v0.7.1

.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

.vscode/launch.json

@@ -12,10 +12,12 @@
                 //"-nw",
                 "-ed",
                 "-emp",
+                "-e2d",
+                "-e2s",
                 "-a",
                 "ed:wark",
                 "-v",
-                "srv::r:aed"
+                "srv::r:aed:cnodupe"
             ]
         },
         {

README.md

@@ -13,6 +13,17 @@ turn your phone or raspi into a portable file server with resumable uploads/down
 * code standard: `black`
 
 
+## quickstart
+
+download [copyparty-sfx.py](https://github.com/9001/copyparty/releases/latest/download/copyparty-sfx.py) and you're all set!
+
+running the sfx without arguments (for example doubleclicking it on Windows) will let anyone access the current folder; see `-h` for help if you want accounts and volumes etc
+
+you may also want these, especially on servers:
+* [contrib/systemd/copyparty.service](contrib/systemd/copyparty.service) to run copyparty as a systemd service
+* [contrib/nginx/copyparty.conf](contrib/nginx/copyparty.conf) to reverse-proxy behind nginx (for legit https)
+
+
 ## notes
 
 * iPhone/iPad: use Firefox to download files
@@ -126,13 +137,15 @@ in the `scripts` folder:
 
 roughly sorted by priority
 
-* up2k handle filename too long
+* reduce up2k roundtrips
-* up2k fails on empty files? alert then stuck
+  * start from a chunk index and just go
+  * terminate client on bad data
 * drop onto folders
-* look into android thumbnail cache file format
+* `os.copy_file_range` for up2k cloning
+* up2k partials ui
 * support pillow-simd
 * cache sha512 chunks on client
-* symlink existing files on upload
 * comment field
+* ~~look into android thumbnail cache file format~~ bad idea
 * figure out the deal with pixel3a not being connectable as hotspot
   * pixel3a having unpredictable 3sec latency in general :||||

bin/README.md

@@ -34,3 +34,8 @@ you could replace winfsp with [dokan](https://github.com/dokan-dev/dokany/releas
 * does the same thing except more correct, `samba` approves
 * **supports Linux** -- expect `18 MiB/s` (wait what)
 * **supports Macos** -- probably
+
+
+
+# copyparty-fuse-streaming.py
+* pretend this doesn't exist

bin/copyparty-fuse.py

@@ -12,7 +12,7 @@ __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 copyparty-fuse.py http://192.168.1.69:3923/  ./music
 
 dependencies:
   python3 -m pip install --user fusepy
@@ -20,6 +20,10 @@ dependencies:
   + 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/master/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
 """
@@ -100,7 +104,7 @@ 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):
@@ -159,7 +163,7 @@ 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"))
@@ -367,7 +371,7 @@ class Gateway(object):
         ret = []
         remainder = b""
         ptn = re.compile(
-            r'^<tr><td>(-|DIR)</td><td><a[^>]* href="([^"]+)"[^>]*>([^<]+)</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:
@@ -405,7 +409,7 @@ class Gateway(object):
                     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])
@@ -658,8 +662,18 @@ class CPPF(Operations):
 
         else:
             if get2 - get1 <= 1024 * 1024:
-                h_ofs = get1 - 256 * 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
@@ -705,6 +719,7 @@ class CPPF(Operations):
             self.dircache.append(cn)
             self.clean_dircache()
 
+        # import pprint; pprint.pprint(ret)
         return ret
 
     def readdir(self, path, fh=None):
@@ -802,7 +817,11 @@ class CPPF(Operations):
                 # dbg("=" + repr(cache_stat))
                 return cache_stat
 
-        info("=ENOENT ({})".format(hexler(path)))
+        fun = info
+        if MACOS and path.split('/')[-1].startswith('._'):
+            fun = dbg
+        
+        fun("=ENOENT ({})".format(hexler(path)))
         raise FuseOSError(errno.ENOENT)
 
     access = None
@@ -906,6 +925,7 @@ class TheArgparseFormatter(
 
 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,

bin/copyparty-fuseb.py

@@ -567,6 +567,8 @@ class CPPF(Fuse):
 
 
 def main():
+    time.strptime("19970815", "%Y%m%d")  # python#7980
+
     server = CPPF()
     server.parser.add_option(mountopt="url", metavar="BASE_URL", default=None)
     server.parse(values=server, errex=1)

contrib/README.md

@@ -9,6 +9,9 @@
 * 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
 
+### [`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))
+
 # OS integration
 init-scripts to start copyparty as a service
 * [`systemd/copyparty.service`](systemd/copyparty.service)

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"

copyparty/__main__.py

@@ -9,6 +9,7 @@ __license__ = "MIT"
 __url__ = "https://github.com/9001/copyparty/"
 
 import os
+import time
 import shutil
 import filecmp
 import locale
@@ -85,6 +86,7 @@ def ensure_cert():
 
 
 def main():
+    time.strptime("19970815", "%Y%m%d")  # python#7980
     if WINDOWS:
         os.system("")  # enables colors
 
@@ -103,17 +105,22 @@ def main():
         epilog=dedent(
             """
             -a takes username:password,
-            -v takes src:dst:permset:permset:... where "permset" is
+            -v takes src:dst:permset:permset:cflag:cflag:...
-               accesslevel followed by username (no separator)
+               where "permset" is accesslevel followed by username (no separator)
+               and "cflag" is config flags to set on this volume
+            
+            list of cflags:
+              cnodupe rejects existing files (instead of symlinking them)
 
             example:\033[35m
-              -a ed:hunter2 -v .::r:aed -v ../inc:dump:w:aed  \033[36m
+              -a ed:hunter2 -v .::r:aed -v ../inc:dump:w:aed:cnodupe  \033[36m
               mount current directory at "/" with
                * r (read-only) for everyone
                * a (read+write) for ed
               mount ../inc at "/dump" with
                * w (write-only) for everyone
-               * a (read+write) for ed  \033[0m
+               * a (read+write) for ed
+               * reject duplicate files  \033[0m
             
             if no accounts or volumes are configured,
             current folder will be read/write for everyone
@@ -123,24 +130,26 @@ def main():
             """
         ),
     )
-    ap.add_argument(
+    # fmt: off
-        "-c", metavar="PATH", type=str, action="append", help="add config file"
+    ap.add_argument("-c", metavar="PATH", type=str, action="append", help="add config file")
-    )
     ap.add_argument("-i", metavar="IP", type=str, default="0.0.0.0", help="ip to bind")
     ap.add_argument("-p", metavar="PORT", type=int, default=3923, help="port to bind")
     ap.add_argument("-nc", metavar="NUM", type=int, default=64, help="max num clients")
-    ap.add_argument(
+    ap.add_argument("-j", metavar="CORES", type=int, default=1, help="max num cpu cores")
-        "-j", metavar="CORES", type=int, default=1, help="max num cpu cores"
-    )
     ap.add_argument("-a", metavar="ACCT", type=str, action="append", help="add account")
     ap.add_argument("-v", metavar="VOL", type=str, action="append", help="add volume")
     ap.add_argument("-q", action="store_true", help="quiet")
     ap.add_argument("-ed", action="store_true", help="enable ?dots")
     ap.add_argument("-emp", action="store_true", help="enable markdown plugins")
+    ap.add_argument("-e2d", action="store_true", help="enable up2k database")
+    ap.add_argument("-e2s", action="store_true", help="enable up2k db-scanner")
+    ap.add_argument("-mcr", metavar="SEC", type=int, default=60, help="md-editor mod-chk rate")
     ap.add_argument("-nw", action="store_true", help="disable writes (benchmark)")
     ap.add_argument("-nih", action="store_true", help="no info hostname")
     ap.add_argument("-nid", action="store_true", help="no info disk-usage")
+    ap.add_argument("--no-sendfile", action="store_true", help="disable sendfile")
     al = ap.parse_args()
+    # fmt: on
 
     SvcHub(al).run()
 

copyparty/__version__.py

@@ -1,8 +1,8 @@
 # coding: utf-8
 
-VERSION = (0, 5, 6)
+VERSION = (0, 7, 1)
-CODENAME = "fuse jelly"
+CODENAME = "keeping track"
-BUILD_DT = (2020, 11, 29)
+BUILD_DT = (2021, 1, 23)
 
 S_VERSION = ".".join(map(str, VERSION))
 S_BUILD_DT = "{0:04d}-{1:02d}-{2:02d}".format(*BUILD_DT)

copyparty/authsrv.py

@@ -12,11 +12,12 @@ from .util import undot, Pebkac, fsdec, fsenc
 class VFS(object):
     """single level in the virtual fs"""
 
-    def __init__(self, realpath, vpath, uread=[], uwrite=[]):
+    def __init__(self, realpath, vpath, uread=[], uwrite=[], flags={}):
         self.realpath = realpath  # absolute path on host filesystem
         self.vpath = vpath  # absolute path in the virtual filesystem
         self.uread = uread  # users who can read this
         self.uwrite = uwrite  # users who can write this
+        self.flags = flags  # config switches
         self.nodes = {}  # child nodes
 
     def add(self, src, dst):
@@ -36,6 +37,7 @@ class VFS(object):
                 "{}/{}".format(self.vpath, name).lstrip("/"),
                 self.uread,
                 self.uwrite,
+                self.flags,
             )
             self.nodes[name] = vn
             return vn.add(src, dst)
@@ -104,7 +106,7 @@ class VFS(object):
         real.sort()
         if not rem:
             for name, vn2 in sorted(self.nodes.items()):
-                if uname in vn2.uread:
+                if uname in vn2.uread or "*" in vn2.uread:
                     virt_vis[name] = vn2
 
             # no vfs nodes in the list of real inodes
@@ -161,7 +163,7 @@ class AuthSrv(object):
 
         yield prev, True
 
-    def _parse_config_file(self, fd, user, mread, mwrite, mount):
+    def _parse_config_file(self, fd, user, mread, mwrite, mflags, mount):
         vol_src = None
         vol_dst = None
         for ln in [x.decode("utf-8").strip() for x in fd]:
@@ -191,6 +193,7 @@ class AuthSrv(object):
                 mount[vol_dst] = vol_src
                 mread[vol_dst] = []
                 mwrite[vol_dst] = []
+                mflags[vol_dst] = {}
                 continue
 
             lvl, uname = ln.split(" ")
@@ -198,6 +201,9 @@ class AuthSrv(object):
                 mread[vol_dst].append(uname)
             if lvl in "wa":
                 mwrite[vol_dst].append(uname)
+            if lvl == "c":
+                # config option, currently switches only
+                mflags[vol_dst][uname] = True
 
     def reload(self):
         """
@@ -210,6 +216,7 @@ class AuthSrv(object):
         user = {}  # username:password
         mread = {}  # mountpoint:[username]
         mwrite = {}  # mountpoint:[username]
+        mflags = {}  # mountpoint:[flag]
         mount = {}  # dst:src (mountpoint:realpath)
 
         if self.args.a:
@@ -232,9 +239,13 @@ class AuthSrv(object):
                 mount[dst] = src
                 mread[dst] = []
                 mwrite[dst] = []
+                mflags[dst] = {}
 
                 perms = perms.split(":")
                 for (lvl, uname) in [[x[0], x[1:]] for x in perms]:
+                    if lvl == "c":
+                        # config option, currently switches only
+                        mflags[dst][uname] = True
                     if uname == "":
                         uname = "*"
                     if lvl in "ra":
@@ -245,14 +256,15 @@ class AuthSrv(object):
         if self.args.c:
             for cfg_fn in self.args.c:
                 with open(cfg_fn, "rb") as f:
-                    self._parse_config_file(f, user, mread, mwrite, mount)
+                    self._parse_config_file(f, user, mread, mwrite, mflags, mount)
 
+        self.all_writable = []
         if not mount:
             # -h says our defaults are CWD at root and read/write for everyone
             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(os.path.abspath("."), "")
 
         maxdepth = 0
         for dst in sorted(mount.keys(), key=lambda x: (x.count("/"), len(x))):
@@ -262,12 +274,18 @@ class AuthSrv(object):
 
             if dst == "":
                 # rootfs was mapped; fully replaces the default CWD vfs
-                vfs = VFS(mount[dst], dst, mread[dst], mwrite[dst])
+                vfs = VFS(mount[dst], dst, mread[dst], mwrite[dst], mflags[dst])
                 continue
 
             v = vfs.add(mount[dst], dst)
             v.uread = mread[dst]
             v.uwrite = mwrite[dst]
+            v.flags = mflags[dst]
+            if v.uwrite:
+                self.all_writable.append(v)
+
+        if vfs.uwrite and vfs not in self.all_writable:
+            self.all_writable.append(vfs)
 
         missing_users = {}
         for d in [mread, mwrite]:

copyparty/httpcli.py

@@ -28,6 +28,7 @@ class HttpCli(object):
         self.conn = conn
         self.s = conn.s
         self.sr = conn.sr
+        self.ip = conn.addr[0]
         self.addr = conn.addr
         self.args = conn.args
         self.auth = conn.auth
@@ -42,7 +43,7 @@ class HttpCli(object):
         self.log_func(self.log_src, msg)
 
     def _check_nonfatal(self, ex):
-        return ex.code in [404]
+        return ex.code < 400 or ex.code == 404
 
     def _assert_safe_rem(self, rem):
         # sanity check to prevent any disasters
@@ -85,7 +86,8 @@ class HttpCli(object):
 
         v = self.headers.get("x-forwarded-for", None)
         if v is not None and self.conn.addr[0] in ["127.0.0.1", "::1"]:
-            self.log_src = self.conn.set_rproxy(v.split(",")[0])
+            self.ip = v.split(",")[0]
+            self.log_src = self.conn.set_rproxy(self.ip)
 
         self.uname = "*"
         if "cookie" in self.headers:
@@ -305,7 +307,7 @@ class HttpCli(object):
         vfs, rem = self.conn.auth.vfs.get(self.vpath, self.uname, False, True)
         fdir = os.path.join(vfs.realpath, rem)
 
-        addr = self.conn.addr[0].replace(":", ".")
+        addr = self.ip.replace(":", ".")
         fn = "put-{:.6f}-{}.bin".format(time.time(), addr)
         path = os.path.join(fdir, fn)
 
@@ -384,9 +386,11 @@ class HttpCli(object):
 
         vfs, rem = self.conn.auth.vfs.get(self.vpath, self.uname, False, True)
 
-        body["vdir"] = self.vpath
+        body["vtop"] = vfs.vpath
-        body["rdir"] = os.path.join(vfs.realpath, rem)
+        body["ptop"] = vfs.realpath
-        body["addr"] = self.addr[0]
+        body["prel"] = rem
+        body["addr"] = self.ip
+        body["flag"] = vfs.flags
 
         x = self.conn.hsrv.broker.put(True, "up2k.handle_json", body)
         response = x.get()
@@ -408,7 +412,10 @@ class HttpCli(object):
         except KeyError:
             raise Pebkac(400, "need hash and wark headers for binary POST")
 
-        x = self.conn.hsrv.broker.put(True, "up2k.handle_chunk", wark, chash)
+        vfs, _ = self.conn.auth.vfs.get(self.vpath, self.uname, False, True)
+        ptop = vfs.realpath
+
+        x = self.conn.hsrv.broker.put(True, "up2k.handle_chunk", ptop, wark, chash)
         response = x.get()
         chunksize, cstart, path, lastmod = response
 
@@ -453,8 +460,8 @@ class HttpCli(object):
 
                 self.log("clone {} done".format(cstart[0]))
 
-        x = self.conn.hsrv.broker.put(True, "up2k.confirm_chunk", wark, chash)
+        x = self.conn.hsrv.broker.put(True, "up2k.confirm_chunk", ptop, wark, chash)
-        num_left = x.get()
+        num_left, path = x.get()
 
         if not WINDOWS and num_left == 0:
             times = (int(time.time()), int(lastmod))
@@ -568,24 +575,24 @@ class HttpCli(object):
                     self.log("discarding incoming file without filename")
                     # fallthrough
 
-                fn = os.devnull
                 if p_file and not nullwrite:
                     fdir = os.path.join(vfs.realpath, rem)
-                    fn = os.path.join(fdir, sanitize_fn(p_file))
+                    fname = sanitize_fn(p_file)
 
                     if not os.path.isdir(fsenc(fdir)):
                         raise Pebkac(404, "that folder does not exist")
 
-                    # TODO broker which avoid this race and
+                    suffix = ".{:.6f}-{}".format(time.time(), self.ip)
-                    # provides a new filename if taken (same as up2k)
+                    open_args = {"fdir": fdir, "suffix": suffix}
-                    if os.path.exists(fsenc(fn)):
+                else:
-                        fn += ".{:.6f}-{}".format(time.time(), self.addr[0])
+                    open_args = {}
-                        # using current-time instead of t0 cause clients
+                    fname = os.devnull
-                        # may reuse a name for multiple files in one post
+                    fdir = ""
 
                 try:
-                    with open(fsenc(fn), "wb") as f:
+                    with ren_open(fname, "wb", 512 * 1024, **open_args) as f:
-                        self.log("writing to {0}".format(fn))
+                        f, fname = f["orz"]
+                        self.log("writing to {}/{}".format(fdir, fname))
                         sz, sha512_hex, _ = hashcopy(self.conn, p_data, f)
                         if sz == 0:
                             raise Pebkac(400, "empty files in post")
@@ -594,8 +601,14 @@ class HttpCli(object):
                         self.conn.nbyte += sz
 
                 except Pebkac:
-                    if fn != os.devnull:
+                    if fname != os.devnull:
-                        os.rename(fsenc(fn), fsenc(fn + ".PARTIAL"))
+                        fp = os.path.join(fdir, fname)
+                        suffix = ".PARTIAL"
+                        try:
+                            os.rename(fsenc(fp), fsenc(fp + suffix))
+                        except:
+                            fp = fp[: -len(suffix)]
+                            os.rename(fsenc(fp), fsenc(fp + suffix))
 
                     raise
 
@@ -631,7 +644,7 @@ class HttpCli(object):
                         "\n".join(
                             unicode(x)
                             for x in [
-                                ":".join(unicode(x) for x in self.addr),
+                                ":".join(unicode(x) for x in [self.ip, self.addr[1]]),
                                 msg.rstrip(),
                             ]
                         )
@@ -680,7 +693,7 @@ class HttpCli(object):
             return True
 
         fp = os.path.join(vfs.realpath, rem)
-        srv_lastmod = -1
+        srv_lastmod = srv_lastmod3 = -1
         try:
             st = os.stat(fsenc(fp))
             srv_lastmod = st.st_mtime
@@ -731,7 +744,7 @@ class HttpCli(object):
         if p_field != "body":
             raise Pebkac(400, "expected body, got {}".format(p_field))
 
-        with open(fp, "wb") as f:
+        with open(fp, "wb", 512 * 1024) as f:
             sz, sha512, _ = hashcopy(self.conn, p_data, f)
 
         new_lastmod = os.stat(fsenc(fp)).st_mtime
@@ -756,9 +769,12 @@ class HttpCli(object):
                 cli_dt = time.strptime(cli_lastmod, "%a, %d %b %Y %H:%M:%S GMT")
                 cli_ts = calendar.timegm(cli_dt)
                 return file_lastmod, int(file_ts) > int(cli_ts)
-            except:
+            except Exception as ex:
-                self.log("bad lastmod format: {}".format(cli_lastmod))
+                self.log(
-                self.log("   expected format: {}".format(file_lastmod))
+                    "lastmod {}\nremote: [{}]\n local: [{}]".format(
+                        repr(ex), cli_lastmod, file_lastmod
+                    )
+                )
                 return file_lastmod, file_lastmod != cli_lastmod
 
         return file_lastmod, True
@@ -875,6 +891,7 @@ class HttpCli(object):
 
             logtail += " [\033[36m{}-{}\033[0m]".format(lower, upper)
 
+        use_sendfile = False
         if decompress:
             open_func = gzip.open
             open_args = [fsenc(fs_path), "rb"]
@@ -884,6 +901,8 @@ class HttpCli(object):
             open_func = open
             # 512 kB is optimal for huge files, use 64k
             open_args = [fsenc(fs_path), "rb", 64 * 1024]
+            if hasattr(os, "sendfile"):
+                use_sendfile = not self.args.no_sendfile
 
         #
         # send reply
@@ -906,24 +925,13 @@ class HttpCli(object):
 
         ret = True
         with open_func(*open_args) as f:
-            remains = upper - lower
+            if use_sendfile:
-            f.seek(lower)
+                remains = sendfile_kern(lower, upper, f, self.s)
-            while remains > 0:
+            else:
-                # time.sleep(0.01)
+                remains = sendfile_py(lower, upper, f, self.s)
-                buf = f.read(4096)
-                if not buf:
-                    break
 
-                if remains < len(buf):
+        if remains > 0:
-                    buf = buf[:remains]
-
-                try:
-                    self.s.sendall(buf)
-                    remains -= len(buf)
-                except:
             logmsg += " \033[31m" + str(upper - remains) + "\033[0m"
-                    ret = False
-                    break
 
         spd = self._spd((upper - lower) - remains)
         self.log("{},  {}".format(logmsg, spd))
@@ -964,6 +972,7 @@ class HttpCli(object):
             "title": html_escape(self.vpath),
             "lastmod": int(ts_md * 1000),
             "md_plug": "true" if self.args.emp else "false",
+            "md_chk_rate": self.args.mcr,
             "md": "",
         }
         sz_html = len(template.render(**targs).encode("utf-8"))
@@ -1018,6 +1027,10 @@ class HttpCli(object):
             if abspath.endswith(".md") and "raw" not in self.uparam:
                 return self.tx_md(abspath)
 
+            bad = "{0}.hist{0}up2k.".format(os.sep)
+            if abspath.endswith(bad + "db") or abspath.endswith(bad + "snap"):
+                raise Pebkac(403)
+
             return self.tx_file(abspath)
 
         fsroot, vfs_ls, vfs_virt = vn.ls(rem, self.uname)

copyparty/httpconn.py

@@ -65,6 +65,7 @@ class HttpConn(object):
             color = 34
             self.rproxy = ip
 
+        self.ip = ip
         self.log_src = "{} \033[{}m{}".format(ip, color, self.addr[1]).ljust(26)
         return self.log_src
 

copyparty/svchub.py

@@ -9,6 +9,7 @@ from datetime import datetime, timedelta
 import calendar
 
 from .__init__ import PY2, WINDOWS, MACOS, VT100
+from .authsrv import AuthSrv
 from .tcpsrv import TcpSrv
 from .up2k import Up2k
 from .util import mp
@@ -38,6 +39,10 @@ class SvcHub(object):
         self.tcpsrv = TcpSrv(self)
         self.up2k = Up2k(self)
 
+        if self.args.e2d and self.args.e2s:
+            auth = AuthSrv(self.args, self.log)
+            self.up2k.build_indexes(auth.all_writable)
+
         # decide which worker impl to use
         if self.check_mp_enable():
             from .broker_mp import BrokerMp as Broker

copyparty/tcpsrv.py

@@ -36,6 +36,7 @@ class TcpSrv(object):
 
         self.srv = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
         self.srv.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+        self.srv.setsockopt(socket.IPPROTO_TCP, socket.TCP_NODELAY, 1)
         try:
             self.srv.bind((self.args.i, self.args.p))
         except (OSError, socket.error) as ex:

copyparty/up2k.py

@@ -6,14 +6,25 @@ import os
 import re
 import time
 import math
+import json
+import gzip
+import stat
 import shutil
 import base64
 import hashlib
 import threading
 from copy import deepcopy
 
-from .__init__ import WINDOWS
+from .__init__ import WINDOWS, PY2
-from .util import Pebkac, Queue, fsenc, sanitize_fn
+from .util import Pebkac, Queue, fsdec, fsenc, sanitize_fn, ren_open, atomic_move
+
+HAVE_SQLITE3 = False
+try:
+    import sqlite3
+
+    HAVE_SQLITE3 = True
+except:
+    pass
 
 
 class Up2k(object):
@@ -22,20 +33,21 @@ class Up2k(object):
       * documentation
       * registry persistence
         * ~/.config flatfiles for active jobs
-        * wark->path database for finished uploads
     """
 
     def __init__(self, broker):
         self.broker = broker
         self.args = broker.args
         self.log = broker.log
+        self.persist = self.args.e2d
 
         # config
         self.salt = "hunter2"  # TODO: config
 
         # state
-        self.registry = {}
         self.mutex = threading.Lock()
+        self.registry = {}
+        self.db = {}
 
         if WINDOWS:
             # usually fails to set lastmod too quickly
@@ -44,54 +56,301 @@ class Up2k(object):
             thr.daemon = True
             thr.start()
 
+        if self.persist:
+            thr = threading.Thread(target=self._snapshot)
+            thr.daemon = True
+            thr.start()
+
         # static
         self.r_hash = re.compile("^[0-9a-zA-Z_-]{43}$")
 
+        if self.persist and not HAVE_SQLITE3:
+            m = "could not initialize sqlite3, will use in-memory registry only"
+            self.log("up2k", m)
+
+    def _vis_job_progress(self, job):
+        perc = 100 - (len(job["need"]) * 100.0 / len(job["hash"]))
+        path = os.path.join(job["ptop"], job["prel"], job["name"])
+        return "{:5.1f}% {}".format(perc, path)
+
+    def _vis_reg_progress(self, reg):
+        ret = []
+        for _, job in reg.items():
+            ret.append(self._vis_job_progress(job))
+
+        return ret
+
+    def register_vpath(self, ptop):
+        with self.mutex:
+            if ptop in self.registry:
+                return None
+
+            reg = {}
+            path = os.path.join(ptop, ".hist", "up2k.snap")
+            if self.persist and os.path.exists(path):
+                with gzip.GzipFile(path, "rb") as f:
+                    j = f.read().decode("utf-8")
+
+                reg = json.loads(j)
+                for _, job in reg.items():
+                    job["poke"] = time.time()
+
+                m = "loaded snap {} |{}|".format(path, len(reg.keys()))
+                m = [m] + self._vis_reg_progress(reg)
+                self.log("up2k", "\n".join(m))
+
+            self.registry[ptop] = reg
+            if not self.persist or not HAVE_SQLITE3:
+                return None
+
+            try:
+                os.mkdir(os.path.join(ptop, ".hist"))
+            except:
+                pass
+
+            db_path = os.path.join(ptop, ".hist", "up2k.db")
+            if ptop in self.db:
+                # self.db[ptop].close()
+                return None
+
+            try:
+                db = self._open_db(db_path)
+                self.db[ptop] = db
+                return db
+            except Exception as ex:
+                m = "failed to open [{}]: {}".format(ptop, repr(ex))
+                self.log("up2k", m)
+
+            return None
+
+    def build_indexes(self, writeables):
+        tops = [d.realpath for d in writeables]
+        for top in tops:
+            db = self.register_vpath(top)
+            if db:
+                # can be symlink so don't `and d.startswith(top)``
+                excl = set([d for d in tops if d != top])
+                self._build_dir([db, 0], top, excl, top)
+                self._drop_lost(db, top)
+                db.commit()
+
+    def _build_dir(self, dbw, top, excl, cdir):
+        try:
+            inodes = [fsdec(x) for x in os.listdir(fsenc(cdir))]
+        except Exception as ex:
+            self.log("up2k", "listdir: " + repr(ex))
+            return
+
+        histdir = os.path.join(top, ".hist")
+        for inode in inodes:
+            abspath = os.path.join(cdir, inode)
+            try:
+                inf = os.stat(fsenc(abspath))
+            except Exception as ex:
+                self.log("up2k", "stat: " + repr(ex))
+                continue
+
+            if stat.S_ISDIR(inf.st_mode):
+                if abspath in excl or abspath == histdir:
+                    continue
+                # self.log("up2k", " dir: {}".format(abspath))
+                self._build_dir(dbw, top, excl, abspath)
+            else:
+                # self.log("up2k", "file: {}".format(abspath))
+                rp = abspath[len(top) :].replace("\\", "/").strip("/")
+                c = dbw[0].execute("select * from up where rp = ?", (rp,))
+                in_db = list(c.fetchall())
+                if in_db:
+                    _, dts, dsz, _ = in_db[0]
+                    if len(in_db) > 1:
+                        m = "WARN: multiple entries: [{}] => [{}] ({})"
+                        self.log("up2k", m.format(top, rp, len(in_db)))
+                        dts = -1
+
+                    if dts == inf.st_mtime and dsz == inf.st_size:
+                        continue
+
+                    m = "reindex [{}] => [{}] ({}/{}) ({}/{})".format(
+                        top, rp, dts, inf.st_mtime, dsz, inf.st_size
+                    )
+                    self.log("up2k", m)
+                    self.db_rm(dbw[0], rp)
+                    dbw[1] += 1
+                    in_db = None
+
+                self.log("up2k", "file: {}".format(abspath))
+                try:
+                    hashes = self._hashlist_from_file(abspath)
+                except Exception as ex:
+                    self.log("up2k", "hash: " + repr(ex))
+                    continue
+
+                wark = self._wark_from_hashlist(inf.st_size, hashes)
+                self.db_add(dbw[0], wark, rp, inf.st_mtime, inf.st_size)
+                dbw[1] += 1
+                if dbw[1] > 1024:
+                    dbw[0].commit()
+                    dbw[1] = 0
+
+    def _drop_lost(self, db, top):
+        rm = []
+        c = db.execute("select * from up")
+        for dwark, dts, dsz, drp in c:
+            abspath = os.path.join(top, drp)
+            try:
+                if not os.path.exists(fsenc(abspath)):
+                    rm.append(drp)
+            except Exception as ex:
+                self.log("up2k", "stat-rm: " + repr(ex))
+
+        if not rm:
+            return
+
+        self.log("up2k", "forgetting {} deleted files".format(len(rm)))
+        for rp in rm:
+            self.db_rm(db, rp)
+
+    def _open_db(self, db_path):
+        conn = sqlite3.connect(db_path, check_same_thread=False)
+        try:
+            c = conn.execute(r"select * from kv where k = 'sver'")
+            rows = c.fetchall()
+            if rows:
+                ver = rows[0][1]
+            else:
+                self.log("up2k", "WARN: no sver in kv, DB corrupt?")
+                ver = "unknown"
+
+            if ver == "1":
+                try:
+                    nfiles = next(conn.execute("select count(w) from up"))[0]
+                    self.log("up2k", "found DB at {} |{}|".format(db_path, nfiles))
+                    return conn
+                except Exception as ex:
+                    m = "WARN: could not list files, DB corrupt?\n  " + repr(ex)
+                    self.log("up2k", m)
+
+            m = "REPLACING unsupported DB (v.{}) at {}".format(ver, db_path)
+            self.log("up2k", m)
+            conn.close()
+            os.unlink(db_path)
+            conn = sqlite3.connect(db_path, check_same_thread=False)
+        except:
+            pass
+
+        # sqlite is variable-width only, no point in using char/nchar/varchar
+        for cmd in [
+            r"create table kv (k text, v text)",
+            r"create table up (w text, mt int, sz int, rp text)",
+            r"insert into kv values ('sver', '1')",
+            r"create index up_w on up(w)",
+        ]:
+            conn.execute(cmd)
+
+        conn.commit()
+        self.log("up2k", "created DB at {}".format(db_path))
+        return conn
+
     def handle_json(self, cj):
+        self.register_vpath(cj["ptop"])
         cj["name"] = sanitize_fn(cj["name"])
+        cj["poke"] = time.time()
         wark = self._get_wark(cj)
         now = time.time()
+        job = None
         with self.mutex:
-            # TODO use registry persistence here to symlink any matching wark
+            db = self.db.get(cj["ptop"], None)
-            if wark in self.registry:
+            reg = self.registry[cj["ptop"]]
-                job = self.registry[wark]
+            if db:
-                if job["rdir"] != cj["rdir"] or job["name"] != cj["name"]:
+                cur = db.execute(r"select * from up where w = ?", (wark,))
-                    src = os.path.join(job["rdir"], job["name"])
+                for _, dtime, dsize, dp_rel in cur:
-                    dst = os.path.join(cj["rdir"], cj["name"])
+                    dp_abs = os.path.join(cj["ptop"], dp_rel).replace("\\", "/")
+                    # relying on path.exists to return false on broken symlinks
+                    if os.path.exists(fsenc(dp_abs)):
+                        try:
+                            prel, name = dp_rel.rsplit("/", 1)
+                        except:
+                            prel = ""
+                            name = dp_rel
+
+                        job = {
+                            "name": name,
+                            "prel": prel,
+                            "vtop": cj["vtop"],
+                            "ptop": cj["ptop"],
+                            "flag": cj["flag"],
+                            "size": dsize,
+                            "lmod": dtime,
+                            "hash": [],
+                            "need": [],
+                        }
+                        break
+
+                if job and wark in reg:
+                    del reg[wark]
+
+            if job or wark in reg:
+                job = job or reg[wark]
+                if job["prel"] == cj["prel"] and job["name"] == cj["name"]:
+                    # ensure the files haven't been deleted manually
+                    names = [job[x] for x in ["name", "tnam"] if x in job]
+                    for fn in names:
+                        path = os.path.join(job["ptop"], job["prel"], fn)
+                        if not os.path.exists(path):
+                            job = None
+                            break
+                else:
+                    # file contents match, but not the path
+                    src = os.path.join(job["ptop"], job["prel"], job["name"])
+                    dst = os.path.join(cj["ptop"], cj["prel"], cj["name"])
+                    vsrc = os.path.join(job["vtop"], job["prel"], job["name"])
+                    vsrc = vsrc.replace("\\", "/")  # just for prints anyways
                     if job["need"]:
                         self.log("up2k", "unfinished:\n  {0}\n  {1}".format(src, dst))
-                        err = "partial upload exists at a different location; please resume uploading here instead:\n{0}{1} ".format(
+                        err = "partial upload exists at a different location; please resume uploading here instead:\n"
-                            job["vdir"], job["name"]
+                        err += vsrc + " "
-                        )
+                        raise Pebkac(400, err)
+                    elif "nodupe" in job["flag"]:
+                        self.log("up2k", "dupe-reject:\n  {0}\n  {1}".format(src, dst))
+                        err = "upload rejected, file already exists:\n " + vsrc + " "
                         raise Pebkac(400, err)
                     else:
                         # symlink to the client-provided name,
                         # returning the previous upload info
                         job = deepcopy(job)
-                        suffix = self._suffix(dst, now, job["addr"])
+                        for k in ["ptop", "vtop", "prel"]:
-                        job["name"] = cj["name"] + suffix
+                            job[k] = cj[k]
-                        self._symlink(src, dst + suffix)
+
-            else:
+                        pdir = os.path.join(cj["ptop"], cj["prel"])
+                        job["name"] = self._untaken(pdir, cj["name"], now, cj["addr"])
+                        dst = os.path.join(job["ptop"], job["prel"], job["name"])
+                        os.unlink(fsenc(dst))  # TODO ed pls
+                        self._symlink(src, dst)
+
+            if not job:
                 job = {
                     "wark": wark,
                     "t0": now,
-                    "addr": cj["addr"],
-                    "vdir": cj["vdir"],
-                    "rdir": cj["rdir"],
-                    # client-provided, sanitized by _get_wark:
-                    "name": cj["name"],
-                    "size": cj["size"],
-                    "lmod": cj["lmod"],
                     "hash": deepcopy(cj["hash"]),
+                    "need": [],
                 }
-
+                # client-provided, sanitized by _get_wark: name, size, lmod
-                path = os.path.join(job["rdir"], job["name"])
+                for k in [
-                job["name"] += self._suffix(path, now, cj["addr"])
+                    "addr",
+                    "vtop",
+                    "ptop",
+                    "prel",
+                    "flag",
+                    "name",
+                    "size",
+                    "lmod",
+                    "poke",
+                ]:
+                    job[k] = cj[k]
 
                 # one chunk may occur multiple times in a file;
                 # filter to unique values for the list of missing chunks
                 # (preserve order to reduce disk thrashing)
-                job["need"] = []
                 lut = {}
                 for k in cj["hash"]:
                     if k not in lut:
@@ -108,13 +367,12 @@ class Up2k(object):
                 "wark": wark,
             }
 
-    def _suffix(self, fpath, ts, ip):
+    def _untaken(self, fdir, fname, ts, ip):
         # TODO broker which avoid this race and
         # provides a new filename if taken (same as bup)
-        if not os.path.exists(fsenc(fpath)):
+        suffix = ".{:.6f}-{}".format(ts, ip)
-            return ""
+        with ren_open(fname, "wb", fdir=fdir, suffix=suffix) as f:
-
+            return f["orz"][1]
-        return ".{:.6f}-{}".format(ts, ip)
 
     def _symlink(self, src, dst):
         # TODO store this in linktab so we never delete src if there are links to it
@@ -141,40 +399,58 @@ class Up2k(object):
                     lsrc = "../" * (len(lsrc) - 1) + "/".join(lsrc)
             os.symlink(fsenc(lsrc), fsenc(ldst))
         except (AttributeError, OSError) as ex:
-            self.log("up2k", "cannot symlink; creating copy")
+            self.log("up2k", "cannot symlink; creating copy: " + repr(ex))
             shutil.copy2(fsenc(src), fsenc(dst))
 
-    def handle_chunk(self, wark, chash):
+    def handle_chunk(self, ptop, wark, chash):
         with self.mutex:
-            job = self.registry.get(wark)
+            job = self.registry[ptop].get(wark, None)
             if not job:
-                raise Pebkac(404, "unknown wark")
+                raise Pebkac(400, "unknown wark")
 
             if chash not in job["need"]:
                 raise Pebkac(200, "already got that but thanks??")
 
             nchunk = [n for n, v in enumerate(job["hash"]) if v == chash]
             if not nchunk:
-                raise Pebkac(404, "unknown chunk")
+                raise Pebkac(400, "unknown chunk")
+
+        job["poke"] = time.time()
 
         chunksize = self._get_chunksize(job["size"])
         ofs = [chunksize * x for x in nchunk]
 
-        path = os.path.join(job["rdir"], job["name"])
+        path = os.path.join(job["ptop"], job["prel"], job["tnam"])
 
         return [chunksize, ofs, path, job["lmod"]]
 
-    def confirm_chunk(self, wark, chash):
+    def confirm_chunk(self, ptop, wark, chash):
         with self.mutex:
-            job = self.registry[wark]
+            job = self.registry[ptop][wark]
+            pdir = os.path.join(job["ptop"], job["prel"])
+            src = os.path.join(pdir, job["tnam"])
+            dst = os.path.join(pdir, job["name"])
+
             job["need"].remove(chash)
             ret = len(job["need"])
+            if ret > 0:
+                return ret, src
 
-            if WINDOWS and ret == 0:
+            atomic_move(src, dst)
-                path = os.path.join(job["rdir"], job["name"])
-                self.lastmod_q.put([path, (int(time.time()), int(job["lmod"]))])
 
-            return ret
+            if WINDOWS:
+                self.lastmod_q.put([dst, (int(time.time()), int(job["lmod"]))])
+
+            db = self.db.get(job["ptop"], None)
+            if db:
+                rp = os.path.join(job["prel"], job["name"]).replace("\\", "/")
+                self.db_rm(db, rp)
+                self.db_add(db, job["wark"], rp, job["lmod"], job["size"])
+                db.commit()
+                del self.registry[ptop][wark]
+                # in-memory registry is reserved for unfinished uploads
+
+            return ret, dst
 
     def _get_chunksize(self, filesize):
         chunksize = 1024 * 1024
@@ -188,6 +464,13 @@ class Up2k(object):
                 chunksize += stepsize
                 stepsize *= mul
 
+    def db_rm(self, db, rp):
+        db.execute("delete from up where rp = ?", (rp,))
+
+    def db_add(self, db, wark, rp, ts, sz):
+        v = (wark, ts, sz, rp)
+        db.execute("insert into up values (?,?,?,?)", v)
+
     def _get_wark(self, cj):
         if len(cj["name"]) > 1024 or len(cj["hash"]) > 512 * 1024:  # 16TiB
             raise Pebkac(400, "name or numchunks not according to spec")
@@ -204,9 +487,13 @@ class Up2k(object):
         except:
             cj["lmod"] = int(time.time())
 
-        # server-reproducible file identifier, independent of name or location
+        wark = self._wark_from_hashlist(cj["size"], cj["hash"])
-        ident = [self.salt, str(cj["size"])]
+        return wark
-        ident.extend(cj["hash"])
+
+    def _wark_from_hashlist(self, filesize, hashes):
+        """ server-reproducible file identifier, independent of name or location """
+        ident = [self.salt, str(filesize)]
+        ident.extend(hashes)
         ident = "\n".join(ident)
 
         hasher = hashlib.sha512()
@@ -216,10 +503,38 @@ class Up2k(object):
         wark = base64.urlsafe_b64encode(digest)
         return wark.decode("utf-8").rstrip("=")
 
+    def _hashlist_from_file(self, path):
+        fsz = os.path.getsize(path)
+        csz = self._get_chunksize(fsz)
+        ret = []
+        with open(path, "rb", 512 * 1024) as f:
+            while fsz > 0:
+                hashobj = hashlib.sha512()
+                rem = min(csz, fsz)
+                fsz -= rem
+                while rem > 0:
+                    buf = f.read(min(rem, 64 * 1024))
+                    if not buf:
+                        raise Exception("EOF at " + str(f.tell()))
+
+                    hashobj.update(buf)
+                    rem -= len(buf)
+
+                digest = hashobj.digest()[:32]
+                digest = base64.urlsafe_b64encode(digest)
+                ret.append(digest.decode("utf-8").rstrip("="))
+
+        return ret
+
     def _new_upload(self, job):
-        self.registry[job["wark"]] = job
+        self.registry[job["ptop"]][job["wark"]] = job
-        path = os.path.join(job["rdir"], job["name"])
+        pdir = os.path.join(job["ptop"], job["prel"])
-        with open(fsenc(path), "wb") as f:
+        job["name"] = self._untaken(pdir, job["name"], job["t0"], job["addr"])
+
+        tnam = job["name"] + ".PARTIAL"
+        suffix = ".{:.6f}-{}".format(job["t0"], job["addr"])
+        with ren_open(tnam, "wb", fdir=pdir, suffix=suffix) as f:
+            f, job["tnam"] = f["orz"]
             f.seek(job["size"] - 1)
             f.write(b"e")
 
@@ -236,3 +551,58 @@ class Up2k(object):
                     os.utime(fsenc(path), times)
                 except:
                     self.log("lmod", "failed to utime ({}, {})".format(path, times))
+
+    def _snapshot(self):
+        persist_interval = 30  # persist unfinished uploads index every 30 sec
+        discard_interval = 3600  # drop unfinished uploads after 1 hour inactivity
+        prev = {}
+        while True:
+            time.sleep(persist_interval)
+            with self.mutex:
+                for k, reg in self.registry.items():
+                    self._snap_reg(prev, k, reg, discard_interval)
+
+    def _snap_reg(self, prev, k, reg, discard_interval):
+        now = time.time()
+        rm = [x for x in reg.values() if now - x["poke"] > discard_interval]
+        if rm:
+            m = "dropping {} abandoned uploads in {}".format(len(rm), k)
+            vis = [self._vis_job_progress(x) for x in rm]
+            self.log("up2k", "\n".join([m] + vis))
+            for job in rm:
+                del reg[job["wark"]]
+                try:
+                    # remove the filename reservation
+                    path = os.path.join(job["ptop"], job["prel"], job["name"])
+                    if os.path.getsize(path) == 0:
+                        os.unlink(path)
+
+                    if len(job["hash"]) == len(job["need"]):
+                        # PARTIAL is empty, delete that too
+                        path = os.path.join(job["ptop"], job["prel"], job["tnam"])
+                        os.unlink(path)
+                except:
+                    pass
+
+        path = os.path.join(k, ".hist", "up2k.snap")
+        if not reg:
+            if k not in prev or prev[k] is not None:
+                prev[k] = None
+                if os.path.exists(path):
+                    os.unlink(path)
+            return
+
+        newest = max(x["poke"] for _, x in reg.items()) if reg else 0
+        etag = [len(reg), newest]
+        if etag == prev.get(k, None):
+            return
+
+        path2 = "{}.{}".format(path, os.getpid())
+        j = json.dumps(reg, indent=2, sort_keys=True).encode("utf-8")
+        with gzip.GzipFile(path2, "wb") as f:
+            f.write(j)
+
+        atomic_move(path2, path)
+
+        self.log("up2k", "snap: {} |{}|".format(path, len(reg.keys())))
+        prev[k] = etag

copyparty/util.py

@@ -2,14 +2,17 @@
 from __future__ import print_function, unicode_literals
 
 import re
+import os
 import sys
 import time
 import base64
+import select
 import struct
 import hashlib
 import platform
 import threading
 import mimetypes
+import contextlib
 import subprocess as sp  # nosec
 
 from .__init__ import PY2, WINDOWS
@@ -96,6 +99,80 @@ class Unrecv(object):
         self.buf = buf + self.buf
 
 
+@contextlib.contextmanager
+def ren_open(fname, *args, **kwargs):
+    fdir = kwargs.pop("fdir", None)
+    suffix = kwargs.pop("suffix", None)
+
+    if fname == os.devnull:
+        with open(fname, *args, **kwargs) as f:
+            yield {"orz": [f, fname]}
+            return
+    
+    orig_name = fname
+    bname = fname
+    ext = ""
+    while True:
+        ofs = bname.rfind(".")
+        if ofs < 0 or ofs < len(bname) - 7:
+            # doesn't look like an extension anymore
+            break
+
+        ext = bname[ofs:] + ext
+        bname = bname[:ofs]
+
+    b64 = ""
+    while True:
+        try:
+            if fdir:
+                fpath = os.path.join(fdir, fname)
+            else:
+                fpath = fname
+
+            if suffix and os.path.exists(fpath):
+                fpath += suffix
+                fname += suffix
+                ext += suffix
+
+            with open(fsenc(fpath), *args, **kwargs) as f:
+                if b64:
+                    fp2 = "fn-trunc.{}.txt".format(b64)
+                    fp2 = os.path.join(fdir, fp2)
+                    with open(fsenc(fp2), "wb") as f2:
+                        f2.write(orig_name.encode("utf-8"))
+
+                yield {"orz": [f, fname]}
+                return
+
+        except OSError as ex_:
+            ex = ex_
+            if ex.errno != 36:
+                raise
+
+        if not b64:
+            b64 = (bname + ext).encode("utf-8", "replace")
+            b64 = hashlib.sha512(b64).digest()[:12]
+            b64 = base64.urlsafe_b64encode(b64).decode("utf-8").rstrip("=")
+
+        badlen = len(fname)
+        while len(fname) >= badlen:
+            if len(bname) < 8:
+                raise ex
+
+            if len(bname) > len(ext):
+                # drop the last letter of the filename
+                bname = bname[:-1]
+            else:
+                try:
+                    # drop the leftmost sub-extension
+                    _, ext = ext.split(".", 1)
+                except:
+                    # okay do the first letter then
+                    ext = "." + ext[2:]
+
+            fname = "{}~{}{}".format(bname, b64, ext)
+
+
 class MultipartParser(object):
     def __init__(self, log_func, sr, http_headers):
         self.sr = sr
@@ -472,6 +549,16 @@ else:
     fsdec = w8dec
 
 
+def atomic_move(src, dst):
+    if not PY2:
+        os.replace(src, dst)
+    else:
+        if os.path.exists(dst):
+            os.unlink(dst)
+
+        os.rename(src, dst)
+
+
 def read_socket(sr, total_size):
     remains = total_size
     while remains > 0:
@@ -515,6 +602,46 @@ def hashcopy(actor, fin, fout):
     return tlen, hashobj.hexdigest(), digest_b64
 
 
+def sendfile_py(lower, upper, f, s):
+    remains = upper - lower
+    f.seek(lower)
+    while remains > 0:
+        # time.sleep(0.01)
+        buf = f.read(min(4096, remains))
+        if not buf:
+            return remains
+
+        try:
+            s.sendall(buf)
+            remains -= len(buf)
+        except:
+            return remains
+
+    return 0
+
+
+def sendfile_kern(lower, upper, f, s):
+    out_fd = s.fileno()
+    in_fd = f.fileno()
+    ofs = lower
+    while ofs < upper:
+        try:
+            req = min(2 ** 30, upper - ofs)
+            select.select([], [out_fd], [], 10)
+            n = os.sendfile(out_fd, in_fd, ofs, req)
+        except Exception as ex:
+            # print("sendfile: " + repr(ex))
+            n = 0
+        
+        if n <= 0:
+            return upper - ofs
+        
+        ofs += n
+        # print("sendfile: ok, sent {} now, {} total, {} remains".format(n, ofs - lower, upper - ofs))
+
+    return 0
+
+
 def unescape_cookie(orig):
     # mw=idk; doot=qwe%2Crty%3Basd+fgh%2Bjkl%25zxc%26vbn  # qwe,rty;asd fgh+jkl%zxc&vbn
     ret = ""
@@ -595,3 +722,6 @@ class Pebkac(Exception):
     def __init__(self, code, msg=None):
         super(Pebkac, self).__init__(msg or HTTPCODE[code])
         self.code = code
+
+    def __repr__(self):
+        return "Pebkac({}, {})".format(self.code, repr(self.args))

copyparty/web/Makefile

@@ -0,0 +1,12 @@
+# run me to zopfli all the static files
+# which should help on really slow connections
+# but then why are you using copyparty in the first place
+
+pk: $(addsuffix .gz, $(wildcard *.js *.css))
+un: $(addsuffix .un, $(wildcard *.gz))
+
+%.gz: %
+	pigz -11 -J 34 -I 5730 $<
+
+%.un: %
+	pigz -d $<

copyparty/web/browser.js

@@ -8,7 +8,14 @@ function dbg(msg) {
 
 function ev(e) {
 	e = e || window.event;
-	e.preventDefault ? e.preventDefault() : (e.returnValue = false);
+
+	if (e.preventDefault)
+		e.preventDefault()
+
+	if (e.stopPropagation)
+		e.stopPropagation();
+
+	e.returnValue = false;
 	return e;
 }
 
@@ -25,9 +32,9 @@ var mp = (function () {
 		'tracks': tracks,
 		'cover_url': ''
 	};
-	var re_audio = new RegExp('\.(opus|ogg|m4a|aac|mp3|wav|flac)$', 'i');
+	var re_audio = /\.(opus|ogg|m4a|aac|mp3|wav|flac)$/i;
 
-	var trs = document.getElementById('files').getElementsByTagName('tbody')[0].getElementsByTagName('tr');
+	var trs = ebi('files').getElementsByTagName('tbody')[0].getElementsByTagName('tr');
 	for (var a = 0, aa = trs.length; a < aa; a++) {
 		var tds = trs[a].getElementsByTagName('td');
 		var link = tds[1].getElementsByTagName('a')[0];
@@ -70,8 +77,8 @@ var mp = (function () {
 // toggle player widget
 var widget = (function () {
 	var ret = {};
-	var widget = document.getElementById('widget');
+	var widget = ebi('widget');
-	var wtoggle = document.getElementById('wtoggle');
+	var wtoggle = ebi('wtoggle');
 	var touchmode = false;
 	var side_open = false;
 	var was_paused = true;
@@ -315,8 +322,12 @@ var vbar = (function () {
 		var rect = pbar.pcan.getBoundingClientRect();
 		var x = e.clientX - rect.left;
 		var mul = x * 1.0 / rect.width;
+		var seek = mp.au.duration * mul;
+		console.log('seek: ' + seek);
+		if (!isFinite(seek))
+			return;
 
-		mp.au.currentTime = mp.au.duration * mul;
+		mp.au.currentTime = seek;
 
 		if (mp.au === mp.au_native)
 			// hack: ogv.js breaks on .play() during playback
@@ -443,7 +454,8 @@ function play(tid, call_depth) {
 	mp.au.tid = tid;
 	mp.au.src = url;
 	mp.au.volume = mp.expvol();
-	setclass('trk' + tid, 'play act');
+	var oid = 'trk' + tid;
+	setclass(oid, 'play act');
 
 	try {
 		if (hack_attempt_play)
@@ -452,7 +464,11 @@ function play(tid, call_depth) {
 		if (mp.au.paused)
 			autoplay_blocked();
 
-		location.hash = 'trk' + tid;
+		var o = ebi(oid);
+		o.setAttribute('id', 'thx_js');
+		location.hash = oid;
+		o.setAttribute('id', oid);
+
 		pbar.drawbuf();
 		return true;
 	}
@@ -468,7 +484,6 @@ function play(tid, call_depth) {
 function evau_error(e) {
 	var err = '';
 	var eplaya = (e && e.target) || (window.event && window.event.srcElement);
-	var url = eplaya.src;
 
 	switch (eplaya.error.code) {
 		case eplaya.error.MEDIA_ERR_ABORTED:
@@ -516,7 +531,7 @@ function unblocked() {
 
 
 // show ui to manually start playback of a linked song
-function autoplay_blocked(tid) {
+function autoplay_blocked() {
 	show_modal(
 		'<div id="blk_play"><a href="#" id="blk_go"></a></div>' +
 		'<div id="blk_abrt"><a href="#" id="blk_na">Cancel<br />(show file list)</a></div>');

copyparty/web/md.html

@@ -126,7 +126,8 @@ write markdown (most html is 🙆 too)
 var last_modified = {{ lastmod }};
 var md_opt = {
 	link_md_as_html: false,
-	allow_plugins: {{ md_plug }}
+	allow_plugins: {{ md_plug }},
+	modpoll_freq: {{ md_chk_rate }}
 };
 
 (function () {

copyparty/web/md.js

@@ -1,12 +1,12 @@
 "use strict";
 
-var dom_toc = document.getElementById('toc');
+var dom_toc = ebi('toc');
-var dom_wrap = document.getElementById('mw');
+var dom_wrap = ebi('mw');
-var dom_hbar = document.getElementById('mh');
+var dom_hbar = ebi('mh');
-var dom_nav = document.getElementById('mn');
+var dom_nav = ebi('mn');
-var dom_pre = document.getElementById('mp');
+var dom_pre = ebi('mp');
-var dom_src = document.getElementById('mt');
+var dom_src = ebi('mt');
-var dom_navtgl = document.getElementById('navtoggle');
+var dom_navtgl = ebi('navtoggle');
 
 
 // chrome 49 needs this
@@ -161,7 +161,7 @@ function copydom(src, dst, lv) {
 
 
 function md_plug_err(ex, js) {
-    var errbox = document.getElementById('md_errbox');
+    var errbox = ebi('md_errbox');
     if (errbox)
         errbox.parentNode.removeChild(errbox);
 
@@ -299,7 +299,7 @@ function convert_markdown(md_text, dest_dom) {
     }
 
     // separate <code> for each line in <pre>
-    var nodes = md_dom.getElementsByTagName('pre');
+    nodes = md_dom.getElementsByTagName('pre');
     for (var a = nodes.length - 1; a >= 0; a--) {
         var el = nodes[a];
 
@@ -367,7 +367,7 @@ function convert_markdown(md_text, dest_dom) {
 
 
 function init_toc() {
-    var loader = document.getElementById('ml');
+    var loader = ebi('ml');
     loader.parentNode.removeChild(loader);
 
     var anchors = [];  // list of toc entries, complex objects

copyparty/web/md2.css

@@ -77,32 +77,52 @@ html.dark #mt {
     background: #f97;
     border-radius: .15em;
 }
+html.dark #save.force-save {
+    color: #fca;
+    background: #720;
+}
 #save.disabled {
     opacity: .4;
 }
+#helpbox,
+#toast {
+    background: #f7f7f7;
+    border-radius: .4em;
+    z-index: 9001;
+}
 #helpbox {
     display: none;
     position: fixed;
-    background: #f7f7f7;
-    box-shadow: 0 .5em 2em #777;
-    border-radius: .4em;
     padding: 2em;
     top: 4em;
     overflow-y: auto;
+    box-shadow: 0 .5em 2em #777;
     height: calc(100% - 12em);
     left: calc(50% - 15em);
     right: 0;
     width: 30em;
-    z-index: 9001;
 }
 #helpclose {
     display: block;
 }
 html.dark #helpbox {
-    background: #222;
     box-shadow: 0 .5em 2em #444;
+}
+html.dark #helpbox,
+html.dark #toast {
+    background: #222;
     border: 1px solid #079;
     border-width: 1px 0;
 }
+#toast {
+    font-weight: bold;
+    text-align: center;
+    padding: .6em 0;
+    position: fixed;
+    z-index: 9001;
+    top: 30%;
+    transition: opacity 0.2s ease-in-out;
+    opacity: 1;
+}
 
 # mt {opacity: .5;top:1px}

copyparty/web/md2.js

@@ -11,15 +11,15 @@ var js_uni_whitelist = eval('\'' + esc_uni_whitelist + '\'');
 
 
 // dom nodes
-var dom_swrap = document.getElementById('mtw');
+var dom_swrap = ebi('mtw');
-var dom_sbs = document.getElementById('sbs');
+var dom_sbs = ebi('sbs');
-var dom_nsbs = document.getElementById('nsbs');
+var dom_nsbs = ebi('nsbs');
-var dom_tbox = document.getElementById('toolsbox');
+var dom_tbox = ebi('toolsbox');
 var dom_ref = (function () {
     var d = document.createElement('div');
     d.setAttribute('id', 'mtr');
     dom_swrap.appendChild(d);
-    d = document.getElementById('mtr');
+    d = ebi('mtr');
     // hide behind the textarea (offsetTop is not computed if display:none)
     dom_src.style.zIndex = '4';
     d.style.zIndex = '3';
@@ -108,7 +108,7 @@ var draw_md = (function () {
         map_src = genmap(dom_ref, map_src);
         map_pre = genmap(dom_pre, map_pre);
 
-        cls(document.getElementById('save'), 'disabled', src == server_md);
+        cls(ebi('save'), 'disabled', src == server_md);
 
         var t1 = new Date().getTime();
         delay = t1 - t0 > 100 ? 25 : 1;
@@ -144,7 +144,7 @@ redraw = (function () {
         onresize();
     }
     function modetoggle() {
-        mode = dom_nsbs.innerHTML;
+        var mode = dom_nsbs.innerHTML;
         dom_nsbs.innerHTML = mode == 'editor' ? 'preview' : 'editor';
         mode += ' single';
         dom_wrap.setAttribute('class', mode);
@@ -223,14 +223,108 @@ redraw = (function () {
 })();
 
 
+// modification checker
+function Modpoll() {
+    this.skip_one = true;
+    this.disabled = false;
+
+    this.periodic = function () {
+        var that = this;
+        setTimeout(function () {
+            that.periodic();
+        }, 1000 * md_opt.modpoll_freq);
+
+        var skip = null;
+
+        if (ebi('toast'))
+            skip = 'toast';
+
+        else if (this.skip_one)
+            skip = 'saved';
+
+        else if (this.disabled)
+            skip = 'disabled';
+
+        if (skip) {
+            console.log('modpoll skip, ' + skip);
+            this.skip_one = false;
+            return;
+        }
+
+        console.log('modpoll...');
+        var url = (document.location + '').split('?')[0] + '?raw&_=' + new Date().getTime();
+        var xhr = new XMLHttpRequest();
+        xhr.modpoll = this;
+        xhr.open('GET', url, true);
+        xhr.responseType = 'text';
+        xhr.onreadystatechange = this.cb;
+        xhr.send();
+    }
+
+    this.cb = function () {
+        if (this.modpoll.disabled || this.modpoll.skip_one) {
+            console.log('modpoll abort');
+            return;
+        }
+
+        if (this.readyState != XMLHttpRequest.DONE)
+            return;
+
+        if (this.status !== 200) {
+            console.log('modpoll err ' + this.status + ": " + this.responseText);
+            return;
+        }
+
+        if (!this.responseText)
+            return;
+
+        var server_ref = server_md.replace(/\r/g, '');
+        var server_now = this.responseText.replace(/\r/g, '');
+
+        if (server_ref != server_now) {
+            console.log("modpoll diff |" + server_ref.length + "|, |" + server_now.length + "|");
+            this.modpoll.disabled = true;
+            var msg = [
+                "The document has changed on the server.<br />" +
+                "The changes will NOT be loaded into your editor automatically.",
+
+                "Press F5 or CTRL-R to refresh the page,<br />" +
+                "replacing your document with the server copy.",
+
+                "You can click this message to ignore and contnue."
+            ];
+            return toast(false, "box-shadow:0 1em 2em rgba(64,64,64,0.8);font-weight:normal",
+                36, "<p>" + msg.join('</p>\n<p>') + '</p>');
+        }
+
+        console.log('modpoll eq');
+    }
+
+    if (md_opt.modpoll_freq > 0)
+        this.periodic();
+
+    return this;
+}
+var modpoll = new Modpoll();
+
+
+window.onbeforeunload = function (e) {
+    if ((ebi("save").getAttribute('class') + '').indexOf('disabled') >= 0)
+        return; //nice (todo)
+
+    e.preventDefault(); //ff
+    e.returnValue = ''; //chrome
+};
+
+
 // save handler
 function save(e) {
     if (e) e.preventDefault();
-    var save_btn = document.getElementById("save"),
+    var save_btn = ebi("save"),
         save_cls = save_btn.getAttribute('class') + '';
 
     if (save_cls.indexOf('disabled') >= 0) {
-        toast('font-size:2em;color:#fc6;width:9em;', 'no changes');
+        toast(true, ";font-size:2em;color:#c90", 9, "no changes");
         return;
     }
 
@@ -254,6 +348,8 @@ function save(e) {
     xhr.onreadystatechange = save_cb;
     xhr.btn = save_btn;
     xhr.txt = txt;
+
+    modpoll.skip_one = true;  // skip one iteration while we save
     xhr.send(fd);
 }
 
@@ -347,23 +443,44 @@ function savechk_cb() {
     last_modified = this.lastmod;
     server_md = this.txt;
     draw_md();
-    toast('font-size:6em;font-family:serif;color:#cf6;width:4em;',
+    toast(true, ";font-size:6em;font-family:serif;color:#9b4", 4,
         'OK✔️<span style="font-size:.2em;color:#999;position:absolute">' + this.ntry + '</span>');
+
+    modpoll.disabled = false;
 }
 
-function toast(style, msg) {
+function toast(autoclose, style, width, msg) {
-    var ok = document.createElement('div');
+    var ok = ebi("toast");
-    style += 'font-weight:bold;background:#444;border-radius:.3em;padding:.6em 0;position:fixed;top:30%;left:calc(50% - 2em);text-align:center;z-index:9001;transition:opacity 0.2s ease-in-out;opacity:1';
+    if (ok)
+        ok.parentNode.removeChild(ok);
+
+    style = "width:" + width + "em;left:calc(50% - " + (width / 2) + "em);" + style;
+    ok = document.createElement('div');
+    ok.setAttribute('id', 'toast');
     ok.setAttribute('style', style);
     ok.innerHTML = msg;
-    var parent = document.getElementById('m');
+    var parent = ebi('m');
     document.documentElement.appendChild(ok);
+
+    var hide = function (delay) {
+        delay = delay || 0;
+
         setTimeout(function () {
             ok.style.opacity = 0;
-    }, 500);
+        }, delay);
+
         setTimeout(function () {
+            if (ok.parentNode)
                 ok.parentNode.removeChild(ok);
-    }, 750);
+        }, delay + 250);
+    }
+
+    ok.onclick = function () {
+        hide(0);
+    };
+
+    if (autoclose)
+        hide(500);
 }
 
 
@@ -543,6 +660,10 @@ function md_backspace() {
     if (/^\s*$/.test(left))
         return true;
 
+    // same if selection
+    if (o0 != dom_src.selectionEnd)
+        return true;
+
     // same if line is all-whitespace or non-markup
     var v = m[0].replace(/[^ ]/g, " ");
     if (v === m[0] || v.length !== left.length)
@@ -626,7 +747,8 @@ function fmt_table(e) {
         lpipe = tab[1].indexOf('|') < tab[1].indexOf('-'),
         rpipe = tab[1].lastIndexOf('|') > tab[1].lastIndexOf('-'),
         re_lpipe = lpipe ? /^\s*\|\s*/ : /^\s*/,
-        re_rpipe = rpipe ? /\s*\|\s*$/ : /\s*$/;
+        re_rpipe = rpipe ? /\s*\|\s*$/ : /\s*$/,
+        ncols;
 
     // the second row defines the table,
     // need to process that first
@@ -751,8 +873,7 @@ function mark_uni(e) {
     dom_tbox.setAttribute('class', '');
 
     var txt = dom_src.value,
-        ptn = new RegExp('([^' + js_uni_whitelist + ']+)', 'g');
+        ptn = new RegExp('([^' + js_uni_whitelist + ']+)', 'g'),
-
         mod = txt.replace(/\r/g, "").replace(ptn, "\u2588\u2770$1\u2771");
 
     if (txt == mod) {
@@ -789,7 +910,12 @@ function iter_uni(e) {
 // configure whitelist
 function cfg_uni(e) {
     if (e) e.preventDefault();
-    esc_uni_whitelist = prompt("unicode whitelist", esc_uni_whitelist);
+
+    var reply = prompt("unicode whitelist", esc_uni_whitelist);
+    if (reply === null)
+        return;
+
+    esc_uni_whitelist = reply;
     js_uni_whitelist = eval('\'' + esc_uni_whitelist + '\'');
 }
 
@@ -806,7 +932,7 @@ function cfg_uni(e) {
             return false;
         }
         if (ev.code == "Escape" || kc == 27) {
-            var d = document.getElementById('helpclose');
+            var d = ebi('helpclose');
             if (d)
                 d.click();
         }
@@ -863,22 +989,22 @@ function cfg_uni(e) {
         }
     }
     document.onkeydown = keydown;
-    document.getElementById('save').onclick = save;
+    ebi('save').onclick = save;
 })();
 
 
-document.getElementById('tools').onclick = function (e) {
+ebi('tools').onclick = function (e) {
     if (e) e.preventDefault();
     var is_open = dom_tbox.getAttribute('class') != 'open';
     dom_tbox.setAttribute('class', is_open ? 'open' : '');
 };
 
 
-document.getElementById('help').onclick = function (e) {
+ebi('help').onclick = function (e) {
     if (e) e.preventDefault();
     dom_tbox.setAttribute('class', '');
 
-    var dom = document.getElementById('helpbox');
+    var dom = ebi('helpbox');
     var dtxt = dom.getElementsByTagName('textarea');
     if (dtxt.length > 0) {
         convert_markdown(dtxt[0].value, dom);
@@ -886,16 +1012,16 @@ document.getElementById('help').onclick = function (e) {
     }
 
     dom.style.display = 'block';
-    document.getElementById('helpclose').onclick = function () {
+    ebi('helpclose').onclick = function () {
         dom.style.display = 'none';
     };
 };
 
 
-document.getElementById('fmt_table').onclick = fmt_table;
+ebi('fmt_table').onclick = fmt_table;
-document.getElementById('mark_uni').onclick = mark_uni;
+ebi('mark_uni').onclick = mark_uni;
-document.getElementById('iter_uni').onclick = iter_uni;
+ebi('iter_uni').onclick = iter_uni;
-document.getElementById('cfg_uni').onclick = cfg_uni;
+ebi('cfg_uni').onclick = cfg_uni;
 
 
 // blame steen
@@ -1009,7 +1135,6 @@ action_stack = (function () {
     }
 
     return {
-        push: push,
         undo: undo,
         redo: redo,
         push: schedule_push,
@@ -1019,7 +1144,7 @@ action_stack = (function () {
 })();
 
 /*
-document.getElementById('help').onclick = function () {
+ebi('help').onclick = function () {
     var c1 = getComputedStyle(dom_src).cssText.split(';');
     var c2 = getComputedStyle(dom_ref).cssText.split(';');
     var max = Math.min(c1.length, c2.length);

copyparty/web/mde.html

@@ -25,7 +25,8 @@
 var last_modified = {{ lastmod }};
 var md_opt = {
 	link_md_as_html: false,
-	allow_plugins: {{ md_plug }}
+	allow_plugins: {{ md_plug }},
+	modpoll_freq: {{ md_chk_rate }}
 };
 
 var lightswitch = (function () {

copyparty/web/mde.js

@@ -1,9 +1,9 @@
 "use strict";
 
-var dom_wrap = document.getElementById('mw');
+var dom_wrap = ebi('mw');
-var dom_nav = document.getElementById('mn');
+var dom_nav = ebi('mn');
-var dom_doc = document.getElementById('m');
+var dom_doc = ebi('m');
-var dom_md = document.getElementById('mt');
+var dom_md = ebi('mt');
 
 (function () {
     var n = document.location + '';
@@ -65,7 +65,7 @@ var mde = (function () {
     mde.codemirror.on("change", function () {
         md_changed(mde);
     });
-    var loader = document.getElementById('ml');
+    var loader = ebi('ml');
     loader.parentNode.removeChild(loader);
     return mde;
 })();
@@ -215,7 +215,7 @@ function save_chk() {
     var ok = document.createElement('div');
     ok.setAttribute('style', 'font-size:6em;font-family:serif;font-weight:bold;color:#cf6;background:#444;border-radius:.3em;padding:.6em 0;position:fixed;top:30%;left:calc(50% - 2em);width:4em;text-align:center;z-index:9001;transition:opacity 0.2s ease-in-out;opacity:1');
     ok.innerHTML = 'OK✔️';
-    var parent = document.getElementById('m');
+    var parent = ebi('m');
     document.documentElement.appendChild(ok);
     setTimeout(function () {
         ok.style.opacity = 0;

copyparty/web/up2k.js

@@ -33,12 +33,12 @@ function goto(dest) {
     for (var a = obj.length - 1; a >= 0; a--)
         obj[a].classList.remove('act');
 
-    var obj = document.querySelectorAll('#ops>a');
+    obj = document.querySelectorAll('#ops>a');
     for (var a = obj.length - 1; a >= 0; a--)
         obj[a].classList.remove('act');
 
     if (dest) {
-        document.getElementById('op_' + dest).classList.add('act');
+        ebi('op_' + dest).classList.add('act');
         document.querySelector('#ops>a[data-dest=' + dest + ']').classList.add('act');
 
         var fn = window['goto_' + dest];
@@ -66,7 +66,7 @@ function goto_up2k() {
         if (op !== null && op !== '.')
             goto(op);
     }
-    document.getElementById('ops').style.display = 'block';
+    ebi('ops').style.display = 'block';
 })();
 
 
@@ -130,7 +130,7 @@ function up2k_init(have_crypto) {
             else
                 ebi('u2foot').innerHTML = 'seems like ' + shame + ' so do that if you want more performance';
         }
-    };
+    }
 
     // show uploader if the user only has write-access
     if (!ebi('files'))
@@ -262,6 +262,7 @@ function up2k_init(have_crypto) {
 
         more_one_file();
         var bad_files = [];
+        var good_files = [];
         for (var a = 0; a < files.length; a++) {
             var fobj = files[a];
             if (is_itemlist) {
@@ -275,9 +276,32 @@ function up2k_init(have_crypto) {
                     throw 1;
             }
             catch (ex) {
-                bad_files.push([a, fobj.name]);
+                bad_files.push(fobj.name);
                 continue;
             }
+            good_files.push(fobj);
+        }
+
+        if (bad_files.length > 0) {
+            var msg = 'These {0} files (of {1} total) were skipped because they are empty:\n'.format(bad_files.length, files.length);
+            for (var a = 0; a < bad_files.length; a++)
+                msg += '-- ' + bad_files[a] + '\n';
+
+            if (files.length - bad_files.length <= 1 && /(android)/i.test(navigator.userAgent))
+                msg += '\nFirefox-Android has a bug which prevents selecting multiple files. Try selecting one file at a time. For more info, see firefox bug 1456557';
+
+            alert(msg);
+        }
+
+        var msg = ['upload these ' + good_files.length + ' files?'];
+        for (var a = 0; a < good_files.length; a++)
+            msg.push(good_files[a].name);
+
+        if (!confirm(msg.join('\n')))
+            return;
+
+        for (var a = 0; a < good_files.length; a++) {
+            var fobj = good_files[a];
             var now = new Date().getTime();
             var lmod = fobj.lastModified || now;
             var entry = {
@@ -307,17 +331,6 @@ function up2k_init(have_crypto) {
             st.files.push(entry);
             st.todo.hash.push(entry);
         }
-
-        if (bad_files.length > 0) {
-            var msg = 'These {0} files (of {1} total) were skipped because they are empty:\n'.format(bad_files.length, files.length);
-            for (var a = 0; a < bad_files.length; a++)
-                msg += '-- ' + bad_files[a][1] + '\n';
-
-            if (files.length - bad_files.length <= 1 && /(android)/i.test(navigator.userAgent))
-                msg += '\nFirefox-Android has a bug which prevents selecting multiple files. Try selecting one file at a time. For more info, see firefox bug 1456557';
-
-            alert(msg);
-        }
     }
     ebi('u2btn').addEventListener('drop', gotfile, false);
 
@@ -336,16 +349,17 @@ function up2k_init(have_crypto) {
     //
 
     function handshakes_permitted() {
-        return multitask || (
+        var lim = multitask ? 1 : 0;
-            st.todo.upload.length == 0 &&
+        return lim >=
-            st.busy.upload.length == 0);
+            st.todo.upload.length +
+            st.busy.upload.length;
     }
 
     function hashing_permitted() {
-        return multitask || (
+        var lim = multitask ? 1 : 0;
-            handshakes_permitted() &&
+        return handshakes_permitted() && lim >=
-            st.todo.handshake.length == 0 &&
+            st.todo.handshake.length +
-            st.busy.handshake.length == 0);
+            st.busy.handshake.length;
     }
 
     var tasker = (function () {
@@ -396,17 +410,6 @@ function up2k_init(have_crypto) {
     ///   hashing
     //
 
-    // https://gist.github.com/jonleighton/958841
-    function buf2b64_maybe_fucky(buffer) {
-        var ret = '';
-        var view = new DataView(buffer);
-        for (var i = 0; i < view.byteLength; i++) {
-            ret += String.fromCharCode(view.getUint8(i));
-        }
-        return window.btoa(ret).replace(
-            /\+/g, '-').replace(/\//g, '_').replace(/=+$/, '');
-    }
-
     // https://gist.github.com/jonleighton/958841
     function buf2b64(arrayBuffer) {
         var base64 = '';
@@ -447,20 +450,6 @@ function up2k_init(have_crypto) {
         return base64;
     }
 
-    // https://developer.mozilla.org/en-US/docs/Web/API/SubtleCrypto/digest
-    function buf2hex(buffer) {
-        var hexCodes = [];
-        var view = new DataView(buffer);
-        for (var i = 0; i < view.byteLength; i += 4) {
-            var value = view.getUint32(i) // 4 bytes per iter
-            var stringValue = value.toString(16) // doesn't pad
-            var padding = '00000000'
-            var paddedValue = (padding + stringValue).slice(-padding.length)
-            hexCodes.push(paddedValue);
-        }
-        return hexCodes.join("");
-    }
-
     function get_chunksize(filesize) {
         var chunksize = 1024 * 1024;
         var stepsize = 512 * 1024;
@@ -692,12 +681,30 @@ function up2k_init(have_crypto) {
                 }
                 tasker();
             }
-            else
+            else {
+                var err = "";
+                var rsp = (xhr.responseText + '');
+                if (rsp.indexOf('partial upload exists') !== -1 ||
+                    rsp.indexOf('file already exists') !== -1) {
+                    err = rsp;
+                    var ofs = err.lastIndexOf(' : ');
+                    if (ofs > 0)
+                        err = err.slice(0, ofs);
+                }
+                if (err != "") {
+                    ebi('f{0}t'.format(t.n)).innerHTML = "ERROR";
+                    ebi('f{0}p'.format(t.n)).innerHTML = err;
+
+                    st.busy.handshake.splice(st.busy.handshake.indexOf(t), 1);
+                    tasker();
+                    return;
+                }
                 alert("server broke (error {0}):\n\"{1}\"\n".format(
                     xhr.status,
                     (xhr.response && xhr.response.err) ||
                     (xhr.responseText && xhr.responseText) ||
                     "no further information"));
+            }
         };
         xhr.open('POST', post_url + 'handshake.php', true);
         xhr.responseType = 'text';

copyparty/web/util.js

@@ -36,7 +36,7 @@ function vis_exh(msg, url, lineNo, columnNo, error) {
     document.body.style.fontSize = '0.8em';
     document.body.style.padding = '0 1em 1em 1em';
     hcroak(html.join('\n'));
-};
+}
 
 
 function ebi(id) {

docs/notes.sh

@@ -3,6 +3,14 @@ echo not a script
 exit 1
 
 
+##
+## delete all partial uploads
+##  (supports linux/macos, probably windows+msys2)
+
+gzip -d < .hist/up2k.snap | jq -r '.[].tnam' | while IFS= read -r f; do rm -f -- "$f"; done
+gzip -d < .hist/up2k.snap | jq -r '.[].name' | while IFS= read -r f; do wc -c -- "$f" | grep -qiE '^[^0-9a-z]*0' & rm -f -- "$f"; done
+
+
 ##
 ## create a test payload
 

scripts/copyparty-repack.sh

@@ -1,4 +1,5 @@
 #!/bin/bash
+repacker=1
 set -e
 
 # -- download latest copyparty (source.tgz and sfx),
@@ -19,20 +20,32 @@ set -e
 # -rwxr-xr-x  0 ed ed  183808 Nov 19 00:43 copyparty-extras/sfx-lite/copyparty-sfx.py
 
 
-command -v gtar && tar=gtar || tar=tar
+command -v gtar && tar() { gtar "$@"; }
+command -v gsed && sed() { gsed "$@"; }
 td="$(mktemp -d)"
 od="$(pwd)"
 cd "$td"
 pwd
 
 
-# debug: if cache exists, use that instead of bothering github
+dl_text() {
+	command -v curl && exec curl "$@"
+	exec wget -O- "$@"
+}
+dl_files() {
+	command -v curl && exec curl -L --remote-name-all "$@"
+	exec wget "$@"
+}
+export -f dl_files
+
+
+# if cache exists, use that instead of bothering github
 cache="$od/.copyparty-repack.cache"
 [ -e "$cache" ] &&
-	$tar -xvf "$cache" ||
+	tar -xf "$cache" ||
 {
 	# get download links from github
-	curl https://api.github.com/repos/9001/copyparty/releases/latest |
+	dl_text https://api.github.com/repos/9001/copyparty/releases/latest |
 	(
 		# prefer jq if available
 		jq -r '.assets[]|select(.name|test("-sfx|tar.gz")).browser_download_url' ||
@@ -41,10 +54,10 @@ cache="$od/.copyparty-repack.cache"
 		awk -F\" '/"browser_download_url".*(\.tar\.gz|-sfx\.)/ {print$4}'
 	) |
 	tee /dev/stderr |
-	tr -d '\r' | tr '\n' '\0' | xargs -0 curl -L --remote-name-all
+	tr -d '\r' | tr '\n' '\0' |
+	xargs -0 bash -c 'dl_files "$@"' _
 
-	# debug: create cache
+	tar -czf "$cache" *
-	#$tar -czvf "$cache" *
 }
 
 
@@ -57,10 +70,21 @@ mv copyparty-*.tar.gz copyparty-extras/
 
 # unpack the source code
 ( cd copyparty-extras/
-$tar -xvf *.tar.gz
+tar -xf *.tar.gz
 )
 
 
+# use repacker from release if that is newer
+p_other=copyparty-extras/copyparty-*/scripts/copyparty-repack.sh
+other=$(awk -F= 'BEGIN{v=-1} NR<10&&/^repacker=/{v=$NF} END{print v}' <$p_other) 
+[ $repacker -lt $other ] &&
+  cat $p_other >"$od/$0" && cd "$od" && rm -rf "$td" && exec "$0" "$@"
+
+
+# now drop the cache
+rm -f "$cache"
+
+
 # fix permissions
 chmod 755 \
   copyparty-extras/sfx-full/* \
@@ -87,13 +111,15 @@ rm -rf copyparty-{0..9}*.*.*{0..9}
 )
 
 
- # and include the repacker itself too
+# and include the repacker itself too
-cp -pv "$od/$0" copyparty-extras/ 
+cp -av "$od/$0" copyparty-extras/ ||
+cp -av "$0" copyparty-extras/ ||
+true
 
 
 # create the bundle
 fn=copyparty-$(date +%Y-%m%d-%H%M%S).tgz
-$tar -czvf "$od/$fn" *
+tar -czvf "$od/$fn" *
 cd "$od"
 rm -rf "$td"
 

scripts/fusefuzz.py

@@ -4,10 +4,10 @@ import os
 import time
 
 """
-mkdir -p /dev/shm/fusefuzz/{r,v}
+td=/dev/shm/; [ -e $td ] || td=$HOME; mkdir -p $td/fusefuzz/{r,v}
-PYTHONPATH=.. python3 -m copyparty -v /dev/shm/fusefuzz/r::r -i 127.0.0.1
+PYTHONPATH=.. python3 -m copyparty -v $td/fusefuzz/r::r -i 127.0.0.1
-../bin/copyparty-fuse.py /dev/shm/fusefuzz/v http://127.0.0.1:3923/ 2 0
+../bin/copyparty-fuse.py http://127.0.0.1:3923/ $td/fusefuzz/v -cf 2 -cd 0.5
-(d="$PWD"; cd /dev/shm/fusefuzz && "$d"/fusefuzz.py)
+(d="$PWD"; cd $td/fusefuzz && "$d"/fusefuzz.py)
 """
 
 

scripts/make-sfx.sh

@@ -18,13 +18,16 @@ echo
 #   (the fancy markdown editor)
 
 
-command -v gtar  >/dev/null &&
+# port install gnutar findutils gsed coreutils
-command -v gfind >/dev/null && {
+gtar=$(command -v gtar || command -v gnutar) || true
-	tar()  { gtar  "$@"; }
+[ ! -z "$gtar" ] && command -v gfind >/dev/null && {
+	tar()  { $gtar "$@"; }
 	sed()  { gsed  "$@"; }
 	find() { gfind "$@"; }
 	sort() { gsort "$@"; }
 	unexpand() { gunexpand "$@"; }
+	command -v grealpath >/dev/null &&
+		realpath() { grealpath "$@"; }
 }
 
 [ -e copyparty/__main__.py ] || cd ..
@@ -95,7 +98,7 @@ cd sfx
 }
 
 ver=
-command -v git >/dev/null && {
+git describe --tags >/dev/null 2>/dev/null && {
 	git_ver="$(git describe --tags)";  # v0.5.5-2-gb164aa0
 	ver="$(printf '%s\n' "$git_ver" | sed -r 's/^v//; s/-g?/./g')";
 	t_ver=
@@ -115,7 +118,7 @@ command -v git >/dev/null && {
 		exit 1
 	}
 
-	dt="$(git log -1 --format=%cd --date=format:'%Y, %m, %d')"
+	dt="$(git log -1 --format=%cd --date=format:'%Y,%m,%d' | sed -E 's/,0?/, /g')"
 	printf 'git %3s: \033[36m%s\033[0m\n' ver "$ver" dt "$dt"
 	sed -ri '
 		s/^(VERSION =)(.*)/#\1\2\n\1 ('"$t_ver"')/;

scripts/make-tgz-release.sh

@@ -16,15 +16,15 @@ which md5sum 2>/dev/null >/dev/null &&
 
 ver="$1"
 
-[[ "x$ver" == x ]] &&
+[ "x$ver" = x ] &&
 {
 	echo "need argument 1:  version"
 	echo
 	exit 1
 }
 
-[[ -e copyparty/__main__.py ]] || cd ..
+[ -e copyparty/__main__.py ] || cd ..
-[[ -e copyparty/__main__.py ]] ||
+[ -e copyparty/__main__.py ] ||
 {
 	echo "run me from within the project root folder"
 	echo
@@ -35,8 +35,8 @@ mkdir -p dist
 zip_path="$(pwd)/dist/copyparty-$ver.zip"
 tgz_path="$(pwd)/dist/copyparty-$ver.tar.gz"
 
-[[ -e "$zip_path" ]] ||
+[ -e "$zip_path" ] ||
-[[ -e "$tgz_path" ]] &&
+[ -e "$tgz_path" ] &&
 {
 	echo "found existing archives for this version"
 	echo "  $zip_path"

tests/test_vfs.py

@@ -6,6 +6,7 @@ import os
 import time
 import json
 import shutil
+import tempfile
 import unittest
 import subprocess as sp  # nosec
 
@@ -31,9 +32,6 @@ class TestVFS(unittest.TestCase):
         response = self.unfoo(response)
         self.assertEqual(util.undot(query), response)
 
-    def absify(self, root, names):
-        return ["{}/{}".format(root, x).replace("//", "/") for x in names]
-
     def ls(self, vfs, vpath, uname):
         """helper for resolving and listing a folder"""
         vn, rem = vfs.get(vpath, uname, True, False)
@@ -60,23 +58,31 @@ class TestVFS(unittest.TestCase):
 
         if os.path.exists("/Volumes"):
             devname, _ = self.chkcmd("hdiutil", "attach", "-nomount", "ram://8192")
+            devname = devname.strip()
+            print("devname: [{}]".format(devname))
             for _ in range(10):
                 try:
-                    _, _ = self.chkcmd("diskutil", "eraseVolume", "HFS+", "cptd", devname)
+                    _, _ = self.chkcmd(
+                        "diskutil", "eraseVolume", "HFS+", "cptd", devname
+                    )
                     return "/Volumes/cptd"
-                except:
+                except Exception as ex:
-                    print('lol macos')
+                    print(repr(ex))
                     time.sleep(0.25)
 
             raise Exception("ramdisk creation failed")
 
-        raise Exception("TODO support windows")
+        ret = os.path.join(tempfile.gettempdir(), "copyparty-test")
+        try:
+            os.mkdir(ret)
+        finally:
+            return ret
 
     def log(self, src, msg):
         pass
 
     def test(self):
-        td = self.get_ramdisk() + "/vfs"
+        td = os.path.join(self.get_ramdisk(), "vfs")
         try:
             shutil.rmtree(td)
         except OSError:
@@ -107,7 +113,7 @@ class TestVFS(unittest.TestCase):
         vfs = AuthSrv(Namespace(c=None, a=[], v=["a/ab/::r"]), self.log).vfs
         self.assertEqual(vfs.nodes, {})
         self.assertEqual(vfs.vpath, "")
-        self.assertEqual(vfs.realpath, td + "/a/ab")
+        self.assertEqual(vfs.realpath, os.path.join(td, "a", "ab"))
         self.assertEqual(vfs.uread, ["*"])
         self.assertEqual(vfs.uwrite, [])
 
@@ -117,7 +123,7 @@ class TestVFS(unittest.TestCase):
         ).vfs
         self.assertEqual(vfs.nodes, {})
         self.assertEqual(vfs.vpath, "")
-        self.assertEqual(vfs.realpath, td + "/a/aa")
+        self.assertEqual(vfs.realpath, os.path.join(td, "a", "aa"))
         self.assertEqual(vfs.uread, ["*"])
         self.assertEqual(vfs.uwrite, [])
 
@@ -146,42 +152,63 @@ class TestVFS(unittest.TestCase):
         n = n.nodes["acb"]
         self.assertEqual(n.nodes, {})
         self.assertEqual(n.vpath, "a/ac/acb")
-        self.assertEqual(n.realpath, td + "/a/ac/acb")
+        self.assertEqual(n.realpath, os.path.join(td, "a", "ac", "acb"))
         self.assertEqual(n.uread, ["k"])
         self.assertEqual(n.uwrite, ["*", "k"])
 
+        # something funky about the windows path normalization,
+        # doesn't really matter but makes the test messy, TODO?
+
         fsdir, real, virt = self.ls(vfs, "/", "*")
         self.assertEqual(fsdir, td)
         self.assertEqual(real, ["b", "c"])
         self.assertEqual(list(virt), ["a"])
 
         fsdir, real, virt = self.ls(vfs, "a", "*")
-        self.assertEqual(fsdir, td + "/a")
+        self.assertEqual(fsdir, os.path.join(td, "a"))
         self.assertEqual(real, ["aa", "ab"])
         self.assertEqual(list(virt), ["ac"])
 
         fsdir, real, virt = self.ls(vfs, "a/ab", "*")
-        self.assertEqual(fsdir, td + "/a/ab")
+        self.assertEqual(fsdir, os.path.join(td, "a", "ab"))
         self.assertEqual(real, ["aba", "abb", "abc"])
         self.assertEqual(list(virt), [])
 
         fsdir, real, virt = self.ls(vfs, "a/ac", "*")
-        self.assertEqual(fsdir, td + "/a/ac")
+        self.assertEqual(fsdir, os.path.join(td, "a", "ac"))
         self.assertEqual(real, ["aca", "acc"])
         self.assertEqual(list(virt), [])
 
         fsdir, real, virt = self.ls(vfs, "a/ac", "k")
-        self.assertEqual(fsdir, td + "/a/ac")
+        self.assertEqual(fsdir, os.path.join(td, "a", "ac"))
         self.assertEqual(real, ["aca", "acc"])
         self.assertEqual(list(virt), ["acb"])
 
         self.assertRaises(util.Pebkac, vfs.get, "a/ac/acb", "*", True, False)
 
         fsdir, real, virt = self.ls(vfs, "a/ac/acb", "k")
-        self.assertEqual(fsdir, td + "/a/ac/acb")
+        self.assertEqual(fsdir, os.path.join(td, "a", "ac", "acb"))
         self.assertEqual(real, ["acba", "acbb", "acbc"])
         self.assertEqual(list(virt), [])
 
+        # admin-only rootfs with all-read-only subfolder
+        vfs = AuthSrv(Namespace(c=None, a=["k:k"], v=[".::ak", "a:a:r"]), self.log,).vfs
+        self.assertEqual(len(vfs.nodes), 1)
+        self.assertEqual(vfs.vpath, "")
+        self.assertEqual(vfs.realpath, td)
+        self.assertEqual(vfs.uread, ["k"])
+        self.assertEqual(vfs.uwrite, ["k"])
+        n = vfs.nodes["a"]
+        self.assertEqual(len(vfs.nodes), 1)
+        self.assertEqual(n.vpath, "a")
+        self.assertEqual(n.realpath, os.path.join(td, "a"))
+        self.assertEqual(n.uread, ["*"])
+        self.assertEqual(n.uwrite, [])
+        self.assertEqual(vfs.can_access("/", "*"), [False, False])
+        self.assertEqual(vfs.can_access("/", "k"), [True, True])
+        self.assertEqual(vfs.can_access("/a", "*"), [True, False])
+        self.assertEqual(vfs.can_access("/a", "k"), [True, False])
+
         # breadth-first construction
         vfs = AuthSrv(
             Namespace(
@@ -215,20 +242,20 @@ class TestVFS(unittest.TestCase):
         self.assertEqual(list(v1), ["a"])
 
         fsp, r1, v1 = self.ls(vfs, "a", "*")
-        self.assertEqual(fsp, td + "/a")
+        self.assertEqual(fsp, os.path.join(td, "a"))
         self.assertEqual(r1, ["aa", "ab"])
         self.assertEqual(list(v1), ["ac"])
 
         fsp1, r1, v1 = self.ls(vfs, "a/ac", "*")
         fsp2, r2, v2 = self.ls(vfs, "b", "*")
-        self.assertEqual(fsp1, td + "/b")
+        self.assertEqual(fsp1, os.path.join(td, "b"))
-        self.assertEqual(fsp2, td + "/b")
+        self.assertEqual(fsp2, os.path.join(td, "b"))
         self.assertEqual(r1, ["ba", "bb", "bc"])
         self.assertEqual(r1, r2)
         self.assertEqual(list(v1), list(v2))
 
         # config file parser
-        cfg_path = self.get_ramdisk() + "/test.cfg"
+        cfg_path = os.path.join(self.get_ramdisk(), "test.cfg")
         with open(cfg_path, "wb") as f:
             f.write(
                 dedent(
@@ -256,10 +283,11 @@ class TestVFS(unittest.TestCase):
         self.assertEqual(len(n.nodes), 1)
         n = n.nodes["dst"]
         self.assertEqual(n.vpath, "dst")
-        self.assertEqual(n.realpath, td + "/src")
+        self.assertEqual(n.realpath, os.path.join(td, "src"))
         self.assertEqual(n.uread, ["a", "asd"])
         self.assertEqual(n.uwrite, ["asd"])
         self.assertEqual(len(n.nodes), 0)
 
+        os.chdir(tempfile.gettempdir())
         shutil.rmtree(td)
         os.unlink(cfg_path)