Author: futatuki
Date: Sat Feb 22 17:39:12 2020
New Revision: 1874393
URL: http://svn.apache.org/viewvc?rev=1874393&view=rev
Log:
Fix handling for non-ascii characters on internal path and stderr
* tools/hook-scripts/validate-files.py
(docstring): Add note to handle non-ascii path and/or stderr output.
(Commands.svnlook_changed):
+ Decode changed paths as 'utf-8' explicitly on Python 3 and leave them as
('utf-8' encoded) bytes in Python 2.7
+ Decode stderr out message from svnlook as sys.stderr.encoding
(Commands.user_command): Decode stderr out message from user commands as
sys.stderr.encoding
Modified:
subversion/trunk/tools/hook-scripts/validate-files.py
Modified: subversion/trunk/tools/hook-scripts/validate-files.py
URL:
http://svn.apache.org/viewvc/subversion/trunk/tools/hook-scripts/validate-files.py?rev=1874393&r1=1874392&r2=1874393&view=diff
==============================================================================
--- subversion/trunk/tools/hook-scripts/validate-files.py (original)
+++ subversion/trunk/tools/hook-scripts/validate-files.py Sat Feb 22 17:39:12
2020
@@ -19,7 +19,13 @@
"""Subversion pre-commit hook script that runs user configured commands
to validate files in the commit and reject the commit if the commands
exit with a non-zero exit code. The script expects a validate-files.conf
-file placed in the conf dir under the repo the commit is for."""
+file placed in the conf dir under the repo the commit is for.
+
+Note: As changed file paths $FILE are always represented as a Unicode (Py3)
+ or UTF-8 (Py2) strings, you might need to set apropriate locale and
+ PYTHONIOENCODING environment variable for this script and
+ commands to handle non-ascii path and command outputs, especially
+ you want to use svnlook cat command to inspect file contents."""
import sys
import os
@@ -82,18 +88,26 @@ class Commands:
line = p.stdout.readline()
if not line:
break
- line = line.decode().strip()
+ line = line.strip()
text_mod = line[0:1]
# Only if the contents of the file changed (by addition or update)
# directories always end in / in the svnlook changed output
- if line[-1] != "/" and (text_mod == "A" or text_mod == "U"):
- changed.append(line[4:])
+ if line[-1:] != b"/" and (text_mod == b"A" or text_mod == b"U"):
+ changed_path = line[4:]
+ if not isinstance(changed_path, str):
+ # svnlook always uses UTF-8 for internal path
+ changed_path = changed_path.decode('utf-8')
+ changed.append(changed_path)
# wait on the command to finish so we can get the
# returncode/stderr output
data = p.communicate()
if p.returncode != 0:
- sys.stderr.write(data[1].decode())
+ err_mesg = data[1]
+ if sys.stderr.encoding:
+ err_mesg =err_mesg.decode(sys.stderr.encoding,
+ 'backslashreplace')
+ sys.stderr.write(err_mesg)
sys.exit(2)
return changed
@@ -111,7 +125,11 @@ class Commands:
cmd_env['FILE'] = fn
p = subprocess.Popen(cmd, shell=True, env=cmd_env,
stderr=subprocess.PIPE)
data = p.communicate()
- return (p.returncode, data[1].decode())
+ err_mesg = data[1]
+ if sys.stderr.encoding:
+ err_mesg = err_mesg.decode(sys.stderr.encoding,
+ 'backslashreplace')
+ return (p.returncode, err_mesg)
def main(repo, txn):
exitcode = 0
