v0.11.11

.vscode/launch.py

@@ -3,14 +3,12 @@
 # launches 10x faster than mspython debugpy
 # and is stoppable with ^C
 
+import re
 import os
 import sys
 import shlex
-
-sys.path.insert(0, os.getcwd())
-
 import jstyleson
-from copyparty.__main__ import main as copyparty
+import subprocess as sp
 
 with open(".vscode/launch.json", "r", encoding="utf-8") as f:
     tj = f.read()
@@ -25,11 +23,19 @@ except:
     pass
 
 argv = [os.path.expanduser(x) if x.startswith("~") else x for x in argv]
-try:
-    copyparty(["a"] + argv)
-except SystemExit as ex:
-    if ex.code:
-        raise
+
+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)

README.md

@@ -68,12 +68,16 @@ you may also want these, especially on servers:
 
 ## notes
 
+general:
+* paper-printing is affected by dark/light-mode! use lightmode for color, darkmode for grayscale
+  * because no browsers currently implement the media-query to do this properly orz
+
+browser-specific:
 * iPhone/iPad: use Firefox to download files
 * Android-Chrome: increase "parallel uploads" for higher speed (android bug)
 * Android-Firefox: takes a while to select files (their fix for ☝️)
 * Desktop-Firefox: ~~may use gigabytes of RAM if your files are massive~~ *seems to be OK now*
-* paper-printing is affected by dark/light-mode! use lightmode for color, darkmode for grayscale
-  * because no browsers currently implement the media-query to do this properly orz
+* Desktop-Firefox: may stop you from deleting folders you've uploaded until you visit `about:memory` and click `Minimize memory usage`
 
 
 ## status
@@ -397,6 +401,8 @@ quick summary of more eccentric web-browsers trying to view a directory index:
   * cross-platform python client available in [./bin/](bin/)
   * [rclone](https://rclone.org/) as client can give ~5x performance, see [./docs/rclone.md](docs/rclone.md)
 
+* sharex (screenshot utility): see [./contrib/sharex.sxcu](contrib/#sharexsxcu)
+
 copyparty returns a truncated sha512sum of your PUT/POST as base64; you can generate the same checksum locally to verify uplaods:
 
     b512(){ printf "$((sha512sum||shasum -a512)|sed -E 's/ .*//;s/(..)/\\x\1/g')"|base64|head -c43;}
@@ -520,20 +526,25 @@ in the `scripts` folder:
 
 roughly sorted by priority
 
-* separate sqlite table per tag
-* audio fingerprinting
 * readme.md as epilogue
+* single sha512 across all up2k chunks? maybe
 * reduce up2k roundtrips
   * start from a chunk index and just go
   * terminate client on bad data
-* `os.copy_file_range` for up2k cloning
-* single sha512 across all up2k chunks? maybe
-* figure out the deal with pixel3a not being connectable as hotspot
-  * pixel3a having unpredictable 3sec latency in general :||||
 
 discarded ideas
 
+* separate sqlite table per tag
+  * performance fixed by skipping some indexes (`+mt.k`)
+* audio fingerprinting
+  * only makes sense if there can be a wasm client and that doesn't exist yet (except for olaf which is agpl hence counts as not existing)
+* `os.copy_file_range` for up2k cloning
+  * almost never hit this path anyways
 * up2k partials ui
+  * feels like there isn't much point
 * cache sha512 chunks on client
+  * too dangerous
 * comment field
+  * nah
 * look into android thumbnail cache file format
+  * absolutely not

bin/README.md

@@ -45,3 +45,18 @@ you could replace winfsp with [dokan](https://github.com/dokan-dev/dokany/releas
 # [`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 recommends to wipe the DB and reindex because it now collects additional metadata during analysis, but you have some really expensive `-mtp` parsers and want to copy over the tags from the old db
+
+for that example (upgrading to v0.11.0), first move the old db aside, launch copyparty, let it rebuild the db until the point where it starts running mtp (colored messages as it adds the mtp tags), then 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`:
+
+```
+~/bin/dbtool.py -ls up2k.db
+~/bin/dbtool.py -src up2k.db.v0.10.22 up2k.db -cmp
+~/bin/dbtool.py -src up2k.db.v0.10.22 up2k.db -rm-mtp-flag -copy key
+~/bin/dbtool.py -src up2k.db.v0.10.22 up2k.db -rm-mtp-flag -copy .bpm -vac
+```

bin/copyparty-fuse.py

@@ -54,6 +54,12 @@ 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:
     from fuse import FUSE, FuseOSError, Operations
 except:

bin/dbtool.py

@@ -0,0 +1,198 @@
+#!/usr/bin/env python3
+
+import os
+import sys
+import sqlite3
+import argparse
+
+DB_VER = 3
+
+
+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 w, 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)
+
+        q = "select w from up where substr(w,1,16) = ?"
+        hit = d2.execute(q, (w[:16],)).fetchone()
+        if not hit:
+            miss += 1
+            if verbose:
+                print(f"file in {n1} missing in {n2}: [{w}] {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 w, 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)
+
+        v2 = d2.execute("select v from mt where w = ? and +k = ?", (w, 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:
+                q = "select rd, fn from up where substr(w,1,16) = ?"
+                rd, fn = d1.execute(q, (w,)).fetchone()
+                print(f"missing in {n2}: [{w}] [{rd}/{fn}] {k} = {v}")
+
+    for k, v in sorted(miss.items()):
+        if v:
+            print(f"{n1} has {v:6} more {k:<6} 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
+    ndone = 0
+    for w, 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 {ndone} copied\033[0m"
+            print(m)
+
+        hit = d2.execute("select v from mt where w = ? and +k = ?", (w, k)).fetchone()
+        if hit:
+            hit = hit[0]
+
+        if hit != v:
+            ndone += 1
+            if hit is not None:
+                d2.execute("delete from mt where w = ? and +k = ?", (w, k))
+
+            d2.execute("insert into mt values (?,?,?)", (w, k, v))
+            if rm:
+                d2.execute("delete from mt where w = ? and +k = 't:mtp'", (w,))
+
+    d2.commit()
+    print(f"copied {ndone} {tag} tags over")
+
+
+def main():
+    os.system("")
+    print()
+    
+    ap = argparse.ArgumentParser()
+    ap.add_argument("db", help="database to work on")
+    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 as done, so copyparty won't run mtp on it",
+    )
+    ap2.add_argument("-vac", action="store_true", help="optimize DB")
+
+    ar = ap.parse_args()
+
+    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
+
+    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_VER:
+            m = f"{n} db is version {ver}, this tool only supports version {DB_VER}, 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()

contrib/README.md

@@ -9,6 +9,16 @@
 * 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)
+
 ### [`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 copyparty-fuse))

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": "$json:files[0].url$"
+}

copyparty/__main__.py

@@ -225,6 +225,20 @@ def run_argparse(argv, formatter):
             --ciphers help = available ssl/tls ciphers,
             --ssl-ver help = available ssl/tls versions,
               default is what python considers safe, usually >= TLS1
+            
+            values for --ls:
+              "USR" is a user to browse as; * is anonymous, ** is all users
+              "VOL" is a single volume to scan, default is * (all vols)
+              "FLAG" is flags;
+                "v" in addition to realpaths, print usernames and vpaths
+                "ln" only prints symlinks leaving the volume mountpoint
+                "p" exits 1 if any such symlinks are found
+                "r" resumes startup after the listing
+            examples:
+              --ls '**'          # list all files which are possible to read
+              --ls '**,*,ln'     # check for dangerous symlinks
+              --ls '**,*,ln,p,r' # check, then start normally if safe
+            \033[0m
             """
         ),
     )
@@ -261,7 +275,7 @@ def run_argparse(argv, formatter):
     ap2.add_argument("--th-no-jpg", action="store_true", help="disable jpg output")
     ap2.add_argument("--th-no-webp", action="store_true", help="disable webp output")
     ap2.add_argument("--th-poke", metavar="SEC", type=int, default=300, help="activity labeling cooldown")
-    ap2.add_argument("--th-clean", metavar="SEC", type=int, default=1800, help="cleanup interval")
+    ap2.add_argument("--th-clean", metavar="SEC", type=int, default=43200, help="cleanup interval")
     ap2.add_argument("--th-maxage", metavar="SEC", type=int, default=604800, help="max folder age")
 
     ap2 = ap.add_argument_group('database options')
@@ -288,6 +302,7 @@ def run_argparse(argv, formatter):
     ap2.add_argument("--ssl-log", metavar="PATH", help="log master secrets")
 
     ap2 = ap.add_argument_group('debug options')
+    ap2.add_argument("--ls", metavar="U[,V[,F]]", help="scan all volumes")
     ap2.add_argument("--log-conn", action="store_true", help="print tcp-server msgs")
     ap2.add_argument("--no-sendfile", action="store_true", help="disable sendfile")
     ap2.add_argument("--no-scandir", action="store_true", help="disable scandir")
@@ -363,6 +378,9 @@ def main(argv=None):
             + "  (if you crash with codec errors then that is why)"
         )
 
+    if WINDOWS and sys.version_info < (3, 6):
+        al.no_scandir = True
+
     # signal.signal(signal.SIGINT, sighandler)
 
     SvcHub(al).run()

copyparty/__version__.py

@@ -1,8 +1,8 @@
 # coding: utf-8
 
-VERSION = (0, 11, 1)
+VERSION = (0, 11, 11)
 CODENAME = "the grid"
-BUILD_DT = (2021, 5, 29)
+BUILD_DT = (2021, 6, 8)
 
 S_VERSION = ".".join(map(str, VERSION))
 S_BUILD_DT = "{0:04d}-{1:02d}-{2:02d}".format(*BUILD_DT)

copyparty/authsrv.py

@@ -7,7 +7,7 @@ import sys
 import stat
 import threading
 
-from .__init__ import PY2, WINDOWS
+from .__init__ import WINDOWS
 from .util import IMPLICATIONS, undot, Pebkac, fsdec, fsenc, statdir, nuprint
 
 
@@ -22,7 +22,7 @@ class VFS(object):
         self.uadm = uadm  # users who are regular admins
         self.flags = flags  # config switches
         self.nodes = {}  # child nodes
-        self.all_vols = {vpath: self}  # flattened recursive
+        self.all_vols = {vpath: self} if realpath else {}  # flattened recursive
 
     def __repr__(self):
         return "VFS({})".format(
@@ -96,6 +96,7 @@ class VFS(object):
         ]
 
     def get(self, vpath, uname, will_read, will_write):
+        # type: (str, str, bool, bool) -> tuple[VFS, str]
         """returns [vfsnode,fs_remainder] if user has the requested permissions"""
         vn, rem = self._find(vpath)
 
@@ -135,7 +136,8 @@ class VFS(object):
             #
             return os.path.realpath(rp)
 
-    def ls(self, rem, uname, scandir, lstat=False):
+    def ls(self, rem, uname, scandir, incl_wo=False, lstat=False):
+        # type: (str, str, bool, bool, bool) -> tuple[str, str, dict[str, VFS]]
         """return user-readable [fsdir,real,virt] items at vpath"""
         virt_vis = {}  # nodes readable by user
         abspath = self.canonical(rem)
@@ -143,12 +145,12 @@ class VFS(object):
         real.sort()
         if not rem:
             for name, vn2 in sorted(self.nodes.items()):
-                if (
-                    uname in vn2.uread
-                    or "*" in vn2.uread
-                    or uname in vn2.uwrite
-                    or "*" in vn2.uwrite
-                ):
+                ok = uname in vn2.uread or "*" in vn2.uread
+
+                if not ok and incl_wo:
+                    ok = uname in vn2.uwrite or "*" in vn2.uwrite
+
+                if ok:
                     virt_vis[name] = vn2
 
             # no vfs nodes in the list of real inodes
@@ -156,13 +158,21 @@ class VFS(object):
 
         return [abspath, real, virt_vis]
 
-    def walk(self, rel, rem, uname, dots, scandir, lstat=False):
+    def walk(self, rel, rem, seen, uname, dots, scandir, lstat):
         """
         recursively yields from ./rem;
         rel is a unix-style user-defined vpath (not vfs-related)
         """
 
-        fsroot, vfs_ls, vfs_virt = self.ls(rem, uname, scandir, lstat)
+        fsroot, vfs_ls, vfs_virt = self.ls(
+            rem, uname, scandir, incl_wo=False, lstat=lstat
+        )
+
+        if seen and not fsroot.startswith(seen[-1]) and fsroot in seen:
+            print("bailing from symlink loop,\n  {}\n  {}".format(seen[-1], fsroot))
+            return
+
+        seen = seen[:] + [fsroot]
         rfiles = [x for x in vfs_ls if not stat.S_ISDIR(x[1].st_mode)]
         rdirs = [x for x in vfs_ls if stat.S_ISDIR(x[1].st_mode)]
 
@@ -177,7 +187,7 @@ class VFS(object):
 
             wrel = (rel + "/" + rdir).lstrip("/")
             wrem = (rem + "/" + rdir).lstrip("/")
-            for x in self.walk(wrel, wrem, uname, scandir, lstat):
+            for x in self.walk(wrel, wrem, seen, uname, dots, scandir, lstat):
                 yield x
 
         for n, vfs in sorted(vfs_virt.items()):
@@ -185,14 +195,16 @@ class VFS(object):
                 continue
 
             wrel = (rel + "/" + n).lstrip("/")
-            for x in vfs.walk(wrel, "", uname, scandir, lstat):
+            for x in vfs.walk(wrel, "", seen, uname, dots, scandir, lstat):
                 yield x
 
     def zipgen(self, vrem, flt, uname, dots, scandir):
         if flt:
             flt = {k: True for k in flt}
 
-        for vpath, apath, files, rd, vd in self.walk("", vrem, uname, dots, scandir):
+        for vpath, apath, files, rd, vd in self.walk(
+            "", vrem, [], uname, dots, scandir, False
+        ):
             if flt:
                 files = [x for x in files if x[0] in flt]
 
@@ -228,21 +240,19 @@ class VFS(object):
             for f in [{"vp": v, "ap": a, "st": n[1]} for v, a, n in files]:
                 yield f
 
-    def user_tree(self, uname, readable=False, writable=False, admin=False):
-        ret = []
-        opt1 = readable and (uname in self.uread or "*" in self.uread)
-        opt2 = writable and (uname in self.uwrite or "*" in self.uwrite)
-        if admin:
-            if opt1 and opt2:
-                ret.append(self.vpath)
-        else:
-            if opt1 or opt2:
-                ret.append(self.vpath)
+    def user_tree(self, uname, readable, writable, admin):
+        is_readable = False
+        if uname in self.uread or "*" in self.uread:
+            readable.append(self.vpath)
+            is_readable = True
+
+        if uname in self.uwrite or "*" in self.uwrite:
+            writable.append(self.vpath)
+            if is_readable:
+                admin.append(self.vpath)
 
         for _, vn in sorted(self.nodes.items()):
-            ret.extend(vn.user_tree(uname, readable, writable, admin))
-
-        return ret
+            vn.user_tree(uname, readable, writable, admin)
 
 
 class AuthSrv(object):
@@ -420,7 +430,7 @@ class AuthSrv(object):
             vfs = VFS(os.path.abspath("."), "", ["*"], ["*"])
         elif "" not in mount:
             # there's volumes but no root; make root inaccessible
-            vfs = VFS(os.path.abspath("."), "")
+            vfs = VFS(None, "")
             vfs.flags["d2d"] = True
 
         maxdepth = 0
@@ -559,3 +569,90 @@ class AuthSrv(object):
 
         # import pprint
         # pprint.pprint({"usr": user, "rd": mread, "wr": mwrite, "mnt": mount})
+
+    def dbg_ls(self):
+        users = self.args.ls
+        vols = "*"
+        flags = []
+
+        try:
+            users, vols = users.split(",", 1)
+        except:
+            pass
+
+        try:
+            vols, flags = vols.split(",", 1)
+            flags = flags.split(",")
+        except:
+            pass
+
+        if users == "**":
+            users = list(self.user.keys()) + ["*"]
+        else:
+            users = [users]
+
+        for u in users:
+            if u not in self.user and u != "*":
+                raise Exception("user not found: " + u)
+
+        if vols == "*":
+            vols = ["/" + x for x in self.vfs.all_vols.keys()]
+        else:
+            vols = [vols]
+
+        for v in vols:
+            if not v.startswith("/"):
+                raise Exception("volumes must start with /")
+
+            if v[1:] not in self.vfs.all_vols:
+                raise Exception("volume not found: " + v)
+
+        self.log({"users": users, "vols": vols, "flags": flags})
+        for k, v in self.vfs.all_vols.items():
+            self.log("/{}: read({}) write({})".format(k, v.uread, v.uwrite))
+
+        flag_v = "v" in flags
+        flag_ln = "ln" in flags
+        flag_p = "p" in flags
+        flag_r = "r" in flags
+
+        n_bads = 0
+        for v in vols:
+            v = v[1:]
+            vtop = "/{}/".format(v) if v else "/"
+            for u in users:
+                self.log("checking /{} as {}".format(v, u))
+                try:
+                    vn, _ = self.vfs.get(v, u, True, False)
+                except:
+                    continue
+
+                atop = vn.realpath
+                g = vn.walk("", "", [], u, True, not self.args.no_scandir, False)
+                for vpath, apath, files, _, _ in g:
+                    fnames = [n[0] for n in files]
+                    vpaths = [vpath + "/" + n for n in fnames] if vpath else fnames
+                    vpaths = [vtop + x for x in vpaths]
+                    apaths = [os.path.join(apath, n) for n in fnames]
+                    files = [[vpath + "/", apath + os.sep]] + list(zip(vpaths, apaths))
+
+                    if flag_ln:
+                        files = [x for x in files if not x[1].startswith(atop + os.sep)]
+                        n_bads += len(files)
+
+                    if flag_v:
+                        msg = [
+                            '# user "{}", vpath "{}"\n{}'.format(u, vp, ap)
+                            for vp, ap in files
+                        ]
+                    else:
+                        msg = [x[1] for x in files]
+
+                    if msg:
+                        nuprint("\n".join(msg))
+
+                if n_bads and flag_p:
+                    raise Exception("found symlink leaving volume, and strict is set")
+
+        if not flag_r:
+            sys.exit(0)

copyparty/broker_mp.py

@@ -44,7 +44,9 @@ class BrokerMp(object):
             proc.clients = {}
             proc.workload = 0
 
-            thr = threading.Thread(target=self.collector, args=(proc,))
+            thr = threading.Thread(
+                target=self.collector, args=(proc,), name="mp-collector"
+            )
             thr.daemon = True
             thr.start()
 
@@ -52,14 +54,19 @@ class BrokerMp(object):
             proc.start()
 
         if not self.args.q:
-            thr = threading.Thread(target=self.debug_load_balancer)
+            thr = threading.Thread(
+                target=self.debug_load_balancer, name="mp-dbg-loadbalancer"
+            )
             thr.daemon = True
             thr.start()
 
     def shutdown(self):
         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:

copyparty/broker_mpw.py

@@ -27,7 +27,7 @@ class MpWorker(object):
         self.retpend = {}
         self.retpend_mutex = threading.Lock()
         self.mutex = threading.Lock()
-        self.workload_thr_active = False
+        self.workload_thr_alive = False
 
         # we inherited signal_handler from parent,
         # replace it with something harmless
@@ -35,12 +35,12 @@ class MpWorker(object):
             signal.signal(signal.SIGINT, self.signal_handler)
 
         # instantiate all services here (TODO: inheritance?)
-        self.httpsrv = HttpSrv(self)
+        self.httpsrv = HttpSrv(self, True)
         self.httpsrv.disconnect_func = self.httpdrop
 
         # on winxp and some other platforms,
         # use thr.join() to block all signals
-        thr = threading.Thread(target=self.main)
+        thr = threading.Thread(target=self.main, name="mpw-main")
         thr.daemon = True
         thr.start()
         thr.join()
@@ -75,13 +75,15 @@ class MpWorker(object):
 
                 if self.args.log_conn:
                     self.log("%s %s" % addr, "|%sC-qpop" % ("-" * 4,), c="1;30")
-                
+
                 self.httpsrv.accept(sck, addr)
 
                 with self.mutex:
-                    if not self.workload_thr_active:
+                    if not self.workload_thr_alive:
                         self.workload_thr_alive = True
-                        thr = threading.Thread(target=self.thr_workload)
+                        thr = threading.Thread(
+                            target=self.thr_workload, name="mpw-workload"
+                        )
                         thr.daemon = True
                         thr.start()
 

copyparty/httpcli.py

@@ -16,6 +16,7 @@ import calendar
 
 from .__init__ import E, PY2, WINDOWS, ANYWIN
 from .util import *  # noqa  # pylint: disable=unused-wildcard-import
+from .authsrv import AuthSrv
 from .szip import StreamZip
 from .star import StreamTar
 
@@ -35,12 +36,13 @@ class HttpCli(object):
     def __init__(self, conn):
         self.t0 = time.time()
         self.conn = conn
-        self.s = conn.s
-        self.sr = conn.sr
+        self.s = conn.s  # type: socket
+        self.sr = conn.sr  # type: Unrecv
         self.ip = conn.addr[0]
-        self.addr = conn.addr
+        self.addr = conn.addr  # type: tuple[str, int]
         self.args = conn.args
-        self.auth = conn.auth
+        self.is_mp = conn.is_mp
+        self.auth = conn.auth  # type: AuthSrv
         self.ico = conn.ico
         self.thumbcli = conn.thumbcli
         self.log_func = conn.log_func
@@ -153,10 +155,8 @@ class HttpCli(object):
 
         pwd = uparam.get("pw")
         self.uname = self.auth.iuser.get(pwd, "*")
-        if self.uname:
-            self.rvol = self.auth.vfs.user_tree(self.uname, readable=True)
-            self.wvol = self.auth.vfs.user_tree(self.uname, writable=True)
-            self.avol = self.auth.vfs.user_tree(self.uname, True, True, True)
+        self.rvol, self.wvol, self.avol = [[], [], []]
+        self.auth.vfs.user_tree(self.uname, self.rvol, self.wvol, self.avol)
 
         ua = self.headers.get("user-agent", "")
         self.is_rclone = ua.startswith("rclone/")
@@ -508,6 +508,7 @@ class HttpCli(object):
         items = items.replace("\r", "").split("\n")
         items = [unquotep(x) for x in items if items]
 
+        self.parser.drop()
         return self.tx_zip(k, v, vn, rem, items, self.args.ed)
 
     def handle_post_json(self):
@@ -600,8 +601,9 @@ class HttpCli(object):
             taglist = {}
         else:
             # search by query params
-            self.log("qj: " + repr(body))
-            hits, taglist = idx.search(vols, body)
+            q = body["q"]
+            self.log("qj: " + q)
+            hits, taglist = idx.search(vols, q)
             msg = len(hits)
 
         idx.p_end = time.time()
@@ -850,14 +852,28 @@ class HttpCli(object):
             status = "ERROR"
 
         msg = "{} // {} bytes // {:.3f} MiB/s\n".format(status, sz_total, spd)
+        jmsg = {"status": status, "sz": sz_total, "mbps": round(spd, 3), "files": []}
 
         for sz, sha512, ofn, lfn in files:
-            vpath = self.vpath + "/" + lfn
+            vpath = (self.vpath + "/" if self.vpath else "") + lfn
             msg += 'sha512: {} // {} bytes // <a href="/{}">{}</a>\n'.format(
                 sha512[:56], sz, quotep(vpath), html_escape(ofn, crlf=True)
             )
             # truncated SHA-512 prevents length extension attacks;
             # using SHA-512/224, optionally SHA-512/256 = :64
+            jpart = {
+                "url": "{}://{}/{}".format(
+                    "https" if self.tls else "http",
+                    self.headers.get("host", "copyparty"),
+                    vpath,
+                ),
+                "sha512": sha512[:56],
+                "sz": sz,
+                "fn": lfn,
+                "fn_orig": ofn,
+                "path": vpath,
+            }
+            jmsg["files"].append(jpart)
 
         vspd = self._spd(sz_total, False)
         self.log("{} {}".format(vspd, msg))
@@ -869,7 +885,12 @@ class HttpCli(object):
                 ft = "{}\n{}\n{}\n".format(ft, msg.rstrip(), errmsg)
                 f.write(ft.encode("utf-8"))
 
-        self.redirect(self.vpath, msg=msg, flavor="return to", click=False)
+        if "j" in self.uparam:
+            jtxt = json.dumps(jmsg, indent=2, sort_keys=True)
+            self.reply(jtxt.encode("utf-8", "replace"), mime="application/json")
+        else:
+            self.redirect(self.vpath, msg=msg, flavor="return to", click=False)
+
         self.parser.drop()
         return True
 
@@ -973,6 +994,8 @@ class HttpCli(object):
         cli_lastmod = self.headers.get("if-modified-since")
         if cli_lastmod:
             try:
+                # some browser append "; length=573"
+                cli_lastmod = cli_lastmod.split(";")[0].strip()
                 cli_dt = time.strptime(cli_lastmod, HTTP_TS_FMT)
                 cli_ts = calendar.timegm(cli_dt)
                 return file_lastmod, int(file_ts) > int(cli_ts)
@@ -1142,7 +1165,8 @@ class HttpCli(object):
             if use_sendfile:
                 remains = sendfile_kern(lower, upper, f, self.s)
             else:
-                remains = sendfile_py(lower, upper, f, self.s)
+                actor = self.conn if self.is_mp else None
+                remains = sendfile_py(lower, upper, f, self.s, actor)
 
         if remains > 0:
             logmsg += " \033[31m" + unicode(upper - remains) + "\033[0m"
@@ -1310,20 +1334,23 @@ class HttpCli(object):
 
     def tx_mounts(self):
         suf = self.urlq(rm=["h"])
-        rvol = [x + "/" if x else x for x in self.rvol]
-        wvol = [x + "/" if x else x for x in self.wvol]
+        rvol, wvol, avol = [
+            [("/" + x).rstrip("/") + "/" for x in y]
+            for y in [self.rvol, self.wvol, self.avol]
+        ]
 
         vstate = {}
         if self.avol and not self.args.no_rescan:
             x = self.conn.hsrv.broker.put(True, "up2k.get_volstate")
             vstate = json.loads(x.get())
+            vstate = {("/" + k).rstrip("/") + "/": v for k, v in vstate.items()}
 
         html = self.j2(
             "splash",
             this=self,
             rvol=rvol,
             wvol=wvol,
-            avol=self.avol,
+            avol=avol,
             vstate=vstate,
             url_suf=suf,
         )
@@ -1355,15 +1382,31 @@ class HttpCli(object):
         if self.args.no_stack:
             raise Pebkac(403, "disabled by argv")
 
-        ret = []
+        threads = {}
         names = dict([(t.ident, t.name) for t in threading.enumerate()])
         for tid, stack in sys._current_frames().items():
-            ret.append("\n\n# {} ({:x})".format(names.get(tid), tid))
+            name = "{} ({:x})".format(names.get(tid), tid)
+            threads[name] = stack
+
+        rret = []
+        bret = []
+        for name, stack in sorted(threads.items()):
+            ret = ["\n\n# {}".format(name)]
+            pad = None
             for fn, lno, name, line in traceback.extract_stack(stack):
+                fn = os.sep.join(fn.split(os.sep)[-3:])
                 ret.append('File: "{}", line {}, in {}'.format(fn, lno, name))
                 if line:
                     ret.append("  " + str(line.strip()))
+                    if "self.not_empty.wait()" in line:
+                        pad = " " * 4
 
+            if pad:
+                bret += [ret[0]] + [pad + x for x in ret[1:]]
+            else:
+                rret += ret
+
+        ret = rret + bret
         ret = ("<pre>" + "\n".join(ret)).encode("utf-8")
         self.reply(ret)
 
@@ -1396,7 +1439,9 @@ class HttpCli(object):
 
         try:
             vn, rem = self.auth.vfs.get(top, self.uname, True, False)
-            fsroot, vfs_ls, vfs_virt = vn.ls(rem, self.uname, not self.args.no_scandir)
+            fsroot, vfs_ls, vfs_virt = vn.ls(
+                rem, self.uname, not self.args.no_scandir, incl_wo=True
+            )
         except:
             vfs_ls = []
             vfs_virt = {}
@@ -1561,7 +1606,9 @@ class HttpCli(object):
             if v is not None:
                 return self.tx_zip(k, v, vn, rem, [], self.args.ed)
 
-        fsroot, vfs_ls, vfs_virt = vn.ls(rem, self.uname, not self.args.no_scandir)
+        fsroot, vfs_ls, vfs_virt = vn.ls(
+            rem, self.uname, not self.args.no_scandir, incl_wo=True
+        )
         stats = {k: v for k, v in vfs_ls}
         vfs_ls = [x[0] for x in vfs_ls]
         vfs_ls.extend(vfs_virt.keys())

copyparty/httpconn.py

@@ -35,6 +35,7 @@ class HttpConn(object):
 
         self.args = hsrv.args
         self.auth = hsrv.auth
+        self.is_mp = hsrv.is_mp
         self.cert_path = hsrv.cert_path
 
         enth = HAVE_PIL and not self.args.no_thumb
@@ -174,6 +175,11 @@ class HttpConn(object):
             self.sr = Unrecv(self.s)
 
         while True:
+            if self.is_mp:
+                self.workload += 50
+                if self.workload >= 2 ** 31:
+                    self.workload = 100
+
             cli = HttpCli(self)
             if not cli.run():
                 return

copyparty/httpsrv.py

@@ -25,8 +25,8 @@ except ImportError:
     sys.exit(1)
 
 from .__init__ import E, MACOS
-from .httpconn import HttpConn
 from .authsrv import AuthSrv
+from .httpconn import HttpConn
 
 
 class HttpSrv(object):
@@ -35,8 +35,9 @@ class HttpSrv(object):
     relying on MpSrv for performance (HttpSrv is just plain threads)
     """
 
-    def __init__(self, broker):
+    def __init__(self, broker, is_mp=False):
         self.broker = broker
+        self.is_mp = is_mp
         self.args = broker.args
         self.log = broker.log
 
@@ -66,7 +67,11 @@ class HttpSrv(object):
         if self.args.log_conn:
             self.log("%s %s" % addr, "|%sC-cthr" % ("-" * 5,), c="1;30")
 
-        thr = threading.Thread(target=self.thr_client, args=(sck, addr))
+        thr = threading.Thread(
+            target=self.thr_client,
+            args=(sck, addr),
+            name="httpsrv-{}-{}".format(addr[0].split(".", 2)[-1][-6:], addr[1]),
+        )
         thr.daemon = True
         thr.start()
 
@@ -84,13 +89,16 @@ class HttpSrv(object):
         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()
+            if self.is_mp:
+                self.workload += 50
+                if not self.workload_thr_alive:
+                    self.workload_thr_alive = True
+                    thr = threading.Thread(
+                        target=self.thr_workload, name="httpsrv-workload"
+                    )
+                    thr.daemon = True
+                    thr.start()
 
         try:
             if self.args.log_conn:
@@ -99,6 +107,7 @@ class HttpSrv(object):
             cli.run()
 
         finally:
+            sck = cli.s
             if self.args.log_conn:
                 self.log("%s %s" % addr, "|%sC-cdone" % ("-" * 7,), c="1;30")
 

copyparty/ico.py

@@ -1,3 +1,6 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
 import hashlib
 import colorsys
 

copyparty/star.py

@@ -1,3 +1,6 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
 import os
 import tarfile
 import threading
@@ -42,7 +45,7 @@ class StreamTar(object):
         fmt = tarfile.GNU_FORMAT
         self.tar = tarfile.open(fileobj=self.qfile, mode="w|", format=fmt)
 
-        w = threading.Thread(target=self._gen)
+        w = threading.Thread(target=self._gen, name="star-gen")
         w.daemon = True
         w.start()
 

copyparty/sutil.py

@@ -1,3 +1,6 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
 import os
 import time
 import tempfile

copyparty/svchub.py

@@ -39,6 +39,8 @@ class SvcHub(object):
 
         # jank goes here
         auth = AuthSrv(self.args, self.log, False)
+        if args.ls:
+            auth.dbg_ls()
 
         # initiate all services to manage
         self.tcpsrv = TcpSrv(self)
@@ -69,7 +71,7 @@ class SvcHub(object):
         self.broker = Broker(self)
 
     def run(self):
-        thr = threading.Thread(target=self.tcpsrv.run)
+        thr = threading.Thread(target=self.tcpsrv.run, name="svchub-main")
         thr.daemon = True
         thr.start()
 
@@ -93,9 +95,11 @@ class SvcHub(object):
                         break
 
                     if n == 3:
-                        print("waiting for thumbsrv...")
+                        print("waiting for thumbsrv (10sec)...")
 
-            print("nailed it")
+            print("nailed it", end="")
+        finally:
+            print("\033[0m")
 
     def _log_disabled(self, src, msg, c=0):
         pass

copyparty/szip.py

@@ -1,3 +1,6 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
 import os
 import time
 import zlib

copyparty/th_cli.py

@@ -1,3 +1,6 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
 import os
 import time
 

copyparty/th_srv.py

@@ -1,3 +1,6 @@
+# coding: utf-8
+from __future__ import print_function, unicode_literals
+
 import os
 import sys
 import time
@@ -114,8 +117,10 @@ class ThumbSrv(object):
         self.stopping = False
         self.nthr = os.cpu_count() if hasattr(os, "cpu_count") else 4
         self.q = Queue(self.nthr * 4)
-        for _ in range(self.nthr):
-            t = threading.Thread(target=self.worker)
+        for n in range(self.nthr):
+            t = threading.Thread(
+                target=self.worker, name="thumb-{}-{}".format(n, self.nthr)
+            )
             t.daemon = True
             t.start()
 
@@ -129,9 +134,9 @@ class ThumbSrv(object):
 
             msg = "cannot create video thumbnails because some of the required programs are not available: "
             msg += ", ".join(missing)
-            self.log(msg, c=1)
+            self.log(msg, c=3)
 
-        t = threading.Thread(target=self.cleaner)
+        t = threading.Thread(target=self.cleaner, name="thumb-cleaner")
         t.daemon = True
         t.start()
 
@@ -316,10 +321,10 @@ class ThumbSrv(object):
             time.sleep(interval)
             for vol in self.vols:
                 vol += "/.hist/th"
-                self.log("cln {}/".format(vol))
+                self.log("\033[Jcln {}/\033[A".format(vol))
                 self.clean(vol)
 
-            self.log("cln ok")
+            self.log("\033[Jcln ok")
 
     def clean(self, vol):
         # self.log("cln {}".format(vol))

copyparty/u2idx.py

@@ -47,11 +47,11 @@ class U2idx(object):
         fhash = body["hash"]
         wark = up2k_wark_from_hashlist(self.args.salt, fsize, fhash)
 
-        uq = "substr(w,1,16) = ? and w = ?"
+        uq = "where substr(w,1,16) = ? and w = ?"
         uv = [wark[:16], wark]
 
         try:
-            return self.run_query(vols, uq, uv, {})[0]
+            return self.run_query(vols, uq, uv)[0]
         except Exception as ex:
             raise Pebkac(500, repr(ex))
 
@@ -67,37 +67,121 @@ class U2idx(object):
         self.cur[ptop] = cur
         return cur
 
-    def search(self, vols, body):
+    def search(self, vols, uq):
         """search by query params"""
         if not HAVE_SQLITE3:
             return []
 
-        qobj = {}
-        _conv_sz(qobj, body, "sz_min", "up.sz >= ?")
-        _conv_sz(qobj, body, "sz_max", "up.sz <= ?")
-        _conv_dt(qobj, body, "dt_min", "up.mt >= ?")
-        _conv_dt(qobj, body, "dt_max", "up.mt <= ?")
-        for seg, dk in [["path", "up.rd"], ["name", "up.fn"]]:
-            if seg in body:
-                _conv_txt(qobj, body, seg, dk)
+        q = ""
+        va = []
+        joins = ""
+        is_key = True
+        is_size = False
+        is_date = False
+        kw_key = ["(", ")", "and ", "or ", "not "]
+        kw_val = ["==", "=", "!=", ">", ">=", "<", "<=", "like "]
+        ptn_mt = re.compile(r"^\.?[a-z]+$")
+        mt_ctr = 0
+        mt_keycmp = "substr(up.w,1,16)"
+        mt_keycmp2 = None
 
-        uq, uv = _sqlize(qobj)
+        while True:
+            uq = uq.strip()
+            if not uq:
+                break
 
-        qobj = {}
-        if "tags" in body:
-            _conv_txt(qobj, body, "tags", "mt.v")
+            ok = False
+            for kw in kw_key + kw_val:
+                if uq.startswith(kw):
+                    is_key = kw in kw_key
+                    uq = uq[len(kw) :]
+                    ok = True
+                    q += kw
+                    break
 
-        if "adv" in body:
-            _conv_adv(qobj, body, "adv")
+            if ok:
+                continue
+
+            v, uq = (uq + " ").split(" ", 1)
+            if is_key:
+                is_key = False
+
+                if v == "size":
+                    v = "up.sz"
+                    is_size = True
+
+                elif v == "date":
+                    v = "up.mt"
+                    is_date = True
+
+                elif v == "path":
+                    v = "up.rd"
+
+                elif v == "name":
+                    v = "up.fn"
+
+                elif v == "tags" or ptn_mt.match(v):
+                    mt_ctr += 1
+                    mt_keycmp2 = "mt{}.w".format(mt_ctr)
+                    joins += "inner join mt mt{} on {} = {} ".format(
+                        mt_ctr, mt_keycmp, mt_keycmp2
+                    )
+                    mt_keycmp = mt_keycmp2
+                    if v == "tags":
+                        v = "mt{0}.v".format(mt_ctr)
+                    else:
+                        v = "+mt{0}.k = '{1}' and mt{0}.v".format(mt_ctr, v)
+
+                else:
+                    raise Pebkac(400, "invalid key [" + v + "]")
+
+                q += v + " "
+                continue
+
+            head = ""
+            tail = ""
+
+            if is_date:
+                is_date = False
+                v = v.upper().rstrip("Z").replace(",", " ").replace("T", " ")
+                while "  " in v:
+                    v = v.replace("  ", " ")
+
+                for fmt in [
+                    "%Y-%m-%d %H:%M:%S",
+                    "%Y-%m-%d %H:%M",
+                    "%Y-%m-%d %H",
+                    "%Y-%m-%d",
+                ]:
+                    try:
+                        v = datetime.strptime(v, fmt).timestamp()
+                        break
+                    except:
+                        pass
+
+            elif is_size:
+                is_size = False
+                v = int(float(v) * 1024 * 1024)
+
+            else:
+                if v.startswith("*"):
+                    head = "'%'||"
+                    v = v[1:]
+
+                if v.endswith("*"):
+                    tail = "||'%'"
+                    v = v[:-1]
+
+            q += " {}?{} ".format(head, tail)
+            va.append(v)
+            is_key = True
 
         try:
-            return self.run_query(vols, uq, uv, qobj)
+            return self.run_query(vols, joins + "where " + q, va)
         except Exception as ex:
             raise Pebkac(500, repr(ex))
 
-    def run_query(self, vols, uq, uv, targs):
-        self.log("qs: {} {} ,  {}".format(uq, repr(uv), repr(targs)))
-
+    def run_query(self, vols, uq, uv):
         done_flag = []
         self.active_id = "{:.6f}_{}".format(
             time.time(), threading.current_thread().ident
@@ -108,39 +192,19 @@ class U2idx(object):
                 self.active_id,
                 done_flag,
             ),
+            name="u2idx-terminator",
         )
         thr.daemon = True
         thr.start()
 
-        if not targs:
-            if not uq:
-                q = "select * from up"
-                v = ()
-            else:
-                q = "select * from up where " + uq
-                v = tuple(uv)
+        if not uq or not uv:
+            q = "select * from up"
+            v = ()
         else:
-            q = "select up.* from up"
-            keycmp = "substr(up.w,1,16)"
-            where = []
-            v = []
-            ctr = 0
-            for tq, tv in sorted(targs.items()):
-                ctr += 1
-                tq = tq.split("\n")[0]
-                keycmp2 = "mt{}.w".format(ctr)
-                q += " inner join mt mt{} on {} = {}".format(ctr, keycmp, keycmp2)
-                keycmp = keycmp2
-                where.append(tq.replace("mt.", keycmp[:-1]))
-                v.append(tv)
+            q = "select up.* from up " + uq
+            v = tuple(uv)
 
-            if uq:
-                where.append(uq)
-                v.extend(uv)
-
-            q += " where " + (" and ".join(where))
-
-        # self.log("q2: {} {}".format(q, repr(v)))
+        self.log("qs: {!r} {!r}".format(q, v))
 
         ret = []
         lim = 1000
@@ -163,7 +227,7 @@ class U2idx(object):
                 if rd.startswith("//") or fn.startswith("//"):
                     rd, fn = s3dec(rd, fn)
 
-                rp = "/".join([vtop, rd, fn])
+                rp = "/".join([x for x in [vtop, rd, fn] if x])
                 sret.append({"ts": int(ts), "sz": sz, "rp": rp, "w": w[:16]})
 
             for hit in sret:
@@ -204,78 +268,3 @@ def _open(ptop):
     db_path = os.path.join(ptop, ".hist", "up2k.db")
     if os.path.exists(db_path):
         return sqlite3.connect(db_path).cursor()
-
-
-def _conv_sz(q, body, k, sql):
-    if k in body:
-        q[sql] = int(float(body[k]) * 1024 * 1024)
-
-
-def _conv_dt(q, body, k, sql):
-    if k not in body:
-        return
-
-    v = body[k].upper().rstrip("Z").replace(",", " ").replace("T", " ")
-    while "  " in v:
-        v = v.replace("  ", " ")
-
-    for fmt in ["%Y-%m-%d %H:%M:%S", "%Y-%m-%d %H:%M", "%Y-%m-%d %H", "%Y-%m-%d"]:
-        try:
-            ts = datetime.strptime(v, fmt).timestamp()
-            break
-        except:
-            ts = None
-
-    if ts:
-        q[sql] = ts
-
-
-def _conv_txt(q, body, k, sql):
-    for v in body[k].split(" "):
-        inv = ""
-        if v.startswith("-"):
-            inv = "not"
-            v = v[1:]
-
-        if not v:
-            continue
-
-        head = "'%'||"
-        if v.startswith("^"):
-            head = ""
-            v = v[1:]
-
-        tail = "||'%'"
-        if v.endswith("$"):
-            tail = ""
-            v = v[:-1]
-
-        qk = "{} {} like {}?{}".format(sql, inv, head, tail)
-        q[qk + "\n" + v] = u8safe(v)
-
-
-def _conv_adv(q, body, k):
-    ptn = re.compile(r"^(\.?[a-z]+) *(==?|!=|<=?|>=?) *(.*)$")
-
-    parts = body[k].split(" ")
-    parts = [x.strip() for x in parts if x.strip()]
-
-    for part in parts:
-        m = ptn.match(part)
-        if not m:
-            p = html_escape(part)
-            raise Pebkac(400, "invalid argument [" + p + "]")
-
-        k, op, v = m.groups()
-        qk = "mt.k = '{}' and mt.v {} ?".format(k, op)
-        q[qk + "\n" + v] = u8safe(v)
-
-
-def _sqlize(qobj):
-    keys = []
-    values = []
-    for k, v in sorted(qobj.items()):
-        keys.append(k.split("\n")[0])
-        values.append(v)
-
-    return " and ".join(keys), values

copyparty/up2k.py

@@ -83,7 +83,7 @@ class Up2k(object):
         if ANYWIN:
             # usually fails to set lastmod too quickly
             self.lastmod_q = Queue()
-            thr = threading.Thread(target=self._lastmodder)
+            thr = threading.Thread(target=self._lastmodder, name="up2k-lastmod")
             thr.daemon = True
             thr.start()
 
@@ -96,7 +96,9 @@ class Up2k(object):
         if self.args.no_fastboot:
             self.deferred_init(all_vols)
         else:
-            t = threading.Thread(target=self.deferred_init, args=(all_vols,))
+            t = threading.Thread(
+                target=self.deferred_init, args=(all_vols,), name="up2k-deferred-init"
+            )
             t.daemon = True
             t.start()
 
@@ -104,20 +106,20 @@ class Up2k(object):
         have_e2d = self.init_indexes(all_vols)
 
         if have_e2d:
-            thr = threading.Thread(target=self._snapshot)
+            thr = threading.Thread(target=self._snapshot, name="up2k-snapshot")
             thr.daemon = True
             thr.start()
 
-            thr = threading.Thread(target=self._hasher)
+            thr = threading.Thread(target=self._hasher, name="up2k-hasher")
             thr.daemon = True
             thr.start()
 
             if self.mtag:
-                thr = threading.Thread(target=self._tagger)
+                thr = threading.Thread(target=self._tagger, name="up2k-tagger")
                 thr.daemon = True
                 thr.start()
 
-                thr = threading.Thread(target=self._run_all_mtp)
+                thr = threading.Thread(target=self._run_all_mtp, name="up2k-mtp-init")
                 thr.daemon = True
                 thr.start()
 
@@ -132,7 +134,11 @@ class Up2k(object):
             return "cannot initiate; scan is already in progress"
 
         args = (all_vols, scan_vols)
-        t = threading.Thread(target=self.init_indexes, args=args)
+        t = threading.Thread(
+            target=self.init_indexes,
+            args=args,
+            name="up2k-rescan-{}".format(scan_vols[0]),
+        )
         t.daemon = True
         t.start()
         return None
@@ -181,17 +187,22 @@ class Up2k(object):
         for vol in vols:
             try:
                 os.listdir(vol.realpath)
-                if not self.register_vpath(vol.realpath, vol.flags):
-                    raise Exception()
-
-                if vol.vpath in scan_vols or not scan_vols:
-                    live_vols.append(vol)
-
-                if vol.vpath not in self.volstate:
-                    self.volstate[vol.vpath] = "OFFLINE (not initialized)"
             except:
+                self.volstate[vol.vpath] = "OFFLINE (cannot access folder)"
+                self.log("cannot access " + vol.realpath, c=1)
+                continue
+
+            if scan_vols and vol.vpath not in scan_vols:
+                continue
+
+            if not self.register_vpath(vol.realpath, vol.flags):
                 # self.log("db not enabled for {}".format(m, vol.realpath))
-                pass
+                continue
+
+            live_vols.append(vol)
+
+            if vol.vpath not in self.volstate:
+                self.volstate[vol.vpath] = "OFFLINE (pending initialization)"
 
         vols = live_vols
         need_vac = {}
@@ -268,7 +279,7 @@ class Up2k(object):
         if self.mtag:
             m = "online (running mtp)"
             if scan_vols:
-                thr = threading.Thread(target=self._run_all_mtp)
+                thr = threading.Thread(target=self._run_all_mtp, name="up2k-mtp-scan")
                 thr.daemon = True
         else:
             del self.pp
@@ -285,7 +296,10 @@ class Up2k(object):
     def register_vpath(self, ptop, flags):
         db_path = os.path.join(ptop, ".hist", "up2k.db")
         if ptop in self.registry:
-            return [self.cur[ptop], db_path]
+            try:
+                return [self.cur[ptop], db_path]
+            except:
+                return None
 
         _, flags = self._expr_idx_filter(flags)
 
@@ -367,7 +381,8 @@ class Up2k(object):
         self.pp.msg = "a{} {}".format(self.pp.n, cdir)
         histdir = os.path.join(top, ".hist")
         ret = 0
-        for iname, inf in statdir(self.log, not self.args.no_scandir, False, cdir):
+        g = statdir(self.log, not self.args.no_scandir, False, cdir)
+        for iname, inf in sorted(g):
             abspath = os.path.join(cdir, iname)
             lmod = int(inf.st_mtime)
             if stat.S_ISDIR(inf.st_mode):
@@ -549,9 +564,10 @@ class Up2k(object):
                     last_write = time.time()
                     n_buf = 0
 
-            self._stop_mpool(mpool)
-            with self.mutex:
-                n_add += len(self._flush_mpool(c3))
+            if mpool:
+                self._stop_mpool(mpool)
+                with self.mutex:
+                    n_add += len(self._flush_mpool(c3))
 
             conn.commit()
             c3.close()
@@ -585,7 +601,8 @@ class Up2k(object):
 
         del self.pp
         for k in list(self.volstate.keys()):
-            self.volstate[k] = "online, idle"
+            if "OFFLINE" not in self.volstate[k]:
+                self.volstate[k] = "online, idle"
 
     def _run_one_mtp(self, ptop):
         entags = self.entags[ptop]
@@ -748,7 +765,9 @@ class Up2k(object):
 
         mpool = Queue(nw)
         for _ in range(nw):
-            thr = threading.Thread(target=self._tag_thr, args=(mpool,))
+            thr = threading.Thread(
+                target=self._tag_thr, args=(mpool,), name="up2k-mpool"
+            )
             thr.daemon = True
             thr.start()
 
@@ -1240,12 +1259,15 @@ class Up2k(object):
         return wark
 
     def _hashlist_from_file(self, path):
+        pp = self.pp if hasattr(self, "pp") else None
         fsz = os.path.getsize(fsenc(path))
         csz = up2k_chunksize(fsz)
         ret = []
         with open(fsenc(path), "rb", 512 * 1024) as f:
             while fsz > 0:
-                self.pp.msg = "{} MB, {}".format(int(fsz / 1024 / 1024), path)
+                if pp:
+                    pp.msg = "{} MB, {}".format(int(fsz / 1024 / 1024), path)
+
                 hashobj = hashlib.sha512()
                 rem = min(csz, fsz)
                 fsz -= rem

copyparty/util.py

@@ -193,7 +193,7 @@ class ProgressPrinter(threading.Thread):
     """
 
     def __init__(self):
-        threading.Thread.__init__(self)
+        threading.Thread.__init__(self, name="pp")
         self.daemon = True
         self.msg = None
         self.end = False
@@ -208,6 +208,8 @@ class ProgressPrinter(threading.Thread):
 
             msg = self.msg
             uprint(" {}\033[K\r".format(msg))
+            if PY2:
+                sys.stdout.flush()
 
         print("\033[K", end="")
         sys.stdout.flush()  # necessary on win10 even w/ stderr btw
@@ -262,6 +264,11 @@ def ren_open(fname, *args, **kwargs):
             yield {"orz": [f, fname]}
             return
 
+    if suffix:
+        ext = fname.split(".")[-1]
+        if len(ext) < 7:
+            suffix += "." + ext
+
     orig_name = fname
     bname = fname
     ext = ""
@@ -847,13 +854,14 @@ def yieldfile(fn):
 
 
 def hashcopy(actor, fin, fout):
-    u32_lim = int((2 ** 31) * 0.9)
+    is_mp = actor.is_mp
     hashobj = hashlib.sha512()
     tlen = 0
     for buf in fin:
-        actor.workload += 1
-        if actor.workload > u32_lim:
-            actor.workload = 100  # prevent overflow
+        if is_mp:
+            actor.workload += 1
+            if actor.workload > 2 ** 31:
+                actor.workload = 100
 
         tlen += len(buf)
         hashobj.update(buf)
@@ -865,12 +873,17 @@ def hashcopy(actor, fin, fout):
     return tlen, hashobj.hexdigest(), digest_b64
 
 
-def sendfile_py(lower, upper, f, s):
+def sendfile_py(lower, upper, f, s, actor=None):
     remains = upper - lower
     f.seek(lower)
     while remains > 0:
+        if actor:
+            actor.workload += 1
+            if actor.workload > 2 ** 31:
+                actor.workload = 100
+
         # time.sleep(0.01)
-        buf = f.read(min(4096, remains))
+        buf = f.read(min(1024 * 32, remains))
         if not buf:
             return remains
 

copyparty/web/browser.css

@@ -529,6 +529,17 @@ input[type="checkbox"]:checked+label {
 	height: 1em;
 	margin: .2em 0 -1em 1.6em;
 }
+#tq_raw {
+	width: calc(100% - 2em);
+	margin: .3em 0 0 1.4em;
+}
+#tq_raw td+td {
+	width: 100%;
+}
+#op_search #q_raw {
+	width: 100%;
+	display: block;
+}
 #files td div span {
 	color: #fff;
 	padding: 0 .4em;

copyparty/web/browser.js

@@ -803,7 +803,10 @@ var thegrid = (function () {
 			r.sz = v;
 			swrite('gridsz', r.sz);
 		}
-		document.documentElement.style.setProperty('--grid-sz', r.sz + 'em');
+		try {
+			document.documentElement.style.setProperty('--grid-sz', r.sz + 'em');
+		}
+		catch (ex) { }
 	}
 	setsz();
 
@@ -820,12 +823,25 @@ var thegrid = (function () {
 		this.setAttribute('class', tr.getAttribute('class'));
 	}
 
+	function bgopen(e) {
+		ev(e);
+		var url = this.getAttribute('href');
+		window.open(url, '_blank');
+	}
+
 	r.loadsel = function () {
-		var ths = QSA('#ggrid>a');
+		var ths = QSA('#ggrid>a'),
+			have_sel = !!QS('#files tr.sel');
+
 		for (var a = 0, aa = ths.length; a < aa; a++) {
-			ths[a].onclick = r.sel ? seltgl : null;
+			ths[a].onclick = r.sel ? seltgl : have_sel ? bgopen : null;
 			ths[a].setAttribute('class', ebi(ths[a].getAttribute('ref')).parentNode.parentNode.getAttribute('class'));
 		}
+		var uns = QS('#ggrid a[ref="unsearch"]');
+		if (uns)
+			uns.onclick = function () {
+				ebi('unsearch').click();
+			};
 	}
 
 	function loadgrid() {
@@ -836,9 +852,9 @@ var thegrid = (function () {
 			return r.loadsel();
 
 		var html = [];
-		var tr = lfiles.tBodies[0].rows;
-		for (var a = 0; a < tr.length; a++) {
-			var ao = tr[a].cells[1].firstChild,
+		var files = QSA('#files>tbody>tr>td:nth-child(2) a[id]');
+		for (var a = 0, aa = files.length; a < aa; a++) {
+			var ao = files[a],
 				href = esc(ao.getAttribute('href')),
 				ref = ao.getAttribute('id'),
 				isdir = href.split('?')[0].slice(-1)[0] == '/',
@@ -960,7 +976,7 @@ document.onkeydown = function (e) {
 	if (k == 'KeyT')
 		return ebi('thumbs').click();
 
-	if (window['thegrid'] && thegrid.en) {
+	if (thegrid.en) {
 		if (k == 'KeyS')
 			return ebi('gridsel').click();
 
@@ -1026,6 +1042,7 @@ document.onkeydown = function (e) {
 	for (var a = 0; a < trs.length; a += 2) {
 		html.push('<table>' + (trs[a].concat(trs[a + 1])).join('\n') + '</table>');
 	}
+	html.push('<table id="tq_raw"><tr><td>raw</td><td><input id="q_raw" type="text" name="q" /></td></tr></table>');
 	ebi('srch_form').innerHTML = html.join('\n');
 
 	var o = QSA('#op_search input');
@@ -1050,33 +1067,83 @@ document.onkeydown = function (e) {
 			var chk = ebi(id.slice(0, -1) + 'c');
 			chk.checked = ((v + '').length > 0);
 		}
+
+		if (id != "q_raw")
+			encode_query();
+
 		clearTimeout(search_timeout);
 		if (Date.now() - search_in_progress > 30 * 1000)
 			search_timeout = setTimeout(do_search, 200);
 	}
 
+	function encode_query() {
+		var q = '';
+		for (var a = 0; a < sconf.length; a++) {
+			for (var b = 1; b < sconf[a].length; b++) {
+				var k = sconf[a][b][0],
+					chk = 'srch_' + k + 'c',
+					tvs = ebi('srch_' + k + 'v').value.split(/ /g);
+
+				if (!ebi(chk).checked)
+					continue;
+
+				for (var c = 0; c < tvs.length; c++) {
+					var tv = tvs[c];
+					if (!tv.length)
+						break;
+
+					q += ' and ';
+
+					if (k == 'adv') {
+						q += tv.replace(/ /g, " and ").replace(/([=!><]=?)/, " $1 ");
+						continue;
+					}
+
+					if (k.length == 3) {
+						q += k.replace(/sz/, 'size').replace(/dt/, 'date').replace(/l$/, ' >= ').replace(/u$/, ' <= ') + tv;
+						continue;
+					}
+
+					if (k == 'path' || k == 'name' || k == 'tags') {
+						var not = ' ';
+						if (tv.slice(0, 1) == '-') {
+							tv = tv.slice(1);
+							not = ' not ';
+						}
+
+						if (tv.slice(0, 1) == '^') {
+							tv = tv.slice(1);
+						}
+						else {
+							tv = '*' + tv;
+						}
+
+						if (tv.slice(-1) == '$') {
+							tv = tv.slice(0, -1);
+						}
+						else {
+							tv += '*';
+						}
+
+						q += k + not + 'like ' + tv;
+					}
+				}
+			}
+		}
+		ebi('q_raw').value = q.slice(5);
+	}
+
 	function do_search() {
 		search_in_progress = Date.now();
 		srch_msg(false, "searching...");
 		clearTimeout(search_timeout);
 
-		var params = {},
-			o = QSA('#op_search input[type="text"]');
-
-		for (var a = 0; a < o.length; a++) {
-			var chk = ebi(o[a].getAttribute('id').slice(0, -1) + 'c');
-			if (!chk.checked)
-				continue;
-
-			params[o[a].getAttribute('name')] = o[a].value;
-		}
-		// ebi('srch_q').textContent = JSON.stringify(params, null, 4);
 		var xhr = new XMLHttpRequest();
 		xhr.open('POST', '/?srch', true);
 		xhr.setRequestHeader('Content-Type', 'text/plain');
 		xhr.onreadystatechange = xhr_search_results;
 		xhr.ts = Date.now();
-		xhr.send(JSON.stringify(params));
+		xhr.send(JSON.stringify({ "q": ebi('q_raw').value }));
 	}
 
 	function xhr_search_results() {
@@ -1381,7 +1448,7 @@ var treectl = (function () {
 		if (hpush)
 			get_tree('.', xhr.top);
 
-		enspin('#files');
+		enspin(thegrid.en ? '#gfiles' : '#files');
 	}
 
 	function treegrow(e) {
@@ -1461,6 +1528,7 @@ var treectl = (function () {
 
 		apply_perms(res.perms);
 		despin('#files');
+		despin('#gfiles');
 
 		ebi('pro').innerHTML = res.logues ? res.logues[0] || "" : "";
 		ebi('epi').innerHTML = res.logues ? res.logues[1] || "" : "";
@@ -1995,8 +2063,7 @@ var msel = (function () {
 	}
 	function selui() {
 		clmod(ebi('wtoggle'), 'sel', getsel().length);
-		if (window['thegrid'])
-			thegrid.loadsel();
+		thegrid.loadsel();
 	}
 	function seltgl(e) {
 		ev(e);

copyparty/web/splash.html

@@ -20,13 +20,13 @@
             <tbody>
                 {% for mp in avol %}
                 {%- if mp in vstate and vstate[mp] %}
-                <tr><td>/{{ mp }}</td><td><a href="/{{ mp }}?scan">rescan</a></td><td>{{ vstate[mp] }}</td></tr>
+                <tr><td><a href="{{ mp }}{{ url_suf }}">{{ mp }}</a></td><td><a href="{{ mp }}?scan">rescan</a></td><td>{{ vstate[mp] }}</td></tr>
                 {%- endif %}
                 {% endfor %}
             </tbody>
         </table>
         <div class="btns">
-            <a href="/{{ avol[0] }}?stack">dump stack</a>
+            <a href="{{ avol[0] }}?stack">dump stack</a>
         </div>
         {%- endif %}
 
@@ -34,7 +34,7 @@
         <h1>you can browse these:</h1>
         <ul>
             {% for mp in rvol %}
-            <li><a href="/{{ mp }}{{ url_suf }}">/{{ mp }}</a></li>
+            <li><a href="{{ mp }}{{ url_suf }}">{{ mp }}</a></li>
             {% endfor %}
         </ul>
         {%- endif %}
@@ -43,7 +43,7 @@
         <h1>you can upload to:</h1>
         <ul>
             {% for mp in wvol %}
-            <li><a href="/{{ mp }}{{ url_suf }}">/{{ mp }}</a></li>
+            <li><a href="{{ mp }}{{ url_suf }}">{{ mp }}</a></li>
             {% endfor %}
         </ul>
         {%- endif %}

copyparty/web/up2k.js

@@ -17,6 +17,7 @@ function goto_up2k() {
 // chrome requires https to use crypto.subtle,
 // usually it's undefined but some chromes throw on invoke
 var up2k = null;
+var sha_js = window.WebAssembly ? 'hw' : 'ac';  // ff53,c57,sa11
 try {
     var cf = crypto.subtle || crypto.webkitSubtle;
     cf.digest('SHA-512', new Uint8Array(1)).then(
@@ -430,13 +431,15 @@ function up2k_init(subtle) {
     // upload ui hidden by default, clicking the header shows it
     function init_deps() {
         if (!subtle && !window.asmCrypto) {
-            showmodal('<h1>loading sha512.js</h1><h2>since ' + shame + '</h2><h4>thanks chrome</h4>');
-            import_js('/.cpr/deps/sha512.js', unmodal);
+            var fn = 'sha512.' + sha_js + '.js';
+            showmodal('<h1>loading ' + fn + '</h1><h2>since ' + shame + '</h2><h4>thanks chrome</h4>');
+            import_js('/.cpr/deps/' + fn, unmodal);
 
             if (is_https)
                 ebi('u2foot').innerHTML = shame + ' so <em>this</em> uploader will do like 500kB/s at best';
             else
-                ebi('u2foot').innerHTML = 'seems like ' + shame + ' so do that if you want more performance';
+                ebi('u2foot').innerHTML = 'seems like ' + shame + ' so do that if you want more performance <span style="color:#' +
+                    (sha_js == 'ac' ? 'c84">(expecting 20' : '8a5">(but dont worry too much, expect 100') + ' MiB/s)</span>';
         }
     }
 
@@ -886,6 +889,10 @@ function up2k_init(subtle) {
         return base64;
     }
 
+    function hex2u8(txt) {
+        return new Uint8Array(txt.match(/.{2}/g).map(function (b) { return parseInt(b, 16); }));
+    }
+
     function get_chunksize(filesize) {
         var chunksize = 1024 * 1024,
             stepsize = 512 * 1024;
@@ -987,10 +994,18 @@ function up2k_init(subtle) {
             if (subtle)
                 subtle.digest('SHA-512', buf).then(hash_done);
             else setTimeout(function () {
-                var hasher = new asmCrypto.Sha512();
-                hasher.process(new Uint8Array(buf));
-                hasher.finish();
-                hash_done(hasher.result);
+                var u8buf = new Uint8Array(buf);
+                if (sha_js == 'hw') {
+                    hashwasm.sha512(u8buf).then(function (v) {
+                        hash_done(hex2u8(v))
+                    });
+                }
+                else {
+                    var hasher = new asmCrypto.Sha512();
+                    hasher.process(u8buf);
+                    hasher.finish();
+                    hash_done(hasher.result);
+                }
             }, 1);
         };
 

copyparty/web/upload.css

@@ -238,6 +238,10 @@
 	color: #fff;
 	font-style: italic;
 }
+#u2foot span {
+	color: #999;
+	font-size: .9em;
+}
 #u2footfoot {
 	margin-bottom: -1em;
 }

docs/notes.sh

@@ -80,6 +80,13 @@ command -v gdate && date() { gdate "$@"; }; while true; do t=$(date +%s.%N); (ti
 var t=[]; var b=document.location.href.split('#')[0].slice(0, -1); document.querySelectorAll('#u2tab .prog a').forEach((x) => {t.push(b+encodeURI(x.getAttribute("href")))}); console.log(t.join("\n"));
 
 
+##
+## bash oneliners
+
+# get the size and video-id of all youtube vids in folder, assuming filename ends with -id.ext, and create a copyparty search query
+find -maxdepth 1 -printf '%s %p\n' | sort -n | awk '!/-([0-9a-zA-Z_-]{11})\.(mkv|mp4|webm)$/{next} {sub(/\.[^\.]+$/,"");n=length($0);v=substr($0,n-10);print $1, v}' | tee /dev/stderr | awk 'BEGIN {p="("} {printf("%s name like *-%s.* ",p,$2);p="or"} END {print ")\n"}' | cat >&2
+
+
 ##
 ## sqlite3 stuff
 
@@ -146,6 +153,9 @@ dbg.asyncStore.pendingBreakpoints = {}
 # fix firefox phantom breakpoints
 about:config >> devtools.debugger.prefs-schema-version = -1
 
+# determine server version
+git reset --hard origin/HEAD && git log --format=format:"%H %ai %d" --decorate=full > /dev/shm/revs && cat /dev/shm/revs | while read -r rev extra; do (git reset --hard $rev >/dev/null 2>/dev/null && dsz=$(cat copyparty/web/{util,browser,up2k}.js 2>/dev/null | diff -wNarU0 - <(cat /mnt/Users/ed/Downloads/ref/{util,browser,up2k}.js) | wc -c) && printf '%s %6s %s\n' "$rev" $dsz "$extra") </dev/null; done                
+
 
 ##
 ## http 206

scripts/deps-docker/Dockerfile

@@ -9,6 +9,12 @@ ENV     ver_asmcrypto=5b994303a9d3e27e0915f72a10b6c2c51535a4dc \
         ver_zopfli=1.0.3
 
 
+# TODO
+# sha512.hw.js https://github.com/Daninet/hash-wasm
+# sha512.kc.js https://github.com/chm-diederichs/sha3-wasm
+# awk '/HMAC state/{o=1}  /var HEAP/{o=0}  /function hmac_reset/{o=1}  /return \{/{o=0}  /var __extends =/{o=1}  /var Hash =/{o=0}  /hmac_|pbkdf2_/{next}  o{next}  {gsub(/IllegalStateError/,"Exception")}  {sub(/^ +/,"");sub(/^\/\/ .*/,"");sub(/;$/," ;")}  1' <sha512.ac.js.orig >sha512.ac.js; for fn in sha512.ac.js.orig sha512.ac.js; do wc -c <$fn; wc -c <$fn.gz ; for n in {1..9}; do printf '%8d %d bz\n' $(bzip2 -c$n <$fn | wc -c) $n; done; done
+
+
 # download;
 # the scp url is latin from https://fonts.googleapis.com/css2?family=Source+Code+Pro&display=swap
 RUN     mkdir -p /z/dist/no-pk \

scripts/install-githooks.sh

@@ -0,0 +1,12 @@
+#!/bin/bash
+set -ex
+
+[ -e setup.py ] || ..
+[ -e setup.py ] || {
+    echo u wot
+    exit 1
+}
+
+cd .git/hooks
+rm -f pre-commit
+ln -s ../../scripts/run-tests.sh pre-commit

scripts/make-sfx.sh

@@ -32,6 +32,10 @@ gtar=$(command -v gtar || command -v gnutar) || true
 	[ -e /opt/local/bin/bzip2 ] &&
 		bzip2() { /opt/local/bin/bzip2 "$@"; }
 }
+
+gawk=$(command -v gawk || command -v gnuawk || command -v awk)
+awk() { $gawk "$@"; }
+
 pybin=$(command -v python3 || command -v python) || {
 	echo need python
 	exit 1
@@ -194,17 +198,46 @@ tmv "$f"
 
 # up2k goes from 28k to 22k laff
 echo entabbening
-find | grep -E '\.(js|css|html)$' | while IFS= read -r f; do
+find | grep -E '\.css$' | while IFS= read -r f; do
+	awk '{
+		sub(/^[ \t]+/,"");
+		sub(/[ \t]+$/,"");
+		$0=gensub(/^([a-z-]+) *: *(.*[^ ]) *;$/,"\\1:\\2;","1");
+		sub(/ +\{$/,"{");
+		gsub(/, /,",")
+	}
+	!/\}$/ {printf "%s",$0;next}
+	1
+	' <$f | gsed 's/;\}$/}/' >t
+	tmv "$f"
+done
+find | grep -E '\.(js|html)$' | while IFS= read -r f; do
 	unexpand -t 4 --first-only <"$f" >t
 	tmv "$f"
 done
 
+
+gzres() {
+command -v pigz &&
+	pk='pigz -11 -J 34 -I 100' ||
+	pk='gzip'
+
+echo "$pk"
+find | grep -E '\.(js|css)$' | while IFS= read -r f; do
+	echo -n .
+	$pk "$f"
+done
+echo
+}
+gzres
+
+
 echo gen tarlist
 for d in copyparty dep-j2; do find $d -type f; done |
 sed -r 's/(.*)\.(.*)/\2 \1/' | LC_ALL=C sort |
 sed -r 's/([^ ]*) (.*)/\2.\1/' | grep -vE '/list1?$' > list1
 
-(grep -vE 'gz$' list1; grep -E 'gz$' list1) >list
+(grep -vE '\.(gz|br)$' list1; grep -E '\.(gz|br)$' list1) >list || true
 
 echo creating tar
 args=(--owner=1000 --group=1000)

scripts/profile.py

@@ -0,0 +1,34 @@
+#!/usr/bin/env python3
+
+import sys
+
+sys.path.insert(0, ".")
+cmd = sys.argv[1]
+
+if cmd == "cpp":
+    from copyparty.__main__ import main
+
+    argv = ["__main__", "-v", "srv::r", "-v", "../../yt:yt:r"]
+    main(argv=argv)
+
+elif cmd == "test":
+    from unittest import main
+
+    argv = ["__main__", "discover", "-s", "tests"]
+    main(module=None, argv=argv)
+
+else:
+    raise Exception()
+
+# import dis; print(dis.dis(main))
+
+
+# macos:
+#   option1) python3.9 -m pip install --user -U vmprof==0.4.9
+#   option2) python3.9 -m pip install --user -U https://github.com/vmprof/vmprof-python/archive/refs/heads/master.zip
+#
+# python -m vmprof -o prof --lines ./scripts/profile.py test
+
+# linux: ~/.local/bin/vmprofshow prof tree | grep -vF '[1m  0.'
+# macos: ~/Library/Python/3.9/bin/vmprofshow prof tree | grep -vF '[1m  0.'
+#   win: %appdata%\..\Roaming\Python\Python39\Scripts\vmprofshow.exe prof tree

scripts/run-tests.sh

@@ -0,0 +1,12 @@
+#!/bin/bash
+set -ex
+
+pids=()
+for py in python{2,3}; do
+    $py -m unittest discover -s tests >/dev/null &
+    pids+=($!)
+done
+
+for pid in ${pids[@]}; do
+    wait $pid
+done

scripts/sfx.sh

@@ -47,7 +47,7 @@ grep -E '/(python|pypy)[0-9\.-]*$' >$dir/pys || true
 	printf '\033[1;30mlooking for jinja2 in [%s]\033[0m\n' "$_py" >&2
 	$_py -c 'import jinja2' 2>/dev/null || continue
 	printf '%s\n' "$_py"
-	mv $dir/{,x.}jinja2
+	mv $dir/{,x.}dep-j2
 	break
 done)"
 

tests/test_httpcli.py

@@ -8,13 +8,13 @@ import time
 import shutil
 import pprint
 import tarfile
+import tempfile
 import unittest
-
 from argparse import Namespace
-from copyparty.authsrv import AuthSrv
-from copyparty.httpcli import HttpCli
 
 from tests import util as tu
+from copyparty.authsrv import AuthSrv
+from copyparty.httpcli import HttpCli
 
 
 def hdr(query):
@@ -32,6 +32,8 @@ class Cfg(Namespace):
             no_zip=False,
             no_scandir=False,
             no_sendfile=True,
+            no_rescan=True,
+            ihead=False,
             nih=True,
             mtp=[],
             mte="a",
@@ -40,13 +42,15 @@ class Cfg(Namespace):
 
 
 class TestHttpCli(unittest.TestCase):
-    def test(self):
-        td = os.path.join(tu.get_ramdisk(), "vfs")
-        try:
-            shutil.rmtree(td)
-        except OSError:
-            pass
+    def setUp(self):
+        self.td = tu.get_ramdisk()
 
+    def tearDown(self):
+        os.chdir(tempfile.gettempdir())
+        shutil.rmtree(self.td)
+
+    def test(self):
+        td = os.path.join(self.td, "vfs")
         os.mkdir(td)
         os.chdir(td)
 

tests/test_vfs.py

@@ -7,13 +7,12 @@ import json
 import shutil
 import tempfile
 import unittest
-
 from textwrap import dedent
 from argparse import Namespace
-from copyparty.authsrv import AuthSrv
-from copyparty import util
 
 from tests import util as tu
+from copyparty.authsrv import AuthSrv, VFS
+from copyparty import util
 
 
 class Cfg(Namespace):
@@ -25,6 +24,13 @@ class Cfg(Namespace):
 
 
 class TestVFS(unittest.TestCase):
+    def setUp(self):
+        self.td = tu.get_ramdisk()
+
+    def tearDown(self):
+        os.chdir(tempfile.gettempdir())
+        shutil.rmtree(self.td)
+
     def dump(self, vfs):
         print(json.dumps(vfs, indent=4, sort_keys=True, default=lambda o: o.__dict__))
 
@@ -41,6 +47,7 @@ class TestVFS(unittest.TestCase):
         self.assertEqual(util.undot(query), response)
 
     def ls(self, vfs, vpath, uname):
+        # type: (VFS, str, str) -> tuple[str, str, str]
         """helper for resolving and listing a folder"""
         vn, rem = vfs.get(vpath, uname, True, False)
         r1 = vn.ls(rem, uname, False)
@@ -55,12 +62,7 @@ class TestVFS(unittest.TestCase):
         pass
 
     def test(self):
-        td = os.path.join(tu.get_ramdisk(), "vfs")
-        try:
-            shutil.rmtree(td)
-        except OSError:
-            pass
-
+        td = os.path.join(self.td, "vfs")
         os.mkdir(td)
         os.chdir(td)
 
@@ -227,7 +229,7 @@ class TestVFS(unittest.TestCase):
         self.assertEqual(list(v1), list(v2))
 
         # config file parser
-        cfg_path = os.path.join(tu.get_ramdisk(), "test.cfg")
+        cfg_path = os.path.join(self.td, "test.cfg")
         with open(cfg_path, "wb") as f:
             f.write(
                 dedent(
@@ -249,7 +251,7 @@ class TestVFS(unittest.TestCase):
         n = au.vfs
         # root was not defined, so PWD with no access to anyone
         self.assertEqual(n.vpath, "")
-        self.assertEqual(n.realpath, td)
+        self.assertEqual(n.realpath, None)
         self.assertEqual(n.uread, [])
         self.assertEqual(n.uwrite, [])
         self.assertEqual(len(n.nodes), 1)
@@ -260,6 +262,4 @@ class TestVFS(unittest.TestCase):
         self.assertEqual(n.uwrite, ["asd"])
         self.assertEqual(len(n.nodes), 0)
 
-        os.chdir(tempfile.gettempdir())
-        shutil.rmtree(td)
         os.unlink(cfg_path)

tests/util.py

@@ -1,16 +1,36 @@
 import os
+import sys
 import time
+import shutil
 import jinja2
 import tempfile
+import platform
 import subprocess as sp
 
-from copyparty.util import Unrecv
 
+WINDOWS = platform.system() == "Windows"
+ANYWIN = WINDOWS or sys.platform in ["msys"]
+MACOS = platform.system() == "Darwin"
 
 J2_ENV = jinja2.Environment(loader=jinja2.BaseLoader)
 J2_FILES = J2_ENV.from_string("{{ files|join('\n') }}")
 
 
+def nah(*a, **ka):
+    return False
+
+
+if MACOS:
+    import posixpath
+
+    posixpath.islink = nah
+    os.path.islink = nah
+    # 25% faster; until any tests do symlink stuff
+
+
+from copyparty.util import Unrecv
+
+
 def runcmd(*argv):
     p = sp.Popen(argv, stdout=sp.PIPE, stderr=sp.PIPE)
     stdout, stderr = p.communicate()
@@ -28,18 +48,25 @@ def chkcmd(*argv):
 
 
 def get_ramdisk():
+    def subdir(top):
+        ret = os.path.join(top, "cptd-{}".format(os.getpid()))
+        shutil.rmtree(ret, True)
+        os.mkdir(ret)
+        return ret
+
     for vol in ["/dev/shm", "/Volumes/cptd"]:  # nosec (singleton test)
         if os.path.exists(vol):
-            return vol
+            return subdir(vol)
 
     if os.path.exists("/Volumes"):
-        devname, _ = chkcmd("hdiutil", "attach", "-nomount", "ram://32768")
+        # hdiutil eject /Volumes/cptd/
+        devname, _ = chkcmd("hdiutil", "attach", "-nomount", "ram://65536")
         devname = devname.strip()
         print("devname: [{}]".format(devname))
         for _ in range(10):
             try:
                 _, _ = chkcmd("diskutil", "eraseVolume", "HFS+", "cptd", devname)
-                return "/Volumes/cptd"
+                return subdir("/Volumes/cptd")
             except Exception as ex:
                 print(repr(ex))
                 time.sleep(0.25)
@@ -50,7 +77,7 @@ def get_ramdisk():
     try:
         os.mkdir(ret)
     finally:
-        return ret
+        return subdir(ret)
 
 
 class NullBroker(object):
@@ -89,9 +116,13 @@ class VHttpConn(object):
         self.addr = ("127.0.0.1", "42069")
         self.args = args
         self.auth = auth
+        self.is_mp = False
         self.log_func = log
         self.log_src = "a"
+        self.lf_url = None
         self.hsrv = VHttpSrv()
         self.nbyte = 0
         self.workload = 0
+        self.ico = None
+        self.thumbcli = None
         self.t0 = time.time()