From f6e693f0f553041c4c80d644dc47293426ceee7e Mon Sep 17 00:00:00 2001
From: ed <s@ocv.me>
Date: Wed, 24 Apr 2024 21:18:26 +0000
Subject: [PATCH] reevaluate support for sparse files periodically

if a given filesystem were to disappear (e.g. removable storage)
followed by another filesystem appearing at the same location,
this would not get noticed by up2k in a timely manner

fix this by discarding the mtab cache after `--mtab-age` seconds and
rebuild it from scratch, unless the previous values are definitely
correct (as indicated by identical output from `/bin/mount`)

probably reduces windows performance by an acceptable amount
---
 copyparty/__main__.py |  1 +
 copyparty/fsutil.py   | 30 +++++++++++++++++++++++++-----
 copyparty/up2k.py     |  2 +-
 tests/util.py         |  2 +-
 4 files changed, 28 insertions(+), 7 deletions(-)

diff --git a/copyparty/__main__.py b/copyparty/__main__.py
index 931e187f..17289afa 100755
--- a/copyparty/__main__.py
+++ b/copyparty/__main__.py
@@ -860,6 +860,7 @@ def add_fs(ap):
     ap2.add_argument("--rm-retry", metavar="T/R", type=u, default=rm_re_def, help="if a file cannot be deleted because it is busy, continue trying for \033[33mT\033[0m seconds, retry every \033[33mR\033[0m seconds; disable with 0/0 (volflag=rm_retry)")
     ap2.add_argument("--mv-retry", metavar="T/R", type=u, default=rm_re_def, help="if a file cannot be renamed because it is busy, continue trying for \033[33mT\033[0m seconds, retry every \033[33mR\033[0m seconds; disable with 0/0 (volflag=mv_retry)")
     ap2.add_argument("--iobuf", metavar="BYTES", type=int, default=256*1024, help="file I/O buffer-size; if your volumes are on a network drive, try increasing to \033[32m524288\033[0m or even \033[32m4194304\033[0m (and let me know if that improves your performance)")
+    ap2.add_argument("--mtab-age", metavar="SEC", type=int, default=60, help="rebuild mountpoint cache every \033[33mSEC\033[0m to keep track of sparse-files support; keep low on servers with removable media")
 
 
 def add_upload(ap):
diff --git a/copyparty/fsutil.py b/copyparty/fsutil.py
index 5bf5ab6b..5ba9df43 100644
--- a/copyparty/fsutil.py
+++ b/copyparty/fsutil.py
@@ -1,6 +1,7 @@
 # coding: utf-8
 from __future__ import print_function, unicode_literals
 
+import argparse
 import os
 import re
 import time
@@ -17,20 +18,26 @@ if True:  # pylint: disable=using-constant-test
 
 
 class Fstab(object):
-    def __init__(self, log: "RootLogger"):
+    def __init__(self, log: "RootLogger", args: argparse.Namespace):
         self.log_func = log
 
+        self.warned = False
         self.trusted = False
         self.tab: Optional[VFS] = None
+        self.oldtab: Optional[VFS] = None
+        self.srctab = "a"
         self.cache: dict[str, str] = {}
         self.age = 0.0
+        self.maxage = args.mtab_age
 
     def log(self, msg: str, c: Union[int, str] = 0) -> None:
         self.log_func("fstab", msg, c)
 
     def get(self, path: str) -> str:
-        if len(self.cache) > 9000:
-            self.age = time.time()
+        now = time.time()
+        if now - self.age > self.maxage or len(self.cache) > 9000:
+            self.age = now
+            self.oldtab = self.tab or self.oldtab
             self.tab = None
             self.cache = {}
 
@@ -75,7 +82,7 @@ class Fstab(object):
         self.trusted = False
 
     def build_tab(self) -> None:
-        self.log("building tab")
+        self.log("inspecting mtab for changes")
 
         sptn = r"^.*? on (.*) type ([^ ]+) \(.*"
         if MACOS:
@@ -84,6 +91,7 @@ class Fstab(object):
         ptn = re.compile(sptn)
         so, _ = chkcmd(["mount"])
         tab1: list[tuple[str, str]] = []
+        atab = []
         for ln in so.split("\n"):
             m = ptn.match(ln)
             if not m:
@@ -91,6 +99,15 @@ class Fstab(object):
 
             zs1, zs2 = m.groups()
             tab1.append((str(zs1), str(zs2)))
+            atab.append(ln)
+
+        # keep empirically-correct values if mounttab unchanged
+        srctab = "\n".join(sorted(atab))
+        if srctab == self.srctab:
+            self.tab = self.oldtab
+            return
+
+        self.log("mtab has changed; reevaluating support for sparse files")
 
         tab1.sort(key=lambda x: (len(x[0]), x[0]))
         path1, fs1 = tab1[0]
@@ -99,6 +116,7 @@ class Fstab(object):
             tab.add(fs, path.lstrip("/"))
 
         self.tab = tab
+        self.srctab = srctab
 
     def relabel(self, path: str, nval: str) -> None:
         assert self.tab
@@ -133,7 +151,9 @@ class Fstab(object):
                 self.trusted = True
             except:
                 # prisonparty or other restrictive environment
-                self.log("failed to build tab:\n{}".format(min_ex()), 3)
+                if not self.warned:
+                    self.warned = True
+                    self.log("failed to build tab:\n{}".format(min_ex()), 3)
                 self.build_fallback()
 
         assert self.tab
diff --git a/copyparty/up2k.py b/copyparty/up2k.py
index a62ad77d..db993541 100644
--- a/copyparty/up2k.py
+++ b/copyparty/up2k.py
@@ -186,7 +186,7 @@ class Up2k(object):
             t = "could not initialize sqlite3, will use in-memory registry only"
             self.log(t, 3)
 
-        self.fstab = Fstab(self.log_func)
+        self.fstab = Fstab(self.log_func, self.args)
         self.gen_fk = self._gen_fk if self.args.log_fk else gen_filekey
 
         if self.args.hash_mt < 2:
diff --git a/tests/util.py b/tests/util.py
index 8a5a75e6..2cc02e17 100644
--- a/tests/util.py
+++ b/tests/util.py
@@ -122,7 +122,7 @@ class Cfg(Namespace):
         ex = "hash_mt srch_time u2abort u2j"
         ka.update(**{k: 1 for k in ex.split()})
 
-        ex = "reg_cap s_thead s_tbody th_convt"
+        ex = "mtab_age reg_cap s_thead s_tbody th_convt"
         ka.update(**{k: 9 for k in ex.split()})
 
         ex = "db_act df k304 loris re_maxage rproxy rsp_jtr rsp_slp s_wr_slp snap_wri theme themes turbo"