[PATCH 2 of 2] util: use tuple accessor to get accurate st_mtime value (issue4836)
Yuya Nishihara
yuya at tcha.org
Tue Oct 6 09:44:27 CDT 2015
# HG changeset patch
# User Yuya Nishihara <yuya at tcha.org>
# Date 1443965736 -32400
# Sun Oct 04 22:35:36 2015 +0900
# Node ID 026a9c645639a1b642e42b0dafeda6461f6ebc73
# Parent 869c9d9c777435f94a84d7284d6ba5c6a39e5ded
util: use tuple accessor to get accurate st_mtime value (issue4836)
Because st.st_mtime is computed as 'sec + 1e-9 * nsec' and double is too narrow
to represent nanoseconds, int(st.st_mtime) can be 'sec + 1'. Therefore, that
value could be different from the one got by osutils.listdir().
This patch fixes the problem by accessing to raw st_mtime by tuple index.
It catches TypeError to fall back to st.st_mtime because our osutil.stat does
not support tuple index. In dirstate.normal(), 'st' is always a Python stat,
but in dirstate.status(), it can be either a Python stat or an osutil.stat.
Thanks to vgatien-baron at janestreet.com for finding the root cause of this
subtle problem.
diff --git a/mercurial/util.py b/mercurial/util.py
--- a/mercurial/util.py
+++ b/mercurial/util.py
@@ -19,6 +19,7 @@ import error, osutil, encoding, parsers
import errno, shutil, sys, tempfile, traceback
import re as remod
import os, time, datetime, calendar, textwrap, signal, collections
+import stat
import imp, socket, urllib
import gc
import bz2
@@ -953,7 +954,18 @@ def fstat(fp):
return os.stat(fp.name)
def statmtimesec(st):
- return int(st.st_mtime)
+ """Get mtime as integer of seconds
+
+ 'int(st.st_mtime)' cannot be used because st.st_mtime is computed as
+ 'sec + 1e-9 * nsec' and double-precision floating-point type is too narrow
+ to represent nanoseconds. If 'nsec' is close to 1 sec, 'int(st.st_mtime)'
+ can be 'sec + 1'. (issue4836)
+ """
+ try:
+ return st[stat.ST_MTIME]
+ except TypeError:
+ # osutil.stat doesn't allow index access and its st_mtime is int
+ return st.st_mtime
# File system features
More information about the Mercurial-devel
mailing list