: {} => {}".format(ap, vp))
+        if vp:
+            vp, fn = os.path.split(vp)
+            vfs, rem = self.hub.asrv.vfs.get(vp, self.username, False, True)
+            vfs, rem = vfs.get_dbv(rem)
+            self.hub.up2k.hash_file(
+                vfs.realpath,
+                vfs.flags,
+                rem,
+                fn,
+                self.remote_ip,
+                time.time(),
+            )
+
+        return FTPHandler.log_transfer(
+            self, cmd, filename, receive, completed, elapsed, bytes
+        )
+
+
+try:
+    from pyftpdlib.handlers import TLS_FTPHandler
+
+    class SftpHandler(FtpHandler, TLS_FTPHandler):
+        pass
+
+except:
+    pass
+
+
+class Ftpd(object):
+    def __init__(self, hub: "SvcHub") -> None:
+        self.hub = hub
+        self.args = hub.args
+
+        hs = []
+        if self.args.ftp:
+            hs.append([FtpHandler, self.args.ftp])
+        if self.args.ftps:
+            try:
+                h1 = SftpHandler
+            except:
+                t = "\nftps requires pyopenssl;\nplease run the following:\n\n  {} -m pip install --user pyopenssl\n"
+                print(t.format(sys.executable))
+                sys.exit(1)
+
+            h1.certfile = os.path.join(self.args.E.cfg, "cert.pem")
+            h1.tls_control_required = True
+            h1.tls_data_required = True
+
+            hs.append([h1, self.args.ftps])
+
+        for h_lp in hs:
+            h2, lp = h_lp
+            h2.hub = hub
+            h2.args = hub.args
+            h2.authorizer = FtpAuth(hub)
+
+            if self.args.ftp_pr:
+                p1, p2 = [int(x) for x in self.args.ftp_pr.split("-")]
+                if self.args.ftp and self.args.ftps:
+                    # divide port range in half
+                    d = int((p2 - p1) / 2)
+                    if lp == self.args.ftp:
+                        p2 = p1 + d
+                    else:
+                        p1 += d + 1
+
+                h2.passive_ports = list(range(p1, p2 + 1))
+
+            if self.args.ftp_nat:
+                h2.masquerade_address = self.args.ftp_nat
+
+        lgr = logging.getLogger("pyftpdlib")
+        lgr.setLevel(logging.DEBUG if self.args.ftpv else logging.INFO)
+
+        ioloop = IOLoop()
+        for ip in self.args.i:
+            for h, lp in hs:
+                FTPServer((ip, int(lp)), h, ioloop)
+
+        Daemon(ioloop.loop, "ftp")
+
+
+def join(p1: str, p2: str) -> str:
+    w = os.path.join(p1, p2.replace("\\", "/"))
+    return os.path.normpath(w).replace("\\", "/")

copyparty/httpconn.py

@@ -1,29 +1,38 @@
 # coding: utf-8
 from __future__ import print_function, unicode_literals
 
+import argparse  # typechk
 import os
-import sys
-import ssl
+import re
 import socket
+import threading  # typechk
+import time
 
 try:
-    import jinja2
-except ImportError:
-    print(
-        """\033[1;31m
-  you do not have jinja2 installed,\033[33m
-  choose one of these:\033[0m
-   * apt install python-jinja2
-   * python3 -m pip install --user jinja2
-   * (try another python version, if you have one)
-   * (try copyparty.sfx instead)
-"""
-    )
-    sys.exit(1)
+    HAVE_SSL = True
+    import ssl
+except:
+    HAVE_SSL = False
 
-from .__init__ import E
-from .util import Unrecv
+from . import util as Util
+from .__init__ import TYPE_CHECKING, EnvParams
+from .authsrv import AuthSrv  # typechk
 from .httpcli import HttpCli
+from .ico import Ico
+from .mtag import HAVE_FFMPEG
+from .th_cli import ThumbCli
+from .th_srv import HAVE_PIL, HAVE_VIPS
+from .u2idx import U2idx
+from .util import HMaccas, shut_socket
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Optional, Pattern, Union
+
+if TYPE_CHECKING:
+    from .httpsrv import HttpSrv
+
+
+PTN_HTTP = re.compile(br"[A-Z]{3}[A-Z ]")
 
 
 class HttpConn(object):
@@ -32,90 +41,176 @@ class HttpConn(object):
     creates an HttpCli for each request (Connection: Keep-Alive)
     """
 
-    def __init__(self, sck, addr, hsrv):
+    def __init__(
+        self, sck: socket.socket, addr: tuple[str, int], hsrv: "HttpSrv"
+    ) -> None:
         self.s = sck
+        self.sr: Optional[Util._Unrecv] = None
+        self.cli: Optional[HttpCli] = None
         self.addr = addr
         self.hsrv = hsrv
 
-        self.args = hsrv.args
-        self.auth = hsrv.auth
+        self.mutex: threading.Lock = hsrv.mutex  # mypy404
+        self.args: argparse.Namespace = hsrv.args  # mypy404
+        self.E: EnvParams = self.args.E
+        self.asrv: AuthSrv = hsrv.asrv  # mypy404
         self.cert_path = hsrv.cert_path
+        self.u2fh: Util.FHC = hsrv.u2fh  # mypy404
+        self.iphash: HMaccas = hsrv.broker.iphash
+        self.bans: dict[str, int] = hsrv.bans
+        self.aclose: dict[str, int] = hsrv.aclose
 
-        self.workload = 0
-        self.log_func = hsrv.log
-        self.log_src = "{} \033[36m{}".format(addr[0], addr[1]).ljust(26)
+        enth = (HAVE_PIL or HAVE_VIPS or HAVE_FFMPEG) and not self.args.no_thumb
+        self.thumbcli: Optional[ThumbCli] = ThumbCli(hsrv) if enth else None  # mypy404
+        self.ico: Ico = Ico(self.args)  # mypy404
 
-        env = jinja2.Environment()
-        env.loader = jinja2.FileSystemLoader(os.path.join(E.mod, "web"))
-        self.tpl_mounts = env.get_template("splash.html")
-        self.tpl_browser = env.get_template("browser.html")
-        self.tpl_msg = env.get_template("msg.html")
-        self.tpl_md = env.get_template("md.html")
-        self.tpl_mde = env.get_template("mde.html")
+        self.t0: float = time.time()  # mypy404
+        self.stopping = False
+        self.nreq: int = -1  # mypy404
+        self.nbyte: int = 0  # mypy404
+        self.u2idx: Optional[U2idx] = None
+        self.log_func: "Util.RootLogger" = hsrv.log  # mypy404
+        self.log_src: str = "httpconn"  # mypy404
+        self.lf_url: Optional[Pattern[str]] = (
+            re.compile(self.args.lf_url) if self.args.lf_url else None
+        )  # mypy404
+        self.set_rproxy()
 
-    def respath(self, res_name):
-        return os.path.join(E.mod, "web", res_name)
+    def shutdown(self) -> None:
+        self.stopping = True
+        try:
+            shut_socket(self.log, self.s, 1)
+        except:
+            pass
 
-    def log(self, msg):
-        self.log_func(self.log_src, msg)
+    def set_rproxy(self, ip: Optional[str] = None) -> str:
+        if ip is None:
+            color = 36
+            ip = self.addr[0]
+            self.rproxy = None
+        else:
+            color = 34
+            self.rproxy = ip
 
-    def run(self):
+        self.ip = ip
+        self.log_src = "{} \033[{}m{}".format(ip, color, self.addr[1]).ljust(26)
+        return self.log_src
+
+    def respath(self, res_name: str) -> str:
+        return os.path.join(self.E.mod, "web", res_name)
+
+    def log(self, msg: str, c: Union[int, str] = 0) -> None:
+        self.log_func(self.log_src, msg, c)
+
+    def get_u2idx(self) -> U2idx:
+        # one u2idx per tcp connection;
+        # sqlite3 fully parallelizes under python threads
+        if not self.u2idx:
+            self.u2idx = U2idx(self)
+
+        return self.u2idx
+
+    def _detect_https(self) -> bool:
         method = None
-        self.sr = None
         if self.cert_path:
             try:
                 method = self.s.recv(4, socket.MSG_PEEK)
             except socket.timeout:
-                return
+                return False
             except AttributeError:
                 # jython does not support msg_peek; forget about https
                 method = self.s.recv(4)
-                self.sr = Unrecv(self.s)
+                self.sr = Util.Unrecv(self.s, self.log)
                 self.sr.buf = method
 
                 # jython used to do this, they stopped since it's broken
                 # but reimplementing sendall is out of scope for now
                 if not getattr(self.s, "sendall", None):
-                    self.s.sendall = self.s.send
+                    self.s.sendall = self.s.send  # type: ignore
 
             if len(method) != 4:
                 err = "need at least 4 bytes in the first packet; got {}".format(
                     len(method)
                 )
-                self.log(err)
-                self.s.send(b"HTTP/1.1 400 Bad Request\r\n\r\n" + err.encode("utf-8"))
-                return
+                if method:
+                    self.log(err)
 
-        if method not in [None, b"GET ", b"HEAD", b"POST", b"PUT ", b"OPTI"]:
+                self.s.send(b"HTTP/1.1 400 Bad Request\r\n\r\n" + err.encode("utf-8"))
+                return False
+
+        return not method or not bool(PTN_HTTP.match(method))
+
+    def run(self) -> None:
+        self.s.settimeout(10)
+
+        self.sr = None
+        if self.args.https_only:
+            is_https = True
+        elif self.args.http_only or not HAVE_SSL:
+            is_https = False
+        else:
+            # raise Exception("asdf")
+            is_https = self._detect_https()
+
+        if is_https:
             if self.sr:
-                self.log("\033[1;31mTODO: cannot do https in jython\033[0m")
+                self.log("TODO: cannot do https in jython", c="1;31")
                 return
 
             self.log_src = self.log_src.replace("[36m", "[35m")
             try:
-                self.s = ssl.wrap_socket(
-                    self.s, server_side=True, certfile=self.cert_path
-                )
+                ctx = ssl.create_default_context(ssl.Purpose.CLIENT_AUTH)
+                ctx.load_cert_chain(self.cert_path)
+                if self.args.ssl_ver:
+                    ctx.options &= ~self.args.ssl_flags_en
+                    ctx.options |= self.args.ssl_flags_de
+                    # print(repr(ctx.options))
+
+                if self.args.ssl_log:
+                    try:
+                        ctx.keylog_filename = self.args.ssl_log
+                    except:
+                        self.log("keylog failed; openssl or python too old")
+
+                if self.args.ciphers:
+                    ctx.set_ciphers(self.args.ciphers)
+
+                self.s = ctx.wrap_socket(self.s, server_side=True)
+                msg = [
+                    "\033[1;3{:d}m{}".format(c, s)
+                    for c, s in zip([0, 5, 0], self.s.cipher())  # type: ignore
+                ]
+                self.log(" ".join(msg) + "\033[0m")
+
+                if self.args.ssl_dbg and hasattr(self.s, "shared_ciphers"):
+                    ciphers = self.s.shared_ciphers()
+                    assert ciphers
+                    overlap = [str(y[::-1]) for y in ciphers]
+                    self.log("TLS cipher overlap:" + "\n".join(overlap))
+                    for k, v in [
+                        ["compression", self.s.compression()],
+                        ["ALPN proto", self.s.selected_alpn_protocol()],
+                        ["NPN proto", self.s.selected_npn_protocol()],
+                    ]:
+                        self.log("TLS {}: {}".format(k, v or "nah"))
+
             except Exception as ex:
                 em = str(ex)
 
-                if "ALERT_BAD_CERTIFICATE" in em:
-                    # firefox-linux if there is no exception yet
-                    self.log("client rejected our certificate (nice)")
-
-                elif "ALERT_CERTIFICATE_UNKNOWN" in em:
-                    # chrome-android keeps doing this
+                if "ALERT_CERTIFICATE_UNKNOWN" in em:
+                    # android-chrome keeps doing this
                     pass
 
                 else:
-                    self.log("\033[35mhandshake\033[0m " + em)
+                    self.log("handshake\033[0m " + em, c=5)
 
                 return
 
         if not self.sr:
-            self.sr = Unrecv(self.s)
+            self.sr = Util.Unrecv(self.s, self.log)
 
-        while True:
-            cli = HttpCli(self)
-            if not cli.run():
+        while not self.stopping:
+            self.nreq += 1
+            self.cli = HttpCli(self)
+            if not self.cli.run():
                 return

copyparty/httpsrv.py

@@ -1,14 +1,58 @@
 # coding: utf-8
 from __future__ import print_function, unicode_literals
 
+import base64
+import math
 import os
-import time
 import socket
+import sys
 import threading
+import time
 
-from .__init__ import E, MACOS
+import queue
+
+try:
+    import jinja2
+except ImportError:
+    print(
+        """\033[1;31m
+  you do not have jinja2 installed,\033[33m
+  choose one of these:\033[0m
+   * apt install python-jinja2
+   * {} -m pip install --user jinja2
+   * (try another python version, if you have one)
+   * (try copyparty.sfx instead)
+""".format(
+            os.path.basename(sys.executable)
+        )
+    )
+    sys.exit(1)
+
+from .__init__ import MACOS, TYPE_CHECKING, EnvParams
+from .bos import bos
 from .httpconn import HttpConn
-from .authsrv import AuthSrv
+from .util import (
+    E_SCK,
+    FHC,
+    Daemon,
+    Garda,
+    Magician,
+    Netdev,
+    NetMap,
+    ipnorm,
+    min_ex,
+    shut_socket,
+    spack,
+    start_log_thrs,
+    start_stackmon,
+)
+
+if TYPE_CHECKING:
+    from .broker_util import BrokerCli
+    from .ssdp import SSDPr
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Any, Optional
 
 
 class HttpSrv(object):
@@ -17,105 +61,392 @@ class HttpSrv(object):
     relying on MpSrv for performance (HttpSrv is just plain threads)
     """
 
-    def __init__(self, broker):
+    def __init__(self, broker: "BrokerCli", nid: Optional[int]) -> None:
         self.broker = broker
+        self.nid = nid
         self.args = broker.args
+        self.E: EnvParams = self.args.E
         self.log = broker.log
+        self.asrv = broker.asrv
 
-        self.disconnect_func = None
+        # redefine in case of multiprocessing
+        socket.setdefaulttimeout(120)
+
+        nsuf = "-n{}-i{:x}".format(nid, os.getpid()) if nid else ""
+        self.magician = Magician()
+        self.nm = NetMap([], {})
+        self.ssdp: Optional["SSDPr"] = None
+        self.gpwd = Garda(self.args.ban_pw)
+        self.g404 = Garda(self.args.ban_404)
+        self.bans: dict[str, int] = {}
+        self.aclose: dict[str, int] = {}
+
+        self.ip = ""
+        self.port = 0
+        self.name = "hsrv" + nsuf
         self.mutex = threading.Lock()
+        self.stopping = False
 
-        self.clients = {}
-        self.workload = 0
-        self.workload_thr_alive = False
-        self.auth = AuthSrv(self.args, self.log)
+        self.tp_nthr = 0  # actual
+        self.tp_ncli = 0  # fading
+        self.tp_time = 0.0  # latest worker collect
+        self.tp_q: Optional[queue.LifoQueue[Any]] = (
+            None if self.args.no_htp else queue.LifoQueue()
+        )
+        self.t_periodic: Optional[threading.Thread] = None
 
-        cert_path = os.path.join(E.cfg, "cert.pem")
-        if os.path.exists(cert_path):
+        self.u2fh = FHC()
+        self.srvs: list[socket.socket] = []
+        self.ncli = 0  # exact
+        self.clients: set[HttpConn] = set()  # laggy
+        self.nclimax = 0
+        self.cb_ts = 0.0
+        self.cb_v = ""
+
+        env = jinja2.Environment()
+        env.loader = jinja2.FileSystemLoader(os.path.join(self.E.mod, "web"))
+        jn = ["splash", "svcs", "browser", "browser2", "msg", "md", "mde", "cf"]
+        self.j2 = {x: env.get_template(x + ".html") for x in jn}
+        zs = os.path.join(self.E.mod, "web", "deps", "prism.js.gz")
+        self.prism = os.path.exists(zs)
+
+        if self.args.zs:
+            from .ssdp import SSDPr
+
+            self.ssdp = SSDPr(broker)
+
+        cert_path = os.path.join(self.E.cfg, "cert.pem")
+        if bos.path.exists(cert_path):
             self.cert_path = cert_path
         else:
-            self.cert_path = None
+            self.cert_path = ""
 
-    def accept(self, sck, addr):
+        if self.tp_q:
+            self.start_threads(4)
+
+        if nid:
+            if self.args.stackmon:
+                start_stackmon(self.args.stackmon, nid)
+
+            if self.args.log_thrs:
+                start_log_thrs(self.log, self.args.log_thrs, nid)
+
+        self.th_cfg: dict[str, Any] = {}
+        Daemon(self.post_init, "hsrv-init2")
+
+    def post_init(self) -> None:
+        try:
+            x = self.broker.ask("thumbsrv.getcfg")
+            self.th_cfg = x.get()
+        except:
+            pass
+
+    def set_netdevs(self, netdevs: dict[str, Netdev]) -> None:
+        self.nm = NetMap([self.ip], netdevs)
+
+    def start_threads(self, n: int) -> None:
+        self.tp_nthr += n
+        if self.args.log_htp:
+            self.log(self.name, "workers += {} = {}".format(n, self.tp_nthr), 6)
+
+        for _ in range(n):
+            Daemon(self.thr_poolw, self.name + "-poolw")
+
+    def stop_threads(self, n: int) -> None:
+        self.tp_nthr -= n
+        if self.args.log_htp:
+            self.log(self.name, "workers -= {} = {}".format(n, self.tp_nthr), 6)
+
+        assert self.tp_q
+        for _ in range(n):
+            self.tp_q.put(None)
+
+    def periodic(self) -> None:
+        while True:
+            time.sleep(2 if self.tp_ncli or self.ncli else 10)
+            with self.mutex:
+                self.u2fh.clean()
+                if self.tp_q:
+                    self.tp_ncli = max(self.ncli, self.tp_ncli - 2)
+                    if self.tp_nthr > self.tp_ncli + 8:
+                        self.stop_threads(4)
+
+                if not self.ncli and not self.u2fh.cache and self.tp_nthr <= 8:
+                    self.t_periodic = None
+                    return
+
+    def listen(self, sck: socket.socket, nlisteners: int) -> None:
+        if self.args.j != 1:
+            # lost in the pickle; redefine
+            try:
+                sck.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEPORT, 1)
+            except:
+                pass
+            sck.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+            sck.setsockopt(socket.IPPROTO_TCP, socket.TCP_NODELAY, 1)
+            sck.settimeout(None)  # < does not inherit, ^ does
+
+        self.ip, self.port = sck.getsockname()[:2]
+        self.srvs.append(sck)
+        self.nclimax = math.ceil(self.args.nc * 1.0 / nlisteners)
+        Daemon(
+            self.thr_listen,
+            "httpsrv-n{}-listen-{}-{}".format(self.nid or "0", self.ip, self.port),
+            (sck,),
+        )
+
+    def thr_listen(self, srv_sck: socket.socket) -> None:
+        """listens on a shared tcp server"""
+        ip, port = srv_sck.getsockname()[:2]
+        fno = srv_sck.fileno()
+        hip = "[{}]".format(ip) if ":" in ip else ip
+        msg = "subscribed @ {}:{}  f{} p{}".format(hip, port, fno, os.getpid())
+        self.log(self.name, msg)
+
+        def fun() -> None:
+            self.broker.say("cb_httpsrv_up")
+
+        threading.Thread(target=fun, name="sig-hsrv-up1").start()
+
+        while not self.stopping:
+            if self.args.log_conn:
+                self.log(self.name, "|%sC-ncli" % ("-" * 1,), c="90")
+
+            spins = 0
+            while self.ncli >= self.nclimax:
+                if not spins:
+                    self.log(self.name, "at connection limit; waiting", 3)
+
+                spins += 1
+                time.sleep(0.1)
+                if spins != 50 or not self.args.aclose:
+                    continue
+
+                ipfreq: dict[str, int] = {}
+                with self.mutex:
+                    for c in self.clients:
+                        ip = ipnorm(c.ip)
+                        try:
+                            ipfreq[ip] += 1
+                        except:
+                            ipfreq[ip] = 1
+
+                ip, n = sorted(ipfreq.items(), key=lambda x: x[1], reverse=True)[0]
+                if n < self.nclimax / 2:
+                    continue
+
+                self.aclose[ip] = int(time.time() + self.args.aclose * 60)
+                nclose = 0
+                nloris = 0
+                nconn = 0
+                with self.mutex:
+                    for c in self.clients:
+                        cip = ipnorm(c.ip)
+                        if ip != cip:
+                            continue
+
+                        nconn += 1
+                        try:
+                            if (
+                                c.nreq >= 1
+                                or not c.cli
+                                or c.cli.in_hdr_recv
+                                or c.cli.keepalive
+                            ):
+                                Daemon(c.shutdown)
+                                nclose += 1
+                                if c.nreq <= 0 and (not c.cli or c.cli.in_hdr_recv):
+                                    nloris += 1
+                        except:
+                            pass
+
+                t = "{} downgraded to connection:close for {} min; dropped {}/{} connections"
+                self.log(self.name, t.format(ip, self.args.aclose, nclose, nconn), 1)
+
+                if nloris < nconn / 2:
+                    continue
+
+                t = "slowloris (idle-conn): {} banned for {} min"
+                self.log(self.name, t.format(ip, self.args.loris, nclose), 1)
+                self.bans[ip] = int(time.time() + self.args.loris * 60)
+
+            if self.args.log_conn:
+                self.log(self.name, "|%sC-acc1" % ("-" * 2,), c="90")
+
+            try:
+                sck, saddr = srv_sck.accept()
+                cip, cport = saddr[:2]
+                if cip.startswith("::ffff:"):
+                    cip = cip[7:]
+
+                addr = (cip, cport)
+            except (OSError, socket.error) as ex:
+                if self.stopping:
+                    break
+
+                self.log(self.name, "accept({}): {}".format(fno, ex), c=6)
+                time.sleep(0.02)
+                continue
+
+            if self.args.log_conn:
+                t = "|{}C-acc2 \033[0;36m{} \033[3{}m{}".format(
+                    "-" * 3, ip, port % 8, port
+                )
+                self.log("%s %s" % addr, t, c="90")
+
+            self.accept(sck, addr)
+
+    def accept(self, sck: socket.socket, addr: tuple[str, int]) -> None:
         """takes an incoming tcp connection and creates a thread to handle it"""
-        self.log("%s %s" % addr, "-" * 5 + "C-cthr")
-        thr = threading.Thread(target=self.thr_client, args=(sck, addr))
-        thr.daemon = True
-        thr.start()
+        now = time.time()
+
+        if now - (self.tp_time or now) > 300:
+            t = "httpserver threadpool died: tpt {:.2f}, now {:.2f}, nthr {}, ncli {}"
+            self.log(self.name, t.format(self.tp_time, now, self.tp_nthr, self.ncli), 1)
+            self.tp_time = 0
+            self.tp_q = None
 
-    def num_clients(self):
         with self.mutex:
-            return len(self.clients)
+            self.ncli += 1
+            if not self.t_periodic:
+                name = "hsrv-pt"
+                if self.nid:
+                    name += "-{}".format(self.nid)
 
-    def shutdown(self):
-        self.log("ok bye")
+                self.t_periodic = Daemon(self.periodic, name)
 
-    def thr_client(self, sck, addr):
+            if self.tp_q:
+                self.tp_time = self.tp_time or now
+                self.tp_ncli = max(self.tp_ncli, self.ncli)
+                if self.tp_nthr < self.ncli + 4:
+                    self.start_threads(8)
+
+                self.tp_q.put((sck, addr))
+                return
+
+        if not self.args.no_htp:
+            t = "looks like the httpserver threadpool died; please make an issue on github and tell me the story of how you pulled that off, thanks and dog bless\n"
+            self.log(self.name, t, 1)
+
+        Daemon(
+            self.thr_client,
+            "httpconn-{}-{}".format(addr[0].split(".", 2)[-1][-6:], addr[1]),
+            (sck, addr),
+        )
+
+    def thr_poolw(self) -> None:
+        assert self.tp_q
+        while True:
+            task = self.tp_q.get()
+            if not task:
+                break
+
+            with self.mutex:
+                self.tp_time = 0
+
+            try:
+                sck, addr = task
+                me = threading.current_thread()
+                me.name = "httpconn-{}-{}".format(
+                    addr[0].split(".", 2)[-1][-6:], addr[1]
+                )
+                self.thr_client(sck, addr)
+                me.name = self.name + "-poolw"
+            except Exception as ex:
+                if str(ex).startswith("client d/c "):
+                    self.log(self.name, "thr_client: " + str(ex), 6)
+                else:
+                    self.log(self.name, "thr_client: " + min_ex(), 3)
+
+    def shutdown(self) -> None:
+        self.stopping = True
+        for srv in self.srvs:
+            try:
+                srv.close()
+            except:
+                pass
+
+        thrs = []
+        clients = list(self.clients)
+        for cli in clients:
+            t = threading.Thread(target=cli.shutdown)
+            thrs.append(t)
+            t.start()
+
+        if self.tp_q:
+            self.stop_threads(self.tp_nthr)
+            for _ in range(10):
+                time.sleep(0.05)
+                if self.tp_q.empty():
+                    break
+
+        for t in thrs:
+            t.join()
+
+        self.log(self.name, "ok bye")
+
+    def thr_client(self, sck: socket.socket, addr: tuple[str, int]) -> None:
         """thread managing one tcp client"""
-        sck.settimeout(120)
-
         cli = HttpConn(sck, addr, self)
         with self.mutex:
-            self.clients[cli] = 0
-            self.workload += 50
-
-            if not self.workload_thr_alive:
-                self.workload_thr_alive = True
-                thr = threading.Thread(target=self.thr_workload)
-                thr.daemon = True
-                thr.start()
+            self.clients.add(cli)
 
+        # print("{}\n".format(len(self.clients)), end="")
+        fno = sck.fileno()
         try:
-            self.log("%s %s" % addr, "-" * 6 + "C-crun")
+            if self.args.log_conn:
+                self.log("%s %s" % addr, "|%sC-crun" % ("-" * 4,), c="90")
+
             cli.run()
 
+        except (OSError, socket.error) as ex:
+            if ex.errno not in E_SCK:
+                self.log(
+                    "%s %s" % addr,
+                    "run({}): {}".format(fno, ex),
+                    c=6,
+                )
+
         finally:
-            self.log("%s %s" % addr, "-" * 7 + "C-done")
+            sck = cli.s
+            if self.args.log_conn:
+                self.log("%s %s" % addr, "|%sC-cdone" % ("-" * 5,), c="90")
+
             try:
-                sck.shutdown(socket.SHUT_RDWR)
-                sck.close()
+                fno = sck.fileno()
+                shut_socket(cli.log, sck)
             except (OSError, socket.error) as ex:
                 if not MACOS:
                     self.log(
                         "%s %s" % addr,
-                        "shut_rdwr err:\n  {}\n  {}".format(repr(sck), ex),
+                        "shut({}): {}".format(fno, ex),
+                        c="90",
                     )
-                if ex.errno not in [10038, 10054, 107, 57, 9]:
-                    # 10038 No longer considered a socket
-                    # 10054 Foribly closed by remote
-                    #   107 Transport endpoint not connected
-                    #    57 Socket is not connected
-                    #     9 Bad file descriptor
+                if ex.errno not in E_SCK:
                     raise
             finally:
                 with self.mutex:
-                    del self.clients[cli]
+                    self.clients.remove(cli)
+                    self.ncli -= 1
 
-                if self.disconnect_func:
-                    self.disconnect_func(addr)  # pylint: disable=not-callable
+    def cachebuster(self) -> str:
+        if time.time() - self.cb_ts < 1:
+            return self.cb_v
 
-    def thr_workload(self):
-        """indicates the python interpreter workload caused by this HttpSrv"""
-        # avoid locking in extract_filedata by tracking difference here
-        while True:
-            time.sleep(0.2)
-            with self.mutex:
-                if not self.clients:
-                    # no clients rn, termiante thread
-                    self.workload_thr_alive = False
-                    self.workload = 0
-                    return
+        with self.mutex:
+            if time.time() - self.cb_ts < 1:
+                return self.cb_v
 
-            total = 0
-            with self.mutex:
-                for cli in self.clients.keys():
-                    now = cli.workload
-                    delta = now - self.clients[cli]
-                    if delta < 0:
-                        # was reset in HttpCli to prevent overflow
-                        delta = now
+            v = self.E.t0
+            try:
+                with os.scandir(os.path.join(self.E.mod, "web")) as dh:
+                    for fh in dh:
+                        inf = fh.stat()
+                        v = max(v, inf.st_mtime)
+            except:
+                pass
 
-                    total += delta
-                    self.clients[cli] = now
-
-            self.workload = total
+            v = base64.urlsafe_b64encode(spack(b">xxL", int(v)))
+            self.cb_v = v.decode("ascii")[-4:]
+            self.cb_ts = time.time()
+            return self.cb_v

copyparty/ico.py

@@ -0,0 +1,78 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
+import argparse  # typechk
+import colorsys
+import hashlib
+
+from .__init__ import PY2
+from .th_srv import HAVE_PIL
+from .util import BytesIO
+
+
+class Ico(object):
+    def __init__(self, args: argparse.Namespace) -> None:
+        self.args = args
+
+    def get(self, ext: str, as_thumb: bool, chrome: bool) -> tuple[str, bytes]:
+        """placeholder to make thumbnails not break"""
+
+        zb = hashlib.sha1(ext.encode("utf-8")).digest()[2:4]
+        if PY2:
+            zb = [ord(x) for x in zb]
+
+        c1 = colorsys.hsv_to_rgb(zb[0] / 256.0, 1, 0.3)
+        c2 = colorsys.hsv_to_rgb(zb[0] / 256.0, 1, 1)
+        ci = [int(x * 255) for x in list(c1) + list(c2)]
+        c = "".join(["{:02x}".format(x) for x in ci])
+
+        w = 100
+        h = 30
+        if not self.args.th_no_crop and as_thumb:
+            sw, sh = self.args.th_size.split("x")
+            h = int(100 / (float(sw) / float(sh)))
+            w = 100
+
+        if chrome and as_thumb:
+            # cannot handle more than ~2000 unique SVGs
+            if HAVE_PIL:
+                # svg: 3s, cache: 6s, this: 8s
+                from PIL import Image, ImageDraw
+
+                h = int(64 * h / w)
+                w = 64
+                img = Image.new("RGB", (w, h), "#" + c[:6])
+                pb = ImageDraw.Draw(img)
+                tw, th = pb.textsize(ext)
+                pb.text(((w - tw) // 2, (h - th) // 2), ext, fill="#" + c[6:])
+                img = img.resize((w * 3, h * 3), Image.NEAREST)
+
+                buf = BytesIO()
+                img.save(buf, format="PNG", compress_level=1)
+                return "image/png", buf.getvalue()
+
+            elif False:
+                # 48s, too slow
+                import pyvips
+
+                h = int(192 * h / w)
+                w = 192
+                img = pyvips.Image.text(
+                    ext, width=w, height=h, dpi=192, align=pyvips.Align.CENTRE
+                )
+                img = img.ifthenelse(ci[3:], ci[:3], blend=True)
+                # i = i.resize(3, kernel=pyvips.Kernel.NEAREST)
+                buf = img.write_to_buffer(".png[compression=1]")
+                return "image/png", buf
+
+        svg = """\
+<?xml version="1.0" encoding="UTF-8"?>
+<svg version="1.1" viewBox="0 0 100 {}" xmlns="http://www.w3.org/2000/svg"><g>
+<rect width="100%" height="100%" fill="#{}" />
+<text x="50%" y="50%" dominant-baseline="middle" text-anchor="middle" xml:space="preserve"
+  fill="#{}" font-family="monospace" font-size="14px" style="letter-spacing:.5px">{}</text>
+</g></svg>
+"""
+        svg = svg.format(h, c[:6], c[6:], ext)
+
+        return "image/svg+xml", svg.encode("utf-8")

copyparty/mdns.py

@@ -0,0 +1,507 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
+import random
+import select
+import socket
+import time
+
+from ipaddress import IPv4Network, IPv6Network
+
+from .__init__ import TYPE_CHECKING
+from .__init__ import unicode as U
+from .multicast import MC_Sck, MCast
+from .stolen.dnslib import CLASS as DC
+from .stolen.dnslib import (
+    NSEC,
+    PTR,
+    QTYPE,
+    RR,
+    SRV,
+    TXT,
+    A,
+    AAAA,
+    DNSHeader,
+    DNSQuestion,
+    DNSRecord,
+)
+from .util import CachedSet, Daemon, Netdev, min_ex
+
+if TYPE_CHECKING:
+    from .svchub import SvcHub
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Any, Optional, Union
+
+
+MDNS4 = "224.0.0.251"
+MDNS6 = "ff02::fb"
+
+
+class MDNS_Sck(MC_Sck):
+    def __init__(
+        self,
+        sck: socket.socket,
+        nd: Netdev,
+        grp: str,
+        ip: str,
+        net: Union[IPv4Network, IPv6Network],
+    ):
+        super(MDNS_Sck, self).__init__(sck, nd, grp, ip, net)
+
+        self.bp_probe = b""
+        self.bp_ip = b""
+        self.bp_svc = b""
+        self.bp_bye = b""
+
+        self.last_tx = 0.0
+
+
+class MDNS(MCast):
+    def __init__(self, hub: "SvcHub") -> None:
+        al = hub.args
+        grp4 = "" if al.zm6 else MDNS4
+        grp6 = "" if al.zm4 else MDNS6
+        super(MDNS, self).__init__(
+            hub, MDNS_Sck, al.zm_on, al.zm_off, grp4, grp6, 5353, hub.args.zmv
+        )
+        self.srv: dict[socket.socket, MDNS_Sck] = {}
+
+        self.ttl = 300
+
+        zs = self.args.name + ".local."
+        zs = zs.encode("ascii", "replace").decode("ascii", "replace")
+        self.hn = "-".join(x for x in zs.split("?") if x) or (
+            "vault-{}".format(random.randint(1, 255))
+        )
+        self.lhn = self.hn.lower()
+
+        # requester ip -> (response deadline, srv, body):
+        self.q: dict[str, tuple[float, MDNS_Sck, bytes]] = {}
+        self.rx4 = CachedSet(0.42)  # 3 probes @ 250..500..750 => 500ms span
+        self.rx6 = CachedSet(0.42)
+        self.svcs, self.sfqdns = self.build_svcs()
+        self.lsvcs = {k.lower(): v for k, v in self.svcs.items()}
+        self.lsfqdns = set([x.lower() for x in self.sfqdns])
+
+        self.probing = 0.0
+        self.unsolicited: list[float] = []  # scheduled announces on all nics
+        self.defend: dict[MDNS_Sck, float] = {}  # server -> deadline
+
+    def log(self, msg: str, c: Union[int, str] = 0) -> None:
+        self.log_func("mDNS", msg, c)
+
+    def build_svcs(self) -> tuple[dict[str, dict[str, Any]], set[str]]:
+        zms = self.args.zms
+        http = {"port": 80 if 80 in self.args.p else self.args.p[0]}
+        https = {"port": 443 if 443 in self.args.p else self.args.p[0]}
+        webdav = http.copy()
+        webdavs = https.copy()
+        webdav["u"] = webdavs["u"] = "u"  # KDE requires username
+        ftp = {"port": (self.args.ftp if "f" in zms else self.args.ftps)}
+        smb = {"port": self.args.smb_port}
+
+        # some gvfs require path
+        zs = self.args.zm_ld or "/"
+        if zs:
+            webdav["path"] = zs
+            webdavs["path"] = zs
+
+        if self.args.zm_lh:
+            http["path"] = self.args.zm_lh
+            https["path"] = self.args.zm_lh
+
+        if self.args.zm_lf:
+            ftp["path"] = self.args.zm_lf
+
+        if self.args.zm_ls:
+            smb["path"] = self.args.zm_ls
+
+        svcs: dict[str, dict[str, Any]] = {}
+
+        if "d" in zms:
+            svcs["_webdav._tcp.local."] = webdav
+
+        if "D" in zms:
+            svcs["_webdavs._tcp.local."] = webdavs
+
+        if "h" in zms:
+            svcs["_http._tcp.local."] = http
+
+        if "H" in zms:
+            svcs["_https._tcp.local."] = https
+
+        if "f" in zms.lower():
+            svcs["_ftp._tcp.local."] = ftp
+
+        if "s" in zms.lower():
+            svcs["_smb._tcp.local."] = smb
+
+        sfqdns: set[str] = set()
+        for k, v in svcs.items():
+            name = "{}-c-{}".format(self.args.name, k.split(".")[0][1:])
+            v["name"] = name
+            sfqdns.add("{}.{}".format(name, k))
+
+        return svcs, sfqdns
+
+    def build_replies(self) -> None:
+        for srv in self.srv.values():
+            probe = DNSRecord(DNSHeader(0, 0), q=DNSQuestion(self.hn, QTYPE.ANY))
+            areply = DNSRecord(DNSHeader(0, 0x8400))
+            sreply = DNSRecord(DNSHeader(0, 0x8400))
+            bye = DNSRecord(DNSHeader(0, 0x8400))
+
+            have4 = have6 = False
+            for s2 in self.srv.values():
+                if srv.idx != s2.idx:
+                    continue
+
+                if s2.v6:
+                    have6 = True
+                else:
+                    have4 = True
+
+            for ip in srv.ips:
+                if ":" in ip:
+                    qt = QTYPE.AAAA
+                    ar = {"rclass": DC.F_IN, "rdata": AAAA(ip)}
+                else:
+                    qt = QTYPE.A
+                    ar = {"rclass": DC.F_IN, "rdata": A(ip)}
+
+                r0 = RR(self.hn, qt, ttl=0, **ar)
+                r120 = RR(self.hn, qt, ttl=120, **ar)
+                # rfc-10:
+                #   SHOULD rr ttl 120sec for A/AAAA/SRV
+                #   (and recommend 75min for all others)
+
+                probe.add_auth(r120)
+                areply.add_answer(r120)
+                sreply.add_answer(r120)
+                bye.add_answer(r0)
+
+            for sclass, props in self.svcs.items():
+                sname = props["name"]
+                sport = props["port"]
+                sfqdn = sname + "." + sclass
+
+                k = "_services._dns-sd._udp.local."
+                r = RR(k, QTYPE.PTR, DC.IN, 4500, PTR(sclass))
+                sreply.add_answer(r)
+
+                r = RR(sclass, QTYPE.PTR, DC.IN, 4500, PTR(sfqdn))
+                sreply.add_answer(r)
+
+                r = RR(sfqdn, QTYPE.SRV, DC.F_IN, 120, SRV(0, 0, sport, self.hn))
+                sreply.add_answer(r)
+                areply.add_answer(r)
+
+                r = RR(sfqdn, QTYPE.SRV, DC.F_IN, 0, SRV(0, 0, sport, self.hn))
+                bye.add_answer(r)
+
+                txts = []
+                for k in ("u", "path"):
+                    if k not in props:
+                        continue
+
+                    zb = "{}={}".format(k, props[k]).encode("utf-8")
+                    if len(zb) > 255:
+                        t = "value too long for mdns: [{}]"
+                        raise Exception(t.format(props[k]))
+
+                    txts.append(zb)
+
+                # gvfs really wants txt even if they're empty
+                r = RR(sfqdn, QTYPE.TXT, DC.F_IN, 4500, TXT(txts))
+                sreply.add_answer(r)
+
+            if not (have4 and have6) and not self.args.zm_noneg:
+                ns = NSEC(self.hn, ["AAAA" if have6 else "A"])
+                r = RR(self.hn, QTYPE.NSEC, DC.F_IN, 120, ns)
+                areply.add_ar(r)
+                if len(sreply.pack()) < 1400:
+                    sreply.add_ar(r)
+
+            srv.bp_probe = probe.pack()
+            srv.bp_ip = areply.pack()
+            srv.bp_svc = sreply.pack()
+            srv.bp_bye = bye.pack()
+
+            # since all replies are small enough to fit in one packet,
+            # always send full replies rather than just a/aaaa records
+            srv.bp_ip = srv.bp_svc
+
+    def send_probes(self) -> None:
+        slp = random.random() * 0.25
+        for _ in range(3):
+            time.sleep(slp)
+            slp = 0.25
+            if not self.running:
+                break
+
+            if self.args.zmv:
+                self.log("sending hostname probe...")
+
+            # ipv4: need to probe each ip (each server)
+            # ipv6: only need to probe each set of looped nics
+            probed6: set[str] = set()
+            for srv in self.srv.values():
+                if srv.ip in probed6:
+                    continue
+
+                try:
+                    srv.sck.sendto(srv.bp_probe, (srv.grp, 5353))
+                    if srv.v6:
+                        for ip in srv.ips:
+                            probed6.add(ip)
+                except Exception as ex:
+                    self.log("sendto failed: {} ({})".format(srv.ip, ex), "90")
+
+    def run(self) -> None:
+        try:
+            bound = self.create_servers()
+        except:
+            t = "no server IP matches the mdns config\n{}"
+            self.log(t.format(min_ex()), 1)
+            bound = []
+
+        if not bound:
+            self.log("failed to announce copyparty services on the network", 3)
+            return
+
+        self.build_replies()
+        Daemon(self.send_probes)
+        zf = time.time() + 2
+        self.probing = zf  # cant unicast so give everyone an extra sec
+        self.unsolicited = [zf, zf + 1, zf + 3, zf + 7]  # rfc-8.3
+        last_hop = time.time()
+        ihop = self.args.mc_hop
+        while self.running:
+            timeout = (
+                0.02 + random.random() * 0.07
+                if self.probing or self.q or self.defend or self.unsolicited
+                else (last_hop + ihop if ihop else 180)
+            )
+            rdy = select.select(self.srv, [], [], timeout)
+            rx: list[socket.socket] = rdy[0]  # type: ignore
+            self.rx4.cln()
+            self.rx6.cln()
+            for sck in rx:
+                buf, addr = sck.recvfrom(4096)
+                try:
+                    self.eat(buf, addr, sck)
+                except:
+                    if not self.running:
+                        return
+
+                    t = "{} {} \033[33m|{}| {}\n{}".format(
+                        self.srv[sck].name, addr, len(buf), repr(buf)[2:-1], min_ex()
+                    )
+                    self.log(t, 6)
+
+            if not self.probing:
+                self.process()
+                continue
+
+            if self.probing < time.time():
+                t = "probe ok; announcing [{}]"
+                self.log(t.format(self.hn[:-1]), 2)
+                self.probing = 0
+
+    def stop(self, panic=False) -> None:
+        self.running = False
+        if not panic:
+            for srv in self.srv.values():
+                srv.sck.sendto(srv.bp_bye, (srv.grp, 5353))
+
+        self.srv = {}
+
+    def eat(self, buf: bytes, addr: tuple[str, int], sck: socket.socket) -> None:
+        cip = addr[0]
+        v6 = ":" in cip
+        if cip.startswith("169.254") or v6 and not cip.startswith("fe80"):
+            return
+
+        cache = self.rx6 if v6 else self.rx4
+        if buf in cache.c:
+            return
+
+        cache.add(buf)
+        srv: Optional[MDNS_Sck] = self.srv[sck] if v6 else self.map_client(cip)  # type: ignore
+        if not srv:
+            return
+
+        now = time.time()
+
+        if self.args.zmv and cip != srv.ip and cip not in srv.ips:
+            t = "{} [{}] \033[36m{} \033[0m|{}|"
+            self.log(t.format(srv.name, srv.ip, cip, len(buf)), "90")
+
+        p = DNSRecord.parse(buf)
+        if self.args.zmvv:
+            self.log(str(p))
+
+        # check for incoming probes for our hostname
+        cips = [U(x.rdata) for x in p.auth if U(x.rname).lower() == self.lhn]
+        if cips and self.sips.isdisjoint(cips):
+            if not [x for x in cips if x not in ("::1", "127.0.0.1")]:
+                # avahi broadcasting 127.0.0.1-only packets
+                return
+
+            self.log("someone trying to steal our hostname: {}".format(cips), 3)
+            # immediately unicast
+            if not self.probing:
+                srv.sck.sendto(srv.bp_ip, (cip, 5353))
+
+            # and schedule multicast
+            self.defend[srv] = self.defend.get(srv, now + 0.1)
+            return
+
+        # check for someone rejecting our probe / hijacking our hostname
+        cips = [
+            U(x.rdata)
+            for x in p.rr
+            if U(x.rname).lower() == self.lhn and x.rclass == DC.F_IN
+        ]
+        if cips and self.sips.isdisjoint(cips):
+            if not [x for x in cips if x not in ("::1", "127.0.0.1")]:
+                # avahi broadcasting 127.0.0.1-only packets
+                return
+
+            t = "mdns zeroconf: "
+            if self.probing:
+                t += "Cannot start; hostname '{}' is occupied"
+            else:
+                t += "Emergency stop; hostname '{}' got stolen"
+
+            t += " on {}! Use --name to set another hostname.\n\nName taken by {}\n\nYour IPs: {}\n"
+            self.log(t.format(self.args.name, srv.name, cips, list(self.sips)), 1)
+            self.stop(True)
+            return
+
+        # then rfc-6.7; dns pretending to be mdns (android...)
+        if p.header.id or addr[1] != 5353:
+            rsp: Optional[DNSRecord] = None
+            for r in p.questions:
+                try:
+                    lhn = U(r.qname).lower()
+                except:
+                    self.log("invalid question: {}".format(r))
+                    continue
+
+                if lhn != self.lhn:
+                    continue
+
+                if p.header.id and r.qtype in (QTYPE.A, QTYPE.AAAA):
+                    rsp = rsp or DNSRecord(DNSHeader(p.header.id, 0x8400))
+                    rsp.add_question(r)
+                    for ip in srv.ips:
+                        qt = r.qtype
+                        v6 = ":" in ip
+                        if v6 == (qt == QTYPE.AAAA):
+                            rd = AAAA(ip) if v6 else A(ip)
+                            rr = RR(self.hn, qt, DC.IN, 10, rd)
+                            rsp.add_answer(rr)
+            if rsp:
+                srv.sck.sendto(rsp.pack(), addr[:2])
+                # but don't return in case it's a differently broken client
+
+        # then a/aaaa records
+        for r in p.questions:
+            try:
+                lhn = U(r.qname).lower()
+            except:
+                self.log("invalid question: {}".format(r))
+                continue
+
+            if lhn != self.lhn:
+                continue
+
+            # gvfs keeps repeating itself
+            found = False
+            unicast = False
+            for rr in p.rr:
+                try:
+                    rname = U(rr.rname).lower()
+                except:
+                    self.log("invalid rr: {}".format(rr))
+                    continue
+
+                if rname == self.lhn:
+                    if rr.ttl > 60:
+                        found = True
+                    if rr.rclass == DC.F_IN:
+                        unicast = True
+
+            if unicast:
+                # spec-compliant mDNS-over-unicast
+                srv.sck.sendto(srv.bp_ip, (cip, 5353))
+            elif addr[1] != 5353:
+                # just in case some clients use (and want us to use) invalid ports
+                srv.sck.sendto(srv.bp_ip, addr[:2])
+
+            if not found:
+                self.q[cip] = (0, srv, srv.bp_ip)
+                return
+
+        deadline = now + (0.5 if p.header.tc else 0.02)  # rfc-7.2
+
+        # and service queries
+        for r in p.questions:
+            if not r or not r.qname:
+                continue
+
+            qname = U(r.qname).lower()
+            if qname in self.lsvcs or qname == "_services._dns-sd._udp.local.":
+                self.q[cip] = (deadline, srv, srv.bp_svc)
+                break
+        # heed rfc-7.1 if there was an announce in the past 12sec
+        # (workaround gvfs race-condition where it occasionally
+        #  doesn't read/decode the full response...)
+        if now < srv.last_tx + 12:
+            for rr in p.rr:
+                if not rr.rdata:
+                    continue
+
+                rdata = U(rr.rdata).lower()
+                if rdata in self.lsfqdns:
+                    if rr.ttl > 2250:
+                        self.q.pop(cip, None)
+                    break
+
+    def process(self) -> None:
+        tx = set()
+        now = time.time()
+        cooldown = 0.9  # rfc-6: 1
+        if self.unsolicited and self.unsolicited[0] < now:
+            self.unsolicited.pop(0)
+            cooldown = 0.1
+            for srv in self.srv.values():
+                tx.add(srv)
+
+        for srv, deadline in list(self.defend.items()):
+            if now < deadline:
+                continue
+
+            if self._tx(srv, srv.bp_ip, 0.02):  # rfc-6: 0.25
+                self.defend.pop(srv)
+
+        for cip, (deadline, srv, msg) in list(self.q.items()):
+            if now < deadline:
+                continue
+
+            self.q.pop(cip)
+            self._tx(srv, msg, cooldown)
+
+        for srv in tx:
+            self._tx(srv, srv.bp_svc, cooldown)
+
+    def _tx(self, srv: MDNS_Sck, msg: bytes, cooldown: float) -> bool:
+        now = time.time()
+        if now < srv.last_tx + cooldown:
+            return False
+
+        srv.sck.sendto(msg, (srv.grp, 5353))
+        srv.last_tx = now
+        return True

copyparty/mtag.py

@@ -0,0 +1,573 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
+import argparse
+import json
+import os
+import shutil
+import subprocess as sp
+import sys
+
+from .__init__ import PY2, WINDOWS, E, unicode
+from .bos import bos
+from .util import REKOBO_LKEY, fsenc, min_ex, retchk, runcmd, uncyg
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Any, Union
+
+    from .util import RootLogger
+
+
+def have_ff(scmd: str) -> bool:
+    if PY2:
+        print("# checking {}".format(scmd))
+        acmd = (scmd + " -version").encode("ascii").split(b" ")
+        try:
+            sp.Popen(acmd, stdout=sp.PIPE, stderr=sp.PIPE).communicate()
+            return True
+        except:
+            return False
+    else:
+        return bool(shutil.which(scmd))
+
+
+HAVE_FFMPEG = have_ff("ffmpeg")
+HAVE_FFPROBE = have_ff("ffprobe")
+
+
+class MParser(object):
+    def __init__(self, cmdline: str) -> None:
+        self.tag, args = cmdline.split("=", 1)
+        self.tags = self.tag.split(",")
+
+        self.timeout = 60
+        self.force = False
+        self.kill = "t"  # tree; all children recursively
+        self.capture = 3  # outputs to consume
+        self.audio = "y"
+        self.pri = 0  # priority; higher = later
+        self.ext = []
+
+        while True:
+            try:
+                bp = os.path.expanduser(args)
+                if WINDOWS:
+                    bp = uncyg(bp)
+
+                if bos.path.exists(bp):
+                    self.bin = bp
+                    return
+            except:
+                pass
+
+            arg, args = args.split(",", 1)
+            arg = arg.lower()
+
+            if arg.startswith("a"):
+                self.audio = arg[1:]  # [r]equire [n]ot [d]ontcare
+                continue
+
+            if arg.startswith("k"):
+                self.kill = arg[1:]  # [t]ree [m]ain [n]one
+                continue
+
+            if arg.startswith("c"):
+                self.capture = int(arg[1:])  # 0=none 1=stdout 2=stderr 3=both
+                continue
+
+            if arg == "f":
+                self.force = True
+                continue
+
+            if arg.startswith("t"):
+                self.timeout = int(arg[1:])
+                continue
+
+            if arg.startswith("e"):
+                self.ext.append(arg[1:])
+                continue
+
+            if arg.startswith("p"):
+                self.pri = int(arg[1:] or "1")
+                continue
+
+            raise Exception()
+
+
+def ffprobe(
+    abspath: str, timeout: int = 60
+) -> tuple[dict[str, tuple[int, Any]], dict[str, list[Any]]]:
+    cmd = [
+        b"ffprobe",
+        b"-hide_banner",
+        b"-show_streams",
+        b"-show_format",
+        b"--",
+        fsenc(abspath),
+    ]
+    rc, so, se = runcmd(cmd, timeout=timeout)
+    retchk(rc, cmd, se)
+    return parse_ffprobe(so)
+
+
+def parse_ffprobe(txt: str) -> tuple[dict[str, tuple[int, Any]], dict[str, list[Any]]]:
+    """ffprobe -show_format -show_streams"""
+    streams = []
+    fmt = {}
+    g = {}
+    for ln in [x.rstrip("\r") for x in txt.split("\n")]:
+        try:
+            sk, sv = ln.split("=", 1)
+            g[sk] = sv
+            continue
+        except:
+            pass
+
+        if ln == "[STREAM]":
+            g = {}
+            streams.append(g)
+
+        if ln == "[FORMAT]":
+            g = {"codec_type": "format"}  # heh
+            fmt = g
+
+    streams = [fmt] + streams
+    ret: dict[str, Any] = {}  # processed
+    md: dict[str, list[Any]] = {}  # raw tags
+
+    is_audio = fmt.get("format_name") in ["mp3", "ogg", "flac", "wav"]
+    if fmt.get("filename", "").split(".")[-1].lower() in ["m4a", "aac"]:
+        is_audio = True
+
+    # if audio file, ensure audio stream appears first
+    if (
+        is_audio
+        and len(streams) > 2
+        and streams[1].get("codec_type") != "audio"
+        and streams[2].get("codec_type") == "audio"
+    ):
+        streams = [fmt, streams[2], streams[1]] + streams[3:]
+
+    have = {}
+    for strm in streams:
+        typ = strm.get("codec_type")
+        if typ in have:
+            continue
+
+        have[typ] = True
+        kvm = []
+
+        if typ == "audio":
+            kvm = [
+                ["codec_name", "ac"],
+                ["channel_layout", "chs"],
+                ["sample_rate", ".hz"],
+                ["bit_rate", ".aq"],
+                ["duration", ".dur"],
+            ]
+
+        if typ == "video":
+            if strm.get("DISPOSITION:attached_pic") == "1" or is_audio:
+                continue
+
+            kvm = [
+                ["codec_name", "vc"],
+                ["pix_fmt", "pixfmt"],
+                ["r_frame_rate", ".fps"],
+                ["bit_rate", ".vq"],
+                ["width", ".resw"],
+                ["height", ".resh"],
+                ["duration", ".dur"],
+            ]
+
+        if typ == "format":
+            kvm = [["duration", ".dur"], ["bit_rate", ".q"], ["format_name", "fmt"]]
+
+        for sk, rk in kvm:
+            v1 = strm.get(sk)
+            if v1 is None:
+                continue
+
+            if rk.startswith("."):
+                try:
+                    zf = float(v1)
+                    v2 = ret.get(rk)
+                    if v2 is None or zf > v2:
+                        ret[rk] = zf
+                except:
+                    # sqlite doesnt care but the code below does
+                    if v1 not in ["N/A"]:
+                        ret[rk] = v1
+            else:
+                ret[rk] = v1
+
+    if ret.get("vc") == "ansi":  # shellscript
+        return {}, {}
+
+    for strm in streams:
+        for sk, sv in strm.items():
+            if not sk.startswith("TAG:"):
+                continue
+
+            sk = sk[4:].strip()
+            sv = sv.strip()
+            if sk and sv and sk not in md:
+                md[sk] = [sv]
+
+    for sk in [".q", ".vq", ".aq"]:
+        if sk in ret:
+            ret[sk] /= 1000  # bit_rate=320000
+
+    for sk in [".q", ".vq", ".aq", ".resw", ".resh"]:
+        if sk in ret:
+            ret[sk] = int(ret[sk])
+
+    if ".fps" in ret:
+        fps = ret[".fps"]
+        if "/" in fps:
+            fa, fb = fps.split("/")
+            try:
+                fps = int(fa) * 1.0 / int(fb)
+            except:
+                fps = 9001
+
+        if fps < 1000 and fmt.get("format_name") not in ["image2", "png_pipe"]:
+            ret[".fps"] = round(fps, 3)
+        else:
+            del ret[".fps"]
+
+    if ".dur" in ret:
+        if ret[".dur"] < 0.1:
+            del ret[".dur"]
+            if ".q" in ret:
+                del ret[".q"]
+
+    if "fmt" in ret:
+        ret["fmt"] = ret["fmt"].split(",")[0]
+
+    if ".resw" in ret and ".resh" in ret:
+        ret["res"] = "{}x{}".format(ret[".resw"], ret[".resh"])
+
+    zd = {k: (0, v) for k, v in ret.items()}
+
+    return zd, md
+
+
+class MTag(object):
+    def __init__(self, log_func: "RootLogger", args: argparse.Namespace) -> None:
+        self.log_func = log_func
+        self.args = args
+        self.usable = True
+        self.prefer_mt = not args.no_mtag_ff
+        self.backend = "ffprobe" if args.no_mutagen else "mutagen"
+        self.can_ffprobe = HAVE_FFPROBE and not args.no_mtag_ff
+        mappings = args.mtm
+        or_ffprobe = " or FFprobe"
+
+        if self.backend == "mutagen":
+            self.get = self.get_mutagen
+            try:
+                from mutagen import version  # noqa: F401
+            except:
+                self.log("could not load Mutagen, trying FFprobe instead", c=3)
+                self.backend = "ffprobe"
+
+        if self.backend == "ffprobe":
+            self.usable = self.can_ffprobe
+            self.get = self.get_ffprobe
+            self.prefer_mt = True
+
+            if not HAVE_FFPROBE:
+                pass
+
+            elif args.no_mtag_ff:
+                msg = "found FFprobe but it was disabled by --no-mtag-ff"
+                self.log(msg, c=3)
+
+        if not self.usable:
+            msg = "need Mutagen{} to read media tags so please run this:\n{}{} -m pip install --user mutagen\n"
+            pybin = os.path.basename(sys.executable)
+            self.log(msg.format(or_ffprobe, " " * 37, pybin), c=1)
+            return
+
+        # https://picard-docs.musicbrainz.org/downloads/MusicBrainz_Picard_Tag_Map.html
+        tagmap = {
+            "album": ["album", "talb", "\u00a9alb", "original-album", "toal"],
+            "artist": [
+                "artist",
+                "tpe1",
+                "\u00a9art",
+                "composer",
+                "performer",
+                "arranger",
+                "\u00a9wrt",
+                "tcom",
+                "tpe3",
+                "original-artist",
+                "tope",
+            ],
+            "title": ["title", "tit2", "\u00a9nam"],
+            "circle": [
+                "album-artist",
+                "tpe2",
+                "aart",
+                "conductor",
+                "organization",
+                "band",
+            ],
+            ".tn": ["tracknumber", "trck", "trkn", "track"],
+            "genre": ["genre", "tcon", "\u00a9gen"],
+            "date": [
+                "original-release-date",
+                "release-date",
+                "date",
+                "tdrc",
+                "\u00a9day",
+                "original-date",
+                "original-year",
+                "tyer",
+                "tdor",
+                "tory",
+                "year",
+                "creation-time",
+            ],
+            ".bpm": ["bpm", "tbpm", "tmpo", "tbp"],
+            "key": ["initial-key", "tkey", "key"],
+            "comment": ["comment", "comm", "\u00a9cmt", "comments", "description"],
+        }
+
+        if mappings:
+            for k, v in [x.split("=") for x in mappings]:
+                tagmap[k] = v.split(",")
+
+        self.tagmap = {}
+        for k, vs in tagmap.items():
+            vs2 = []
+            for v in vs:
+                if "-" not in v:
+                    vs2.append(v)
+                    continue
+
+                vs2.append(v.replace("-", " "))
+                vs2.append(v.replace("-", "_"))
+                vs2.append(v.replace("-", ""))
+
+            self.tagmap[k] = vs2
+
+        self.rmap = {
+            v: [n, k] for k, vs in self.tagmap.items() for n, v in enumerate(vs)
+        }
+        # self.get = self.compare
+
+    def log(self, msg: str, c: Union[int, str] = 0) -> None:
+        self.log_func("mtag", msg, c)
+
+    def normalize_tags(
+        self, parser_output: dict[str, tuple[int, Any]], md: dict[str, list[Any]]
+    ) -> dict[str, Union[str, float]]:
+        for sk, tv in dict(md).items():
+            if not tv:
+                continue
+
+            sk = sk.lower().split("::")[0].strip()
+            key_mapping = self.rmap.get(sk)
+            if not key_mapping:
+                continue
+
+            priority, alias = key_mapping
+            if alias not in parser_output or parser_output[alias][0] > priority:
+                parser_output[alias] = (priority, tv[0])
+
+        # take first value (lowest priority / most preferred)
+        ret: dict[str, Union[str, float]] = {
+            sk: unicode(tv[1]).strip() for sk, tv in parser_output.items()
+        }
+
+        # track 3/7 => track 3
+        for sk, zv in ret.items():
+            if sk[0] == ".":
+                sv = str(zv).split("/")[0].strip().lstrip("0")
+                ret[sk] = sv or 0
+
+        # normalize key notation to rkeobo
+        okey = ret.get("key")
+        if okey:
+            key = str(okey).replace(" ", "").replace("maj", "").replace("min", "m")
+            ret["key"] = REKOBO_LKEY.get(key.lower(), okey)
+
+        if self.args.mtag_vv:
+            zl = " ".join("\033[36m{} \033[33m{}".format(k, v) for k, v in ret.items())
+            self.log("norm: {}\033[0m".format(zl), "90")
+
+        return ret
+
+    def compare(self, abspath: str) -> dict[str, Union[str, float]]:
+        if abspath.endswith(".au"):
+            return {}
+
+        print("\n" + abspath)
+        r1 = self.get_mutagen(abspath)
+        r2 = self.get_ffprobe(abspath)
+
+        keys = {}
+        for d in [r1, r2]:
+            for k in d.keys():
+                keys[k] = True
+
+        diffs = []
+        l1 = []
+        l2 = []
+        for k in sorted(keys.keys()):
+            if k in [".q", ".dur"]:
+                continue  # lenient
+
+            v1 = r1.get(k)
+            v2 = r2.get(k)
+            if v1 == v2:
+                print("  ", k, v1)
+            elif v1 != "0000":  # FFprobe date=0
+                diffs.append(k)
+                print(" 1", k, v1)
+                print(" 2", k, v2)
+                if v1:
+                    l1.append(k)
+                if v2:
+                    l2.append(k)
+
+        if diffs:
+            raise Exception()
+
+        return r1
+
+    def get_mutagen(self, abspath: str) -> dict[str, Union[str, float]]:
+        ret: dict[str, tuple[int, Any]] = {}
+
+        if not bos.path.isfile(abspath):
+            return {}
+
+        from mutagen import File
+
+        try:
+            md = File(fsenc(abspath), easy=True)
+            assert md
+            if self.args.mtag_vv:
+                for zd in (md.info.__dict__, dict(md.tags)):
+                    zl = ["\033[36m{} \033[33m{}".format(k, v) for k, v in zd.items()]
+                    self.log("mutagen: {}\033[0m".format(" ".join(zl)), "90")
+            if not md.info.length and not md.info.codec:
+                raise Exception()
+        except:
+            return self.get_ffprobe(abspath) if self.can_ffprobe else {}
+
+        sz = bos.path.getsize(abspath)
+        try:
+            ret[".q"] = (0, int((sz / md.info.length) / 128))
+        except:
+            pass
+
+        for attr, k, norm in [
+            ["codec", "ac", unicode],
+            ["channels", "chs", int],
+            ["sample_rate", ".hz", int],
+            ["bitrate", ".aq", int],
+            ["length", ".dur", int],
+        ]:
+            try:
+                v = getattr(md.info, attr)
+            except:
+                if k != "ac":
+                    continue
+
+                try:
+                    v = str(md.info).split(".")[1]
+                    if v.startswith("ogg"):
+                        v = v[3:]
+                except:
+                    continue
+
+            if not v:
+                continue
+
+            if k == ".aq":
+                v /= 1000
+
+            if k == "ac" and v.startswith("mp4a.40."):
+                v = "aac"
+
+            ret[k] = (0, norm(v))
+
+        return self.normalize_tags(ret, md)
+
+    def get_ffprobe(self, abspath: str) -> dict[str, Union[str, float]]:
+        if not bos.path.isfile(abspath):
+            return {}
+
+        ret, md = ffprobe(abspath, self.args.mtag_to)
+
+        if self.args.mtag_vv:
+            for zd in (ret, dict(md)):
+                zl = ["\033[36m{} \033[33m{}".format(k, v) for k, v in zd.items()]
+                self.log("ffprobe: {}\033[0m".format(" ".join(zl)), "90")
+
+        return self.normalize_tags(ret, md)
+
+    def get_bin(
+        self, parsers: dict[str, MParser], abspath: str, oth_tags: dict[str, Any]
+    ) -> dict[str, Any]:
+        if not bos.path.isfile(abspath):
+            return {}
+
+        env = os.environ.copy()
+        try:
+            pypath = os.path.abspath(os.path.dirname(os.path.dirname(__file__)))
+            zsl = [str(pypath)] + [str(x) for x in sys.path if x]
+            pypath = str(os.pathsep.join(zsl))
+            env["PYTHONPATH"] = pypath
+        except:
+            if not E.ox:
+                raise
+
+        ret: dict[str, Any] = {}
+        for tagname, parser in sorted(parsers.items(), key=lambda x: (x[1].pri, x[0])):
+            try:
+                cmd = [parser.bin, abspath]
+                if parser.bin.endswith(".py"):
+                    cmd = [sys.executable] + cmd
+
+                args = {
+                    "env": env,
+                    "timeout": parser.timeout,
+                    "kill": parser.kill,
+                    "capture": parser.capture,
+                }
+
+                if parser.pri:
+                    zd = oth_tags.copy()
+                    zd.update(ret)
+                    args["sin"] = json.dumps(zd).encode("utf-8", "replace")
+
+                if WINDOWS:
+                    args["creationflags"] = 0x4000
+                else:
+                    cmd = ["nice"] + cmd
+
+                bcmd = [fsenc(x) for x in cmd]
+                rc, v, err = runcmd(bcmd, **args)  # type: ignore
+                retchk(rc, bcmd, err, self.log, 5, self.args.mtag_v)
+                v = v.strip()
+                if not v:
+                    continue
+
+                if "," not in tagname:
+                    ret[tagname] = v
+                else:
+                    zj = json.loads(v)
+                    for tag in tagname.split(","):
+                        if tag and tag in zj:
+                            ret[tag] = zj[tag]
+            except:
+                if self.args.mtag_v:
+                    t = "mtag error: tagname {}, parser {}, file {} => {}"
+                    self.log(t.format(tagname, parser.bin, abspath, min_ex()))
+
+        return ret

copyparty/multicast.py

@@ -0,0 +1,332 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
+import socket
+import time
+
+import ipaddress
+from ipaddress import IPv4Address, IPv4Network, IPv6Address, IPv6Network
+
+from .__init__ import TYPE_CHECKING
+from .util import MACOS, Netdev, min_ex, spack
+
+if TYPE_CHECKING:
+    from .svchub import SvcHub
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Optional, Union
+
+if not hasattr(socket, "IPPROTO_IPV6"):
+    setattr(socket, "IPPROTO_IPV6", 41)
+
+
+class NoIPs(Exception):
+    pass
+
+
+class MC_Sck(object):
+    """there is one socket for each server ip"""
+
+    def __init__(
+        self,
+        sck: socket.socket,
+        nd: Netdev,
+        grp: str,
+        ip: str,
+        net: Union[IPv4Network, IPv6Network],
+    ):
+        self.sck = sck
+        self.idx = nd.idx
+        self.name = nd.name
+        self.grp = grp
+        self.mreq = b""
+        self.ip = ip
+        self.net = net
+        self.ips = {ip: net}
+        self.v6 = ":" in ip
+        self.have4 = ":" not in ip
+        self.have6 = ":" in ip
+
+
+class MCast(object):
+    def __init__(
+        self,
+        hub: "SvcHub",
+        Srv: type[MC_Sck],
+        on: list[str],
+        off: list[str],
+        mc_grp_4: str,
+        mc_grp_6: str,
+        port: int,
+        vinit: bool,
+    ) -> None:
+        """disable ipv%d by setting mc_grp_%d empty"""
+        self.hub = hub
+        self.Srv = Srv
+        self.args = hub.args
+        self.asrv = hub.asrv
+        self.log_func = hub.log
+        self.on = on
+        self.off = off
+        self.grp4 = mc_grp_4
+        self.grp6 = mc_grp_6
+        self.port = port
+        self.vinit = vinit
+
+        self.srv: dict[socket.socket, MC_Sck] = {}  # listening sockets
+        self.sips: set[str] = set()  # all listening ips (including failed attempts)
+        self.b2srv: dict[bytes, MC_Sck] = {}  # binary-ip -> server socket
+        self.b4: list[bytes] = []  # sorted list of binary-ips
+        self.b6: list[bytes] = []  # sorted list of binary-ips
+        self.cscache: dict[str, Optional[MC_Sck]] = {}  # client ip -> server cache
+
+        self.running = True
+
+    def log(self, msg: str, c: Union[int, str] = 0) -> None:
+        self.log_func("multicast", msg, c)
+
+    def create_servers(self) -> list[str]:
+        bound: list[str] = []
+        netdevs = self.hub.tcpsrv.netdevs
+        ips = [x[0] for x in self.hub.tcpsrv.bound]
+
+        if "::" in ips:
+            ips = [x for x in ips if x != "::"] + list(
+                [x.split("/")[0] for x in netdevs if ":" in x]
+            )
+            ips.append("0.0.0.0")
+
+        if "0.0.0.0" in ips:
+            ips = [x for x in ips if x != "0.0.0.0"] + list(
+                [x.split("/")[0] for x in netdevs if ":" not in x]
+            )
+
+        ips = [x for x in ips if x not in ("::1", "127.0.0.1")]
+
+        # ip -> ip/prefix
+        ips = [[x for x in netdevs if x.startswith(y + "/")][0] for y in ips]
+
+        on = self.on[:]
+        off = self.off[:]
+        for lst in (on, off):
+            for av in list(lst):
+                for sk, sv in netdevs.items():
+                    if (av == str(sv.idx) or av == sv.name) and sk not in lst:
+                        lst.append(sk)
+
+        if on:
+            ips = [x for x in ips if x in on]
+        elif off:
+            ips = [x for x in ips if x not in off]
+
+        if not self.grp4:
+            ips = [x for x in ips if ":" in x]
+
+        if not self.grp6:
+            ips = [x for x in ips if ":" not in x]
+
+        ips = list(set(ips))
+        all_selected = ips[:]
+
+        # discard non-linklocal ipv6
+        ips = [x for x in ips if ":" not in x or x.startswith("fe80")]
+
+        if not ips:
+            raise NoIPs()
+
+        for ip in ips:
+            v6 = ":" in ip
+            netdev = netdevs[ip]
+            if not netdev.idx:
+                t = "using INADDR_ANY for ip [{}], netdev [{}]"
+                if not self.srv and ip not in ["::", "0.0.0.0"]:
+                    self.log(t.format(ip, netdev), 3)
+
+            ipv = socket.AF_INET6 if v6 else socket.AF_INET
+            sck = socket.socket(ipv, socket.SOCK_DGRAM, socket.IPPROTO_UDP)
+            sck.settimeout(None)
+            sck.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+            try:
+                sck.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEPORT, 1)
+            except:
+                pass
+
+            # most ipv6 clients expect multicast on linklocal ip only;
+            # add a/aaaa records for the other nic IPs
+            other_ips: set[str] = set()
+            if v6:
+                for nd in netdevs.values():
+                    if nd.idx == netdev.idx and nd.ip in all_selected and ":" in nd.ip:
+                        other_ips.add(nd.ip)
+
+            net = ipaddress.ip_network(ip, False)
+            ip = ip.split("/")[0]
+            srv = self.Srv(sck, netdev, self.grp6 if ":" in ip else self.grp4, ip, net)
+            for oth_ip in other_ips:
+                srv.ips[oth_ip.split("/")[0]] = ipaddress.ip_network(oth_ip, False)
+
+            # gvfs breaks if a linklocal ip appears in a dns reply
+            srv.ips = {k: v for k, v in srv.ips.items() if not k.startswith("fe80")}
+            if not srv.ips:
+                self.log("no routable IPs on {}; skipping [{}]".format(netdev, ip), 3)
+                continue
+
+            try:
+                self.setup_socket(srv)
+                self.srv[sck] = srv
+                bound.append(ip)
+            except:
+                t = "announce failed on {} [{}]:\n{}"
+                self.log(t.format(netdev, ip, min_ex()), 3)
+
+        if self.args.zm_msub:
+            for s1 in self.srv.values():
+                for s2 in self.srv.values():
+                    if s1.idx != s2.idx:
+                        continue
+
+                    if s1.ip not in s2.ips:
+                        s2.ips[s1.ip] = s1.net
+
+        if self.args.zm_mnic:
+            for s1 in self.srv.values():
+                for s2 in self.srv.values():
+                    for ip1, net1 in list(s1.ips.items()):
+                        for ip2, net2 in list(s2.ips.items()):
+                            if net1 == net2 and ip1 != ip2:
+                                s1.ips[ip2] = net2
+
+        self.sips = set([x.split("/")[0] for x in all_selected])
+        for srv in self.srv.values():
+            assert srv.ip in self.sips
+
+        return bound
+
+    def setup_socket(self, srv: MC_Sck) -> None:
+        sck = srv.sck
+        if srv.v6:
+            if self.vinit:
+                zsl = list(srv.ips.keys())
+                self.log("v6({}) idx({}) {}".format(srv.ip, srv.idx, zsl), 6)
+
+            for ip in srv.ips:
+                bip = socket.inet_pton(socket.AF_INET6, ip)
+                self.b2srv[bip] = srv
+                self.b6.append(bip)
+
+            grp = self.grp6 if srv.idx else ""
+            try:
+                if MACOS:
+                    raise Exception()
+
+                sck.bind((grp, self.port, 0, srv.idx))
+            except:
+                sck.bind(("", self.port, 0, srv.idx))
+
+            bgrp = socket.inet_pton(socket.AF_INET6, self.grp6)
+            dev = spack(b"@I", srv.idx)
+            srv.mreq = bgrp + dev
+            if srv.idx != socket.INADDR_ANY:
+                sck.setsockopt(socket.IPPROTO_IPV6, socket.IPV6_MULTICAST_IF, dev)
+
+            try:
+                sck.setsockopt(socket.IPPROTO_IPV6, socket.IPV6_MULTICAST_HOPS, 255)
+                sck.setsockopt(socket.IPPROTO_IPV6, socket.IPV6_MULTICAST_LOOP, 1)
+            except:
+                # macos
+                t = "failed to set IPv6 TTL/LOOP; announcements may not survive multiple switches/routers"
+                self.log(t, 3)
+        else:
+            if self.vinit:
+                self.log("v4({}) idx({})".format(srv.ip, srv.idx), 6)
+
+            bip = socket.inet_aton(srv.ip)
+            self.b2srv[bip] = srv
+            self.b4.append(bip)
+
+            grp = self.grp4 if srv.idx else ""
+            try:
+                if MACOS:
+                    raise Exception()
+
+                sck.bind((grp, self.port))
+            except:
+                sck.bind(("", self.port))
+
+            bgrp = socket.inet_aton(self.grp4)
+            dev = (
+                spack(b"=I", socket.INADDR_ANY)
+                if srv.idx == socket.INADDR_ANY
+                else socket.inet_aton(srv.ip)
+            )
+            srv.mreq = bgrp + dev
+            if srv.idx != socket.INADDR_ANY:
+                sck.setsockopt(socket.IPPROTO_IP, socket.IP_MULTICAST_IF, dev)
+
+            try:
+                sck.setsockopt(socket.IPPROTO_IP, socket.IP_MULTICAST_TTL, 255)
+                sck.setsockopt(socket.IPPROTO_IP, socket.IP_MULTICAST_LOOP, 1)
+            except:
+                # probably can't happen but dontcare if it does
+                t = "failed to set IPv4 TTL/LOOP; announcements may not survive multiple switches/routers"
+                self.log(t, 3)
+
+        self.hop(srv)
+        self.b4.sort(reverse=True)
+        self.b6.sort(reverse=True)
+
+    def hop(self, srv: MC_Sck) -> None:
+        """rejoin to keepalive on routers/switches without igmp-snooping"""
+        sck = srv.sck
+        req = srv.mreq
+        if ":" in srv.ip:
+            try:
+                sck.setsockopt(socket.IPPROTO_IPV6, socket.IPV6_LEAVE_GROUP, req)
+                # linux does leaves/joins twice with 0.2~1.05s spacing
+                time.sleep(1.2)
+            except:
+                pass
+
+            sck.setsockopt(socket.IPPROTO_IPV6, socket.IPV6_JOIN_GROUP, req)
+        else:
+            try:
+                sck.setsockopt(socket.IPPROTO_IP, socket.IP_DROP_MEMBERSHIP, req)
+                time.sleep(1.2)
+            except:
+                pass
+
+            # t = "joining {} from ip {} idx {} with mreq {}"
+            # self.log(t.format(srv.grp, srv.ip, srv.idx, repr(srv.mreq)), 6)
+            sck.setsockopt(socket.IPPROTO_IP, socket.IP_ADD_MEMBERSHIP, req)
+
+    def map_client(self, cip: str) -> Optional[MC_Sck]:
+        try:
+            return self.cscache[cip]
+        except:
+            pass
+
+        ret: Optional[MC_Sck] = None
+        v6 = ":" in cip
+        ci = IPv6Address(cip) if v6 else IPv4Address(cip)
+        for x in self.b6 if v6 else self.b4:
+            srv = self.b2srv[x]
+            if any([x for x in srv.ips.values() if ci in x]):
+                ret = srv
+                break
+
+        if not ret and cip in ("127.0.0.1", "::1"):
+            # just give it something
+            ret = list(self.srv.values())[0]
+
+        if ret:
+            t = "new client on {} ({}): {}"
+            self.log(t.format(ret.name, ret.net, cip), 6)
+        else:
+            t = "could not map client {} to known subnet; maybe forwarded from another network?"
+            self.log(t.format(cip), 3)
+
+        if len(self.cscache) > 9000:
+            self.cscache = {}
+
+        self.cscache[cip] = ret
+        return ret

copyparty/smbd.py

@@ -0,0 +1,321 @@
+# coding: utf-8
+
+import inspect
+import logging
+import os
+import random
+import stat
+import sys
+import time
+from types import SimpleNamespace
+
+from .__init__ import ANYWIN, TYPE_CHECKING
+from .authsrv import LEELOO_DALLAS, VFS
+from .bos import bos
+from .util import Daemon, min_ex
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Any
+
+if TYPE_CHECKING:
+    from .svchub import SvcHub
+
+
+lg = logging.getLogger("smb")
+debug, info, warning, error = (lg.debug, lg.info, lg.warning, lg.error)
+
+
+class SMB(object):
+    def __init__(self, hub: "SvcHub") -> None:
+        self.hub = hub
+        self.args = hub.args
+        self.asrv = hub.asrv
+        self.log = hub.log
+        self.files: dict[int, tuple[float, str]] = {}
+
+        lg.setLevel(logging.DEBUG if self.args.smbvvv else logging.INFO)
+        for x in ["impacket", "impacket.smbserver"]:
+            lgr = logging.getLogger(x)
+            lgr.setLevel(logging.DEBUG if self.args.smbvv else logging.INFO)
+
+        try:
+            from impacket import smbserver
+            from impacket.ntlm import compute_lmhash, compute_nthash
+        except ImportError:
+            m = "\033[36m\n{}\033[31m\n\nERROR: need 'impacket'; please run this command:\033[33m\n {} -m pip install --user impacket\n\033[0m"
+            print(m.format(min_ex(), sys.executable))
+            sys.exit(1)
+
+        # patch vfs into smbserver.os
+        fos = SimpleNamespace()
+        for k in os.__dict__:
+            try:
+                setattr(fos, k, getattr(os, k))
+            except:
+                pass
+        fos.close = self._close
+        fos.listdir = self._listdir
+        fos.mkdir = self._mkdir
+        fos.open = self._open
+        fos.remove = self._unlink
+        fos.rename = self._rename
+        fos.stat = self._stat
+        fos.unlink = self._unlink
+        fos.utime = self._utime
+        smbserver.os = fos
+
+        # ...and smbserver.os.path
+        fop = SimpleNamespace()
+        for k in os.path.__dict__:
+            try:
+                setattr(fop, k, getattr(os.path, k))
+            except:
+                pass
+        fop.exists = self._p_exists
+        fop.getsize = self._p_getsize
+        fop.isdir = self._p_isdir
+        smbserver.os.path = fop
+
+        if not self.args.smb_nwa_2:
+            fop.join = self._p_join
+
+        # other patches
+        smbserver.isInFileJail = self._is_in_file_jail
+        self._disarm()
+
+        ip = next((x for x in self.args.i if ":" not in x), None)
+        if not ip:
+            self.log("smb", "IPv6 not supported for SMB; listening on 0.0.0.0", 3)
+            ip = "0.0.0.0"
+
+        port = int(self.args.smb_port)
+        srv = smbserver.SimpleSMBServer(listenAddress=ip, listenPort=port)
+
+        ro = "no" if self.args.smbw else "yes"  # (does nothing)
+        srv.addShare("A", "/", readOnly=ro)
+        srv.setSMB2Support(not self.args.smb1)
+
+        for name, pwd in self.asrv.acct.items():
+            for u, p in ((name, pwd), (pwd, "k")):
+                lmhash = compute_lmhash(p)
+                nthash = compute_nthash(p)
+                srv.addCredential(u, 0, lmhash, nthash)
+
+        chi = [random.randint(0, 255) for x in range(8)]
+        cha = "".join(["{:02x}".format(x) for x in chi])
+        srv.setSMBChallenge(cha)
+
+        self.srv = srv
+        self.stop = srv.stop
+        self.log("smb", "listening @ {}:{}".format(ip, port))
+
+    def start(self) -> None:
+        Daemon(self.srv.start)
+
+    def _v2a(self, caller: str, vpath: str, *a: Any) -> tuple[VFS, str]:
+        vpath = vpath.replace("\\", "/").lstrip("/")
+        # cf = inspect.currentframe().f_back
+        # c1 = cf.f_back.f_code.co_name
+        # c2 = cf.f_code.co_name
+        debug('%s("%s", %s)\033[K\033[0m', caller, vpath, str(a))
+
+        # TODO find a way to grab `identity` in smbComSessionSetupAndX and smb2SessionSetup
+        vfs, rem = self.asrv.vfs.get(vpath, LEELOO_DALLAS, True, True)
+        return vfs, vfs.canonical(rem)
+
+    def _listdir(self, vpath: str, *a: Any, **ka: Any) -> list[str]:
+        vpath = vpath.replace("\\", "/").lstrip("/")
+        # caller = inspect.currentframe().f_back.f_code.co_name
+        debug('listdir("%s", %s)\033[K\033[0m', vpath, str(a))
+        vfs, rem = self.asrv.vfs.get(vpath, LEELOO_DALLAS, False, False)
+        _, vfs_ls, vfs_virt = vfs.ls(
+            rem, LEELOO_DALLAS, not self.args.no_scandir, [[False, False]]
+        )
+        dirs = [x[0] for x in vfs_ls if stat.S_ISDIR(x[1].st_mode)]
+        fils = [x[0] for x in vfs_ls if x[0] not in dirs]
+        ls = list(vfs_virt.keys()) + dirs + fils
+        if self.args.smb_nwa_1:
+            return ls
+
+        # clients crash somewhere around 65760 byte
+        ret = []
+        sz = 112 * 2  # ['.', '..']
+        for n, fn in enumerate(ls):
+            if sz >= 64000:
+                t = "listing only %d of %d files (%d byte); see impacket#1433"
+                warning(t, n, len(ls), sz)
+                break
+
+            nsz = len(fn.encode("utf-16", "replace"))
+            nsz = ((nsz + 7) // 8) * 8
+            sz += 104 + nsz
+            ret.append(fn)
+
+        return ret
+
+    def _open(
+        self, vpath: str, flags: int, *a: Any, chmod: int = 0o777, **ka: Any
+    ) -> Any:
+        f_ro = os.O_RDONLY
+        if ANYWIN:
+            f_ro |= os.O_BINARY
+
+        wr = flags != f_ro
+        if wr and not self.args.smbw:
+            yeet("blocked write (no --smbw): " + vpath)
+
+        vfs, ap = self._v2a("open", vpath, *a)
+        if wr and not vfs.axs.uwrite:
+            yeet("blocked write (no-write-acc): " + vpath)
+
+        ret = bos.open(ap, flags, *a, mode=chmod, **ka)
+        if wr:
+            now = time.time()
+            nf = len(self.files)
+            if nf > 9000:
+                oldest = min([x[0] for x in self.files.values()])
+                cutoff = oldest + (now - oldest) / 2
+                self.files = {k: v for k, v in self.files.items() if v[0] > cutoff}
+                info("was tracking %d files, now %d", nf, len(self.files))
+
+            vpath = vpath.replace("\\", "/").lstrip("/")
+            self.files[ret] = (now, vpath)
+
+        return ret
+
+    def _close(self, fd: int) -> None:
+        os.close(fd)
+        if fd not in self.files:
+            return
+
+        _, vp = self.files.pop(fd)
+        vp, fn = os.path.split(vp)
+        vfs, rem = self.hub.asrv.vfs.get(vp, LEELOO_DALLAS, False, True)
+        vfs, rem = vfs.get_dbv(rem)
+        self.hub.up2k.hash_file(
+            vfs.realpath,
+            vfs.flags,
+            rem,
+            fn,
+            "1.7.6.2",
+            time.time(),
+        )
+
+    def _rename(self, vp1: str, vp2: str) -> None:
+        if not self.args.smbw:
+            yeet("blocked rename (no --smbw): " + vp1)
+
+        vp1 = vp1.lstrip("/")
+        vp2 = vp2.lstrip("/")
+
+        vfs2, ap2 = self._v2a("rename", vp2, vp1)
+        if not vfs2.axs.uwrite:
+            yeet("blocked rename (no-write-acc): " + vp2)
+
+        vfs1, _ = self.asrv.vfs.get(vp1, LEELOO_DALLAS, True, True)
+        if not vfs1.axs.umove:
+            yeet("blocked rename (no-move-acc): " + vp1)
+
+        self.hub.up2k.handle_mv(LEELOO_DALLAS, vp1, vp2)
+        try:
+            bos.makedirs(ap2)
+        except:
+            pass
+
+    def _mkdir(self, vpath: str) -> None:
+        if not self.args.smbw:
+            yeet("blocked mkdir (no --smbw): " + vpath)
+
+        vfs, ap = self._v2a("mkdir", vpath)
+        if not vfs.axs.uwrite:
+            yeet("blocked mkdir (no-write-acc): " + vpath)
+
+        return bos.mkdir(ap)
+
+    def _stat(self, vpath: str, *a: Any, **ka: Any) -> os.stat_result:
+        return bos.stat(self._v2a("stat", vpath, *a)[1], *a, **ka)
+
+    def _unlink(self, vpath: str) -> None:
+        if not self.args.smbw:
+            yeet("blocked delete (no --smbw): " + vpath)
+
+        # return bos.unlink(self._v2a("stat", vpath, *a)[1])
+        vfs, ap = self._v2a("delete", vpath)
+        if not vfs.axs.udel:
+            yeet("blocked delete (no-del-acc): " + vpath)
+
+        vpath = vpath.replace("\\", "/").lstrip("/")
+        self.hub.up2k.handle_rm(LEELOO_DALLAS, "1.7.6.2", [vpath], [])
+
+    def _utime(self, vpath: str, times: tuple[float, float]) -> None:
+        if not self.args.smbw:
+            yeet("blocked utime (no --smbw): " + vpath)
+
+        vfs, ap = self._v2a("utime", vpath)
+        if not vfs.axs.uwrite:
+            yeet("blocked utime (no-write-acc): " + vpath)
+
+        return bos.utime(ap, times)
+
+    def _p_exists(self, vpath: str) -> bool:
+        try:
+            bos.stat(self._v2a("p.exists", vpath)[1])
+            return True
+        except:
+            return False
+
+    def _p_getsize(self, vpath: str) -> int:
+        st = bos.stat(self._v2a("p.getsize", vpath)[1])
+        return st.st_size
+
+    def _p_isdir(self, vpath: str) -> bool:
+        try:
+            st = bos.stat(self._v2a("p.isdir", vpath)[1])
+            return stat.S_ISDIR(st.st_mode)
+        except:
+            return False
+
+    def _p_join(self, *a) -> str:
+        # impacket.smbserver reads globs from queryDirectoryRequest['Buffer']
+        # where somehow `fds.*` becomes `fds"*` so lets fix that
+        ret = os.path.join(*a)
+        return ret.replace('"', ".")  # type: ignore
+
+    def _hook(self, *a: Any, **ka: Any) -> None:
+        src = inspect.currentframe().f_back.f_code.co_name
+        error("\033[31m%s:hook(%s)\033[0m", src, a)
+        raise Exception("nope")
+
+    def _disarm(self) -> None:
+        from impacket import smbserver
+
+        smbserver.os.chmod = self._hook
+        smbserver.os.chown = self._hook
+        smbserver.os.ftruncate = self._hook
+        smbserver.os.lchown = self._hook
+        smbserver.os.link = self._hook
+        smbserver.os.lstat = self._hook
+        smbserver.os.replace = self._hook
+        smbserver.os.scandir = self._hook
+        smbserver.os.symlink = self._hook
+        smbserver.os.truncate = self._hook
+        smbserver.os.walk = self._hook
+
+        smbserver.os.path.abspath = self._hook
+        smbserver.os.path.expanduser = self._hook
+        smbserver.os.path.getatime = self._hook
+        smbserver.os.path.getctime = self._hook
+        smbserver.os.path.getmtime = self._hook
+        smbserver.os.path.isabs = self._hook
+        smbserver.os.path.isfile = self._hook
+        smbserver.os.path.islink = self._hook
+        smbserver.os.path.realpath = self._hook
+
+    def _is_in_file_jail(self, *a: Any) -> bool:
+        # handled by vfs
+        return True
+
+
+def yeet(msg: str) -> None:
+    info(msg)
+    raise Exception(msg)

copyparty/ssdp.py

@@ -0,0 +1,194 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
+import re
+import select
+import socket
+from email.utils import formatdate
+
+from .__init__ import TYPE_CHECKING
+from .multicast import MC_Sck, MCast
+from .util import CachedSet, min_ex
+
+if TYPE_CHECKING:
+    from .broker_util import BrokerCli
+    from .httpcli import HttpCli
+    from .svchub import SvcHub
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Optional, Union
+
+
+GRP = "239.255.255.250"
+
+
+class SSDP_Sck(MC_Sck):
+    def __init__(self, *a):
+        super(SSDP_Sck, self).__init__(*a)
+        self.hport = 0
+
+
+class SSDPr(object):
+    """generates http responses for httpcli"""
+
+    def __init__(self, broker: "BrokerCli") -> None:
+        self.broker = broker
+        self.args = broker.args
+
+    def reply(self, hc: "HttpCli") -> bool:
+        if hc.vpath.endswith("device.xml"):
+            return self.tx_device(hc)
+
+        hc.reply(b"unknown request", 400)
+        return False
+
+    def tx_device(self, hc: "HttpCli") -> bool:
+        zs = """
+<?xml version="1.0"?>
+<root xmlns="urn:schemas-upnp-org:device-1-0">
+    <specVersion>
+        <major>1</major>
+        <minor>0</minor>
+    </specVersion>
+    <URLBase>{}</URLBase>
+    <device>
+        <presentationURL>{}</presentationURL>
+        <deviceType>urn:schemas-upnp-org:device:Basic:1</deviceType>
+        <friendlyName>{}</friendlyName>
+        <modelDescription>file server</modelDescription>
+        <manufacturer>ed</manufacturer>
+        <manufacturerURL>https://ocv.me/</manufacturerURL>
+        <modelName>copyparty</modelName>
+        <modelURL>https://github.com/9001/copyparty/</modelURL>
+        <UDN>{}</UDN>
+        <serviceList>
+            <service>
+                <serviceType>urn:schemas-upnp-org:device:Basic:1</serviceType>
+                <serviceId>urn:schemas-upnp-org:device:Basic</serviceId>
+                <controlURL>/.cpr/ssdp/services.xml</controlURL>
+                <eventSubURL>/.cpr/ssdp/services.xml</eventSubURL>
+                <SCPDURL>/.cpr/ssdp/services.xml</SCPDURL>
+            </service>
+        </serviceList>
+    </device>
+</root>"""
+
+        sip, sport = hc.s.getsockname()[:2]
+        proto = "https" if self.args.https_only else "http"
+        ubase = "{}://{}:{}".format(proto, sip, sport)
+        zsl = self.args.zsl
+        url = zsl if "://" in zsl else ubase + "/" + zsl.lstrip("/")
+        name = "{} @ {}".format(self.args.doctitle, self.args.name)
+        zs = zs.strip().format(ubase, url, name, self.args.zsid)
+        hc.reply(zs.encode("utf-8", "replace"))
+        return False  # close connectino
+
+
+class SSDPd(MCast):
+    """communicates with ssdp clients over multicast"""
+
+    def __init__(self, hub: "SvcHub") -> None:
+        al = hub.args
+        vinit = al.zsv and not al.zmv
+        super(SSDPd, self).__init__(
+            hub, SSDP_Sck, al.zs_on, al.zs_off, GRP, "", 1900, vinit
+        )
+        self.srv: dict[socket.socket, SSDP_Sck] = {}
+        self.rxc = CachedSet(0.7)
+        self.txc = CachedSet(5)  # win10: every 3 sec
+        self.ptn_st = re.compile(b"\nst: *upnp:rootdevice", re.I)
+
+    def log(self, msg: str, c: Union[int, str] = 0) -> None:
+        self.log_func("SSDP", msg, c)
+
+    def run(self) -> None:
+        try:
+            bound = self.create_servers()
+        except:
+            t = "no server IP matches the ssdp config\n{}"
+            self.log(t.format(min_ex()), 1)
+            bound = []
+
+        if not bound:
+            self.log("failed to announce copyparty services on the network", 3)
+            return
+
+        # find http port for this listening ip
+        for srv in self.srv.values():
+            tcps = self.hub.tcpsrv.bound
+            hp = next((x[1] for x in tcps if x[0] in ("0.0.0.0", srv.ip)), 0)
+            hp = hp or next((x[1] for x in tcps if x[0] == "::"), 0)
+            if not hp:
+                hp = tcps[0][1]
+                self.log("assuming port {} for {}".format(hp, srv.ip), 3)
+            srv.hport = hp
+
+        self.log("listening")
+        while self.running:
+            rdy = select.select(self.srv, [], [], 180)
+            rx: list[socket.socket] = rdy[0]  # type: ignore
+            self.rxc.cln()
+            for sck in rx:
+                buf, addr = sck.recvfrom(4096)
+                try:
+                    self.eat(buf, addr)
+                except:
+                    if not self.running:
+                        return
+
+                    t = "{} {} \033[33m|{}| {}\n{}".format(
+                        self.srv[sck].name, addr, len(buf), repr(buf)[2:-1], min_ex()
+                    )
+                    self.log(t, 6)
+
+    def stop(self) -> None:
+        self.running = False
+        self.srv = {}
+
+    def eat(self, buf: bytes, addr: tuple[str, int]) -> None:
+        cip = addr[0]
+        if cip.startswith("169.254"):
+            return
+
+        if buf in self.rxc.c:
+            return
+
+        self.rxc.add(buf)
+        srv: Optional[SSDP_Sck] = self.map_client(cip)  # type: ignore
+        if not srv:
+            return
+
+        if not buf.startswith(b"M-SEARCH * HTTP/1."):
+            raise Exception("not an ssdp message")
+
+        if not self.ptn_st.search(buf):
+            return
+
+        if self.args.zsv:
+            t = "{} [{}] \033[36m{} \033[0m|{}|"
+            self.log(t.format(srv.name, srv.ip, cip, len(buf)), "90")
+
+        zs = """
+HTTP/1.1 200 OK
+CACHE-CONTROL: max-age=1800
+DATE: {0}
+EXT:
+LOCATION: http://{1}:{2}/.cpr/ssdp/device.xml
+OPT: "http://schemas.upnp.org/upnp/1/0/"; ns=01
+01-NLS: {3}
+SERVER: UPnP/1.0
+ST: upnp:rootdevice
+USN: {3}::upnp:rootdevice
+BOOTID.UPNP.ORG: 0
+CONFIGID.UPNP.ORG: 1
+
+"""
+        zs = zs.format(formatdate(usegmt=True), srv.ip, srv.hport, self.args.zsid)
+        zb = zs[1:].replace("\n", "\r\n").encode("utf-8", "replace")
+        srv.sck.sendto(zb, addr[:2])
+
+        if cip not in self.txc.c:
+            self.log("{} [{}] --> {}".format(srv.name, srv.ip, cip), "6")
+
+        self.txc.add(cip)
+        self.txc.cln()

copyparty/star.py

@@ -0,0 +1,112 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
+import tarfile
+
+from queue import Queue
+
+from .bos import bos
+from .sutil import StreamArc, errdesc
+from .util import Daemon, fsenc, min_ex
+
+if True:  # pylint: disable=using-constant-test
+    from typing import Any, Generator, Optional
+
+    from .util import NamedLogger
+
+
+class QFile(object):  # inherit io.StringIO for painful typing
+    """file-like object which buffers writes into a queue"""
+
+    def __init__(self) -> None:
+        self.q: Queue[Optional[bytes]] = Queue(64)
+        self.bq: list[bytes] = []
+        self.nq = 0
+
+    def write(self, buf: Optional[bytes]) -> None:
+        if buf is None or self.nq >= 240 * 1024:
+            self.q.put(b"".join(self.bq))
+            self.bq = []
+            self.nq = 0
+
+        if buf is None:
+            self.q.put(None)
+        else:
+            self.bq.append(buf)
+            self.nq += len(buf)
+
+
+class StreamTar(StreamArc):
+    """construct in-memory tar file from the given path"""
+
+    def __init__(
+        self,
+        log: "NamedLogger",
+        fgen: Generator[dict[str, Any], None, None],
+        **kwargs: Any
+    ):
+        super(StreamTar, self).__init__(log, fgen)
+
+        self.ci = 0
+        self.co = 0
+        self.qfile = QFile()
+        self.errf: dict[str, Any] = {}
+
+        # python 3.8 changed to PAX_FORMAT as default,
+        # waste of space and don't care about the new features
+        fmt = tarfile.GNU_FORMAT
+        self.tar = tarfile.open(fileobj=self.qfile, mode="w|", format=fmt)  # type: ignore
+
+        Daemon(self._gen, "star-gen")
+
+    def gen(self) -> Generator[Optional[bytes], None, None]:
+        try:
+            while True:
+                buf = self.qfile.q.get()
+                if not buf:
+                    break
+
+                self.co += len(buf)
+                yield buf
+
+            yield None
+        finally:
+            if self.errf:
+                bos.unlink(self.errf["ap"])
+
+    def ser(self, f: dict[str, Any]) -> None:
+        name = f["vp"]
+        src = f["ap"]
+        fsi = f["st"]
+
+        inf = tarfile.TarInfo(name=name)
+        inf.mode = fsi.st_mode
+        inf.size = fsi.st_size
+        inf.mtime = fsi.st_mtime
+        inf.uid = 0
+        inf.gid = 0
+
+        self.ci += inf.size
+        with open(fsenc(src), "rb", 512 * 1024) as fo:
+            self.tar.addfile(inf, fo)
+
+    def _gen(self) -> None:
+        errors = []
+        for f in self.fgen:
+            if "err" in f:
+                errors.append((f["vp"], f["err"]))
+                continue
+
+            try:
+                self.ser(f)
+            except:
+                ex = min_ex(5, True).replace("\n", "\n-- ")
+                errors.append((f["vp"], ex))
+
+        if errors:
+            self.errf, txt = errdesc(errors)
+            self.log("\n".join(([repr(self.errf)] + txt[1:])))
+            self.ser(self.errf)
+
+        self.tar.close()
+        self.qfile.write(None)

copyparty/stolen/dnslib/README.md

@@ -0,0 +1,5 @@
+`dnslib` but heavily simplified/feature-stripped
+
+L: MIT
+Copyright (c) 2010 - 2017 Paul Chakravarti
+https://github.com/paulc/dnslib/

copyparty/stolen/dnslib/__init__.py

@@ -0,0 +1,11 @@
+# coding: utf-8
+
+"""
+L: MIT
+Copyright (c) 2010 - 2017 Paul Chakravarti
+https://github.com/paulc/dnslib/tree/0.9.23
+"""
+
+from .dns import *
+
+version = "0.9.23"

copyparty/stolen/dnslib/bimap.py

@@ -0,0 +1,41 @@
+# coding: utf-8
+
+import types
+
+
+class BimapError(Exception):
+    pass
+
+
+class Bimap(object):
+    def __init__(self, name, forward, error=AttributeError):
+        self.name = name
+        self.error = error
+        self.forward = forward.copy()
+        self.reverse = dict([(v, k) for (k, v) in list(forward.items())])
+
+    def get(self, k, default=None):
+        try:
+            return self.forward[k]
+        except KeyError:
+            return default or str(k)
+
+    def __getitem__(self, k):
+        try:
+            return self.forward[k]
+        except KeyError:
+            if isinstance(self.error, types.FunctionType):
+                return self.error(self.name, k, True)
+            else:
+                raise self.error("%s: Invalid forward lookup: [%s]" % (self.name, k))
+
+    def __getattr__(self, k):
+        try:
+            if k == "__wrapped__":
+                raise AttributeError()
+            return self.reverse[k]
+        except KeyError:
+            if isinstance(self.error, types.FunctionType):
+                return self.error(self.name, k, False)
+            else:
+                raise self.error("%s: Invalid reverse lookup: [%s]" % (self.name, k))

copyparty/stolen/dnslib/bit.py

@@ -0,0 +1,15 @@
+# coding: utf-8
+
+from __future__ import print_function
+
+
+def get_bits(data, offset, bits=1):
+    mask = ((1 << bits) - 1) << offset
+    return (data & mask) >> offset
+
+
+def set_bits(data, value, offset, bits=1):
+    mask = ((1 << bits) - 1) << offset
+    clear = 0xFFFF ^ mask
+    data = (data & clear) | ((value << offset) & mask)
+    return data

copyparty/stolen/dnslib/buffer.py

@@ -0,0 +1,56 @@
+# coding: utf-8
+
+import binascii
+import struct
+
+
+class BufferError(Exception):
+    pass
+
+
+class Buffer(object):
+    def __init__(self, data=b""):
+        self.data = bytearray(data)
+        self.offset = 0
+
+    def remaining(self):
+        return len(self.data) - self.offset
+
+    def get(self, length):
+        if length > self.remaining():
+            raise BufferError(
+                "Not enough bytes [offset=%d,remaining=%d,requested=%d]"
+                % (self.offset, self.remaining(), length)
+            )
+        start = self.offset
+        end = self.offset + length
+        self.offset += length
+        return bytes(self.data[start:end])
+
+    def hex(self):
+        return binascii.hexlify(self.data)
+
+    def pack(self, fmt, *args):
+        self.offset += struct.calcsize(fmt)
+        self.data += struct.pack(fmt, *args)
+
+    def append(self, s):
+        self.offset += len(s)
+        self.data += s
+
+    def update(self, ptr, fmt, *args):
+        s = struct.pack(fmt, *args)
+        self.data[ptr : ptr + len(s)] = s
+
+    def unpack(self, fmt):
+        try:
+            data = self.get(struct.calcsize(fmt))
+            return struct.unpack(fmt, data)
+        except struct.error:
+            raise BufferError(
+                "Error unpacking struct '%s' <%s>"
+                % (fmt, binascii.hexlify(data).decode())
+            )
+
+    def __len__(self):
+        return len(self.data)

copyparty/stolen/dnslib/dns.py

@@ -0,0 +1,775 @@
+# coding: utf-8
+
+from __future__ import print_function
+
+import binascii
+from itertools import chain
+
+from .bimap import Bimap, BimapError
+from .bit import get_bits, set_bits
+from .buffer import BufferError
+from .label import DNSBuffer, DNSLabel
+from .ranges import IP4, IP6, H, I, check_bytes
+
+
+class DNSError(Exception):
+    pass
+
+
+def unknown_qtype(name, key, forward):
+    if forward:
+        try:
+            return "TYPE%d" % (key,)
+        except:
+            raise DNSError("%s: Invalid forward lookup: [%s]" % (name, key))
+    else:
+        if key.startswith("TYPE"):
+            try:
+                return int(key[4:])
+            except:
+                pass
+        raise DNSError("%s: Invalid reverse lookup: [%s]" % (name, key))
+
+
+QTYPE = Bimap(
+    "QTYPE",
+    {1: "A", 12: "PTR", 16: "TXT", 28: "AAAA", 33: "SRV", 47: "NSEC", 255: "ANY"},
+    unknown_qtype,
+)
+
+CLASS = Bimap("CLASS", {1: "IN", 254: "None", 255: "*", 0x8001: "F_IN"}, DNSError)
+
+QR = Bimap("QR", {0: "QUERY", 1: "RESPONSE"}, DNSError)
+
+RCODE = Bimap(
+    "RCODE",
+    {
+        0: "NOERROR",
+        1: "FORMERR",
+        2: "SERVFAIL",
+        3: "NXDOMAIN",
+        4: "NOTIMP",
+        5: "REFUSED",
+        6: "YXDOMAIN",
+        7: "YXRRSET",
+        8: "NXRRSET",
+        9: "NOTAUTH",
+        10: "NOTZONE",
+    },
+    DNSError,
+)
+
+OPCODE = Bimap(
+    "OPCODE", {0: "QUERY", 1: "IQUERY", 2: "STATUS", 4: "NOTIFY", 5: "UPDATE"}, DNSError
+)
+
+
+def label(label, origin=None):
+    if label.endswith("."):
+        return DNSLabel(label)
+    else:
+        return (origin if isinstance(origin, DNSLabel) else DNSLabel(origin)).add(label)
+
+
+class DNSRecord(object):
+    @classmethod
+    def parse(cls, packet) -> "DNSRecord":
+        buffer = DNSBuffer(packet)
+        try:
+            header = DNSHeader.parse(buffer)
+            questions = []
+            rr = []
+            auth = []
+            ar = []
+            for i in range(header.q):
+                questions.append(DNSQuestion.parse(buffer))
+            for i in range(header.a):
+                rr.append(RR.parse(buffer))
+            for i in range(header.auth):
+                auth.append(RR.parse(buffer))
+            for i in range(header.ar):
+                ar.append(RR.parse(buffer))
+            return cls(header, questions, rr, auth=auth, ar=ar)
+        except (BufferError, BimapError) as e:
+            raise DNSError(
+                "Error unpacking DNSRecord [offset=%d]: %s" % (buffer.offset, e)
+            )
+
+    @classmethod
+    def question(cls, qname, qtype="A", qclass="IN"):
+        return DNSRecord(
+            q=DNSQuestion(qname, getattr(QTYPE, qtype), getattr(CLASS, qclass))
+        )
+
+    def __init__(
+        self, header=None, questions=None, rr=None, q=None, a=None, auth=None, ar=None
+    ) -> None:
+        self.header = header or DNSHeader()
+        self.questions: list[DNSQuestion] = questions or []
+        self.rr: list[RR] = rr or []
+        self.auth: list[RR] = auth or []
+        self.ar: list[RR] = ar or []
+
+        if q:
+            self.questions.append(q)
+        if a:
+            self.rr.append(a)
+        self.set_header_qa()
+
+    def reply(self, ra=1, aa=1):
+        return DNSRecord(
+            DNSHeader(id=self.header.id, bitmap=self.header.bitmap, qr=1, ra=ra, aa=aa),
+            q=self.q,
+        )
+
+    def add_question(self, *q) -> None:
+        self.questions.extend(q)
+        self.set_header_qa()
+
+    def add_answer(self, *rr) -> None:
+        self.rr.extend(rr)
+        self.set_header_qa()
+
+    def add_auth(self, *auth) -> None:
+        self.auth.extend(auth)
+        self.set_header_qa()
+
+    def add_ar(self, *ar) -> None:
+        self.ar.extend(ar)
+        self.set_header_qa()
+
+    def set_header_qa(self) -> None:
+        self.header.q = len(self.questions)
+        self.header.a = len(self.rr)
+        self.header.auth = len(self.auth)
+        self.header.ar = len(self.ar)
+
+    def get_q(self):
+        return self.questions[0] if self.questions else DNSQuestion()
+
+    q = property(get_q)
+
+    def get_a(self):
+        return self.rr[0] if self.rr else RR()
+
+    a = property(get_a)
+
+    def pack(self) -> bytes:
+        self.set_header_qa()
+        buffer = DNSBuffer()
+        self.header.pack(buffer)
+        for q in self.questions:
+            q.pack(buffer)
+        for rr in self.rr:
+            rr.pack(buffer)
+        for auth in self.auth:
+            auth.pack(buffer)
+        for ar in self.ar:
+            ar.pack(buffer)
+        return buffer.data
+
+    def truncate(self):
+        return DNSRecord(DNSHeader(id=self.header.id, bitmap=self.header.bitmap, tc=1))
+
+    def format(self, prefix="", sort=False):
+        s = sorted if sort else lambda x: x
+        sections = [repr(self.header)]
+        sections.extend(s([repr(q) for q in self.questions]))
+        sections.extend(s([repr(rr) for rr in self.rr]))
+        sections.extend(s([repr(rr) for rr in self.auth]))
+        sections.extend(s([repr(rr) for rr in self.ar]))
+        return prefix + ("\n" + prefix).join(sections)
+
+    short = format
+
+    def __repr__(self):
+        return self.format()
+
+    __str__ = __repr__
+
+
+class DNSHeader(object):
+    id = H("id")
+    bitmap = H("bitmap")
+    q = H("q")
+    a = H("a")
+    auth = H("auth")
+    ar = H("ar")
+
+    @classmethod
+    def parse(cls, buffer):
+        try:
+            (id, bitmap, q, a, auth, ar) = buffer.unpack("!HHHHHH")
+            return cls(id, bitmap, q, a, auth, ar)
+        except (BufferError, BimapError) as e:
+            raise DNSError(
+                "Error unpacking DNSHeader [offset=%d]: %s" % (buffer.offset, e)
+            )
+
+    def __init__(self, id=None, bitmap=None, q=0, a=0, auth=0, ar=0, **args) -> None:
+        self.id = id if id else 0
+        if bitmap is None:
+            self.bitmap = 0
+        else:
+            self.bitmap = bitmap
+        self.q = q
+        self.a = a
+        self.auth = auth
+        self.ar = ar
+        for k, v in args.items():
+            if k.lower() == "qr":
+                self.qr = v
+            elif k.lower() == "opcode":
+                self.opcode = v
+            elif k.lower() == "aa":
+                self.aa = v
+            elif k.lower() == "tc":
+                self.tc = v
+            elif k.lower() == "rd":
+                self.rd = v
+            elif k.lower() == "ra":
+                self.ra = v
+            elif k.lower() == "z":
+                self.z = v
+            elif k.lower() == "ad":
+                self.ad = v
+            elif k.lower() == "cd":
+                self.cd = v
+            elif k.lower() == "rcode":
+                self.rcode = v
+
+    def get_qr(self):
+        return get_bits(self.bitmap, 15)
+
+    def set_qr(self, val):
+        self.bitmap = set_bits(self.bitmap, val, 15)
+
+    qr = property(get_qr, set_qr)
+
+    def get_opcode(self):
+        return get_bits(self.bitmap, 11, 4)
+
+    def set_opcode(self, val):
+        self.bitmap = set_bits(self.bitmap, val, 11, 4)
+
+    opcode = property(get_opcode, set_opcode)
+
+    def get_aa(self):
+        return get_bits(self.bitmap, 10)
+
+    def set_aa(self, val):
+        self.bitmap = set_bits(self.bitmap, val, 10)
+
+    aa = property(get_aa, set_aa)
+
+    def get_tc(self):
+        return get_bits(self.bitmap, 9)
+
+    def set_tc(self, val):
+        self.bitmap = set_bits(self.bitmap, val, 9)
+
+    tc = property(get_tc, set_tc)
+
+    def get_rd(self):
+        return get_bits(self.bitmap, 8)
+
+    def set_rd(self, val):
+        self.bitmap = set_bits(self.bitmap, val, 8)
+
+    rd = property(get_rd, set_rd)
+
+    def get_ra(self):
+        return get_bits(self.bitmap, 7)
+
+    def set_ra(self, val):
+        self.bitmap = set_bits(self.bitmap, val, 7)
+
+    ra = property(get_ra, set_ra)
+
+    def get_z(self):
+        return get_bits(self.bitmap, 6)
+
+    def set_z(self, val):
+        self.bitmap = set_bits(self.bitmap, val, 6)
+
+    z = property(get_z, set_z)
+
+    def get_ad(self):
+        return get_bits(self.bitmap, 5)
+
+    def set_ad(self, val):
+        self.bitmap = set_bits(self.bitmap, val, 5)
+
+    ad = property(get_ad, set_ad)
+
+    def get_cd(self):
+        return get_bits(self.bitmap, 4)
+
+    def set_cd(self, val):
+        self.bitmap = set_bits(self.bitmap, val, 4)
+
+    cd = property(get_cd, set_cd)
+
+    def get_rcode(self):
+        return get_bits(self.bitmap, 0, 4)
+
+    def set_rcode(self, val):
+        self.bitmap = set_bits(self.bitmap, val, 0, 4)
+
+    rcode = property(get_rcode, set_rcode)
+
+    def pack(self, buffer):
+        buffer.pack("!HHHHHH", self.id, self.bitmap, self.q, self.a, self.auth, self.ar)
+
+    def __repr__(self):
+        f = [
+            self.aa and "AA",
+            self.tc and "TC",
+            self.rd and "RD",
+            self.ra and "RA",
+            self.z and "Z",
+            self.ad and "AD",
+            self.cd and "CD",
+        ]
+        if OPCODE.get(self.opcode) == "UPDATE":
+            f1 = "zo"
+            f2 = "pr"
+            f3 = "up"
+            f4 = "ad"
+        else:
+            f1 = "q"
+            f2 = "a"
+            f3 = "ns"
+            f4 = "ar"
+        return (
+            "<DNS Header: id=0x%x type=%s opcode=%s flags=%s "
+            "rcode='%s' %s=%d %s=%d %s=%d %s=%d>"
+            % (
+                self.id,
+                QR.get(self.qr),
+                OPCODE.get(self.opcode),
+                ",".join(filter(None, f)),
+                RCODE.get(self.rcode),
+                f1,
+                self.q,
+                f2,
+                self.a,
+                f3,
+                self.auth,
+                f4,
+                self.ar,
+            )
+        )
+
+    __str__ = __repr__
+
+
+class DNSQuestion(object):
+    @classmethod
+    def parse(cls, buffer):
+        try:
+            qname = buffer.decode_name()
+            qtype, qclass = buffer.unpack("!HH")
+            return cls(qname, qtype, qclass)
+        except (BufferError, BimapError) as e:
+            raise DNSError(
+                "Error unpacking DNSQuestion [offset=%d]: %s" % (buffer.offset, e)
+            )
+
+    def __init__(self, qname=None, qtype=1, qclass=1) -> None:
+        self.qname = qname
+        self.qtype = qtype
+        self.qclass = qclass
+
+    def set_qname(self, qname):
+        if isinstance(qname, DNSLabel):
+            self._qname = qname
+        else:
+            self._qname = DNSLabel(qname)
+
+    def get_qname(self):
+        return self._qname
+
+    qname = property(get_qname, set_qname)
+
+    def pack(self, buffer):
+        buffer.encode_name(self.qname)
+        buffer.pack("!HH", self.qtype, self.qclass)
+
+    def __repr__(self):
+        return "<DNS Question: '%s' qtype=%s qclass=%s>" % (
+            self.qname,
+            QTYPE.get(self.qtype),
+            CLASS.get(self.qclass),
+        )
+
+    __str__ = __repr__
+
+
+class RR(object):
+    rtype = H("rtype")
+    rclass = H("rclass")
+    ttl = I("ttl")
+    rdlength = H("rdlength")
+
+    @classmethod
+    def parse(cls, buffer):
+        try:
+            rname = buffer.decode_name()
+            rtype, rclass, ttl, rdlength = buffer.unpack("!HHIH")
+            if rdlength:
+                rdata = RDMAP.get(QTYPE.get(rtype), RD).parse(buffer, rdlength)
+            else:
+                rdata = ""
+            return cls(rname, rtype, rclass, ttl, rdata)
+        except (BufferError, BimapError) as e:
+            raise DNSError("Error unpacking RR [offset=%d]: %s" % (buffer.offset, e))
+
+    def __init__(self, rname=None, rtype=1, rclass=1, ttl=0, rdata=None) -> None:
+        self.rname = rname
+        self.rtype = rtype
+        self.rclass = rclass
+        self.ttl = ttl
+        self.rdata = rdata
+
+    def set_rname(self, rname):
+        if isinstance(rname, DNSLabel):
+            self._rname = rname
+        else:
+            self._rname = DNSLabel(rname)
+
+    def get_rname(self):
+        return self._rname
+
+    rname = property(get_rname, set_rname)
+
+    def pack(self, buffer):
+        buffer.encode_name(self.rname)
+        buffer.pack("!HHI", self.rtype, self.rclass, self.ttl)
+        rdlength_ptr = buffer.offset
+        buffer.pack("!H", 0)
+        start = buffer.offset
+        self.rdata.pack(buffer)
+        end = buffer.offset
+        buffer.update(rdlength_ptr, "!H", end - start)
+
+    def __repr__(self):
+        return "<DNS RR: '%s' rtype=%s rclass=%s ttl=%d rdata='%s'>" % (
+            self.rname,
+            QTYPE.get(self.rtype),
+            CLASS.get(self.rclass),
+            self.ttl,
+            self.rdata,
+        )
+
+    __str__ = __repr__
+
+
+class RD(object):
+    @classmethod
+    def parse(cls, buffer, length):
+        try:
+            data = buffer.get(length)
+            return cls(data)
+        except (BufferError, BimapError) as e:
+            raise DNSError("Error unpacking RD [offset=%d]: %s" % (buffer.offset, e))
+
+    def __init__(self, data=b"") -> None:
+        check_bytes("data", data)
+        self.data = bytes(data)
+
+    def pack(self, buffer):
+        buffer.append(self.data)
+
+    def __repr__(self):
+        if len(self.data) > 0:
+            return "\\# %d %s" % (
+                len(self.data),
+                binascii.hexlify(self.data).decode().upper(),
+            )
+        else:
+            return "\\# 0"
+
+    attrs = ("data",)
+
+
+def _force_bytes(x):
+    if isinstance(x, bytes):
+        return x
+    else:
+        return x.encode()
+
+
+class TXT(RD):
+    @classmethod
+    def parse(cls, buffer, length):
+        try:
+            data = list()
+            start_bo = buffer.offset
+            now_length = 0
+            while buffer.offset < start_bo + length:
+                (txtlength,) = buffer.unpack("!B")
+
+                if now_length + txtlength < length:
+                    now_length += txtlength
+                    data.append(buffer.get(txtlength))
+                else:
+                    raise DNSError(
+                        "Invalid TXT record: len(%d) > RD len(%d)" % (txtlength, length)
+                    )
+            return cls(data)
+        except (BufferError, BimapError) as e:
+            raise DNSError("Error unpacking TXT [offset=%d]: %s" % (buffer.offset, e))
+
+    def __init__(self, data) -> None:
+        if type(data) in (tuple, list):
+            self.data = [_force_bytes(x) for x in data]
+        else:
+            self.data = [_force_bytes(data)]
+        if any([len(x) > 255 for x in self.data]):
+            raise DNSError("TXT record too long: %s" % self.data)
+
+    def pack(self, buffer):
+        for ditem in self.data:
+            if len(ditem) > 255:
+                raise DNSError("TXT record too long: %s" % ditem)
+            buffer.pack("!B", len(ditem))
+            buffer.append(ditem)
+
+    def __repr__(self):
+        return ",".join([repr(x) for x in self.data])
+
+
+class A(RD):
+
+    data = IP4("data")
+
+    @classmethod
+    def parse(cls, buffer, length):
+        try:
+            data = buffer.unpack("!BBBB")
+            return cls(data)
+        except (BufferError, BimapError) as e:
+            raise DNSError("Error unpacking A [offset=%d]: %s" % (buffer.offset, e))
+
+    def __init__(self, data) -> None:
+        if type(data) in (tuple, list):
+            self.data = tuple(data)
+        else:
+            self.data = tuple(map(int, data.rstrip(".").split(".")))
+
+    def pack(self, buffer):
+        buffer.pack("!BBBB", *self.data)
+
+    def __repr__(self):
+        return "%d.%d.%d.%d" % self.data
+
+
+def _parse_ipv6(a):
+    l, _, r = a.partition("::")
+    l_groups = list(chain(*[divmod(int(x, 16), 256) for x in l.split(":") if x]))
+    r_groups = list(chain(*[divmod(int(x, 16), 256) for x in r.split(":") if x]))
+    zeros = [0] * (16 - len(l_groups) - len(r_groups))
+    return tuple(l_groups + zeros + r_groups)
+
+
+def _format_ipv6(a):
+    left = []
+    right = []
+    current = "left"
+    for i in range(0, 16, 2):
+        group = (a[i] << 8) + a[i + 1]
+        if current == "left":
+            if group == 0 and i < 14:
+                if (a[i + 2] << 8) + a[i + 3] == 0:
+                    current = "right"
+                else:
+                    left.append("0")
+            else:
+                left.append("%x" % group)
+        else:
+            if group == 0 and len(right) == 0:
+                pass
+            else:
+                right.append("%x" % group)
+    if len(left) < 8:
+        return ":".join(left) + "::" + ":".join(right)
+    else:
+        return ":".join(left)
+
+
+class AAAA(RD):
+    data = IP6("data")
+
+    @classmethod
+    def parse(cls, buffer, length):
+        try:
+            data = buffer.unpack("!16B")
+            return cls(data)
+        except (BufferError, BimapError) as e:
+            raise DNSError("Error unpacking AAAA [offset=%d]: %s" % (buffer.offset, e))
+
+    def __init__(self, data) -> None:
+        if type(data) in (tuple, list):
+            self.data = tuple(data)
+        else:
+            self.data = _parse_ipv6(data)
+
+    def pack(self, buffer):
+        buffer.pack("!16B", *self.data)
+
+    def __repr__(self):
+        return _format_ipv6(self.data)
+
+
+class CNAME(RD):
+    @classmethod
+    def parse(cls, buffer, length):
+        try:
+            label = buffer.decode_name()
+            return cls(label)
+        except (BufferError, BimapError) as e:
+            raise DNSError("Error unpacking CNAME [offset=%d]: %s" % (buffer.offset, e))
+
+    def __init__(self, label=None) -> None:
+        self.label = label
+
+    def set_label(self, label):
+        if isinstance(label, DNSLabel):
+            self._label = label
+        else:
+            self._label = DNSLabel(label)
+
+    def get_label(self):
+        return self._label
+
+    label = property(get_label, set_label)
+
+    def pack(self, buffer):
+        buffer.encode_name(self.label)
+
+    def __repr__(self):
+        return "%s" % (self.label)
+
+    attrs = ("label",)
+
+
+class PTR(CNAME):
+    pass
+
+
+class SRV(RD):
+    priority = H("priority")
+    weight = H("weight")
+    port = H("port")
+
+    @classmethod
+    def parse(cls, buffer, length):
+        try:
+            priority, weight, port = buffer.unpack("!HHH")
+            target = buffer.decode_name()
+            return cls(priority, weight, port, target)
+        except (BufferError, BimapError) as e:
+            raise DNSError("Error unpacking SRV [offset=%d]: %s" % (buffer.offset, e))
+
+    def __init__(self, priority=0, weight=0, port=0, target=None) -> None:
+        self.priority = priority
+        self.weight = weight
+        self.port = port
+        self.target = target
+
+    def set_target(self, target):
+        if isinstance(target, DNSLabel):
+            self._target = target
+        else:
+            self._target = DNSLabel(target)
+
+    def get_target(self):
+        return self._target
+
+    target = property(get_target, set_target)
+
+    def pack(self, buffer):
+        buffer.pack("!HHH", self.priority, self.weight, self.port)
+        buffer.encode_name(self.target)
+
+    def __repr__(self):
+        return "%d %d %d %s" % (self.priority, self.weight, self.port, self.target)
+
+    attrs = ("priority", "weight", "port", "target")
+
+
+def decode_type_bitmap(type_bitmap):
+    rrlist = []
+    buf = DNSBuffer(type_bitmap)
+    while buf.remaining():
+        winnum, winlen = buf.unpack("BB")
+        bitmap = bytearray(buf.get(winlen))
+        for (pos, value) in enumerate(bitmap):
+            for i in range(8):
+                if (value << i) & 0x80:
+                    bitpos = (256 * winnum) + (8 * pos) + i
+                    rrlist.append(QTYPE[bitpos])
+    return rrlist
+
+
+def encode_type_bitmap(rrlist):
+    rrlist = sorted([getattr(QTYPE, rr) for rr in rrlist])
+    buf = DNSBuffer()
+    curWindow = rrlist[0] // 256
+    bitmap = bytearray(32)
+    n = len(rrlist) - 1
+    for i, rr in enumerate(rrlist):
+        v = rr - curWindow * 256
+        bitmap[v // 8] |= 1 << (7 - v % 8)
+
+        if i == n or rrlist[i + 1] >= (curWindow + 1) * 256:
+            while bitmap[-1] == 0:
+                bitmap = bitmap[:-1]
+            buf.pack("BB", curWindow, len(bitmap))
+            buf.append(bitmap)
+
+            if i != n:
+                curWindow = rrlist[i + 1] // 256
+                bitmap = bytearray(32)
+
+    return buf.data
+
+
+class NSEC(RD):
+    @classmethod
+    def parse(cls, buffer, length):
+        try:
+            end = buffer.offset + length
+            name = buffer.decode_name()
+            rrlist = decode_type_bitmap(buffer.get(end - buffer.offset))
+            return cls(name, rrlist)
+        except (BufferError, BimapError) as e:
+            raise DNSError("Error unpacking NSEC [offset=%d]: %s" % (buffer.offset, e))
+
+    def __init__(self, label, rrlist) -> None:
+        self.label = label
+        self.rrlist = rrlist
+
+    def set_label(self, label):
+        if isinstance(label, DNSLabel):
+            self._label = label
+        else:
+            self._label = DNSLabel(label)
+
+    def get_label(self):
+        return self._label
+
+    label = property(get_label, set_label)
+
+    def pack(self, buffer):
+        buffer.encode_name(self.label)
+        buffer.append(encode_type_bitmap(self.rrlist))
+
+    def __repr__(self):
+        return "%s %s" % (self.label, " ".join(self.rrlist))
+
+    attrs = ("label", "rrlist")
+
+
+RDMAP = {"A": A, "AAAA": AAAA, "TXT": TXT, "PTR": PTR, "SRV": SRV, "NSEC": NSEC}

copyparty/stolen/dnslib/label.py

@@ -0,0 +1,154 @@
+# coding: utf-8
+
+from __future__ import print_function
+
+import re
+
+from .bit import get_bits, set_bits
+from .buffer import Buffer, BufferError
+
+LDH = set(range(33, 127))
+ESCAPE = re.compile(r"\\([0-9][0-9][0-9])")
+
+
+class DNSLabelError(Exception):
+    pass
+
+
+class DNSLabel(object):
+    def __init__(self, label):
+        if type(label) == DNSLabel:
+            self.label = label.label
+        elif type(label) in (list, tuple):
+            self.label = tuple(label)
+        else:
+            if not label or label in (b".", "."):
+                self.label = ()
+            elif type(label) is not bytes:
+                if type("") != type(b""):
+
+                    label = ESCAPE.sub(lambda m: chr(int(m[1])), label)
+                self.label = tuple(label.encode("idna").rstrip(b".").split(b"."))
+            else:
+                if type("") == type(b""):
+
+                    label = ESCAPE.sub(lambda m: chr(int(m.groups()[0])), label)
+                self.label = tuple(label.rstrip(b".").split(b"."))
+
+    def add(self, name):
+        new = DNSLabel(name)
+        if self.label:
+            new.label += self.label
+        return new
+
+    def idna(self):
+        return ".".join([s.decode("idna") for s in self.label]) + "."
+
+    def _decode(self, s):
+        if set(s).issubset(LDH):
+
+            return s.decode()
+        else:
+
+            return "".join([(chr(c) if (c in LDH) else "\\%03d" % c) for c in s])
+
+    def __str__(self):
+        return ".".join([self._decode(bytearray(s)) for s in self.label]) + "."
+
+    def __repr__(self):
+        return "<DNSLabel: '%s'>" % str(self)
+
+    def __hash__(self):
+        return hash(tuple(map(lambda x: x.lower(), self.label)))
+
+    def __ne__(self, other):
+        return not self == other
+
+    def __eq__(self, other):
+        if type(other) != DNSLabel:
+            return self.__eq__(DNSLabel(other))
+        else:
+            return [l.lower() for l in self.label] == [l.lower() for l in other.label]
+
+    def __len__(self):
+        return len(b".".join(self.label))
+
+
+class DNSBuffer(Buffer):
+    def __init__(self, data=b""):
+        super(DNSBuffer, self).__init__(data)
+        self.names = {}
+
+    def decode_name(self, last=-1):
+        label = []
+        done = False
+        while not done:
+            (length,) = self.unpack("!B")
+            if get_bits(length, 6, 2) == 3:
+
+                self.offset -= 1
+                pointer = get_bits(self.unpack("!H")[0], 0, 14)
+                save = self.offset
+                if last == save:
+                    raise BufferError(
+                        "Recursive pointer in DNSLabel [offset=%d,pointer=%d,length=%d]"
+                        % (self.offset, pointer, len(self.data))
+                    )
+                if pointer < self.offset:
+                    self.offset = pointer
+                else:
+
+                    raise BufferError(
+                        "Invalid pointer in DNSLabel [offset=%d,pointer=%d,length=%d]"
+                        % (self.offset, pointer, len(self.data))
+                    )
+                label.extend(self.decode_name(save).label)
+                self.offset = save
+                done = True
+            else:
+                if length > 0:
+                    l = self.get(length)
+                    try:
+                        l.decode()
+                    except UnicodeDecodeError:
+                        raise BufferError("Invalid label <%s>" % l)
+                    label.append(l)
+                else:
+                    done = True
+        return DNSLabel(label)
+
+    def encode_name(self, name):
+        if not isinstance(name, DNSLabel):
+            name = DNSLabel(name)
+        if len(name) > 253:
+            raise DNSLabelError("Domain label too long: %r" % name)
+        name = list(name.label)
+        while name:
+            if tuple(name) in self.names:
+
+                pointer = self.names[tuple(name)]
+                pointer = set_bits(pointer, 3, 14, 2)
+                self.pack("!H", pointer)
+                return
+            else:
+                self.names[tuple(name)] = self.offset
+                element = name.pop(0)
+                if len(element) > 63:
+                    raise DNSLabelError("Label component too long: %r" % element)
+                self.pack("!B", len(element))
+                self.append(element)
+        self.append(b"\x00")
+
+    def encode_name_nocompress(self, name):
+        if not isinstance(name, DNSLabel):
+            name = DNSLabel(name)
+        if len(name) > 253:
+            raise DNSLabelError("Domain label too long: %r" % name)
+        name = list(name.label)
+        while name:
+            element = name.pop(0)
+            if len(element) > 63:
+                raise DNSLabelError("Label component too long: %r" % element)
+            self.pack("!B", len(element))
+            self.append(element)
+        self.append(b"\x00")

copyparty/stolen/dnslib/lex.py

@@ -0,0 +1,105 @@
+# coding: utf-8
+
+from __future__ import print_function
+
+import collections
+
+try:
+    from StringIO import StringIO
+except ImportError:
+    from io import StringIO
+
+
+class Lexer(object):
+
+    escape_chars = "\\"
+    escape = {"n": "\n", "t": "\t", "r": "\r"}
+
+    def __init__(self, f, debug=False):
+        if hasattr(f, "read"):
+            self.f = f
+        elif type(f) == str:
+            self.f = StringIO(f)
+        elif type(f) == bytes:
+            self.f = StringIO(f.decode())
+        else:
+            raise ValueError("Invalid input")
+        self.debug = debug
+        self.q = collections.deque()
+        self.state = self.lexStart
+        self.escaped = False
+        self.eof = False
+
+    def __iter__(self):
+        return self.parse()
+
+    def next_token(self):
+        if self.debug:
+            print("STATE", self.state)
+        (tok, self.state) = self.state()
+        return tok
+
+    def parse(self):
+        while self.state is not None and not self.eof:
+            tok = self.next_token()
+            if tok:
+                yield tok
+
+    def read(self, n=1):
+        s = ""
+        while self.q and n > 0:
+            s += self.q.popleft()
+            n -= 1
+        s += self.f.read(n)
+        if s == "":
+            self.eof = True
+        if self.debug:
+            print("Read: >%s<" % repr(s))
+        return s
+
+    def peek(self, n=1):
+        s = ""
+        i = 0
+        while len(self.q) > i and n > 0:
+            s += self.q[i]
+            i += 1
+            n -= 1
+        r = self.f.read(n)
+        if n > 0 and r == "":
+            self.eof = True
+        self.q.extend(r)
+        if self.debug:
+            print("Peek : >%s<" % repr(s + r))
+        return s + r
+
+    def pushback(self, s):
+        p = collections.deque(s)
+        p.extend(self.q)
+        self.q = p
+
+    def readescaped(self):
+        c = self.read(1)
+        if c in self.escape_chars:
+            self.escaped = True
+            n = self.peek(3)
+            if n.isdigit():
+                n = self.read(3)
+                if self.debug:
+                    print("Escape: >%s<" % n)
+                return chr(int(n, 8))
+            elif n[0] in "x":
+                x = self.read(3)
+                if self.debug:
+                    print("Escape: >%s<" % x)
+                return chr(int(x[1:], 16))
+            else:
+                c = self.read(1)
+                if self.debug:
+                    print("Escape: >%s<" % c)
+                return self.escape.get(c, c)
+        else:
+            self.escaped = False
+            return c
+
+    def lexStart(self):
+        return (None, None)