itstool: use proposed patch to port to python-lxml (#5681)

So we can get rid of libxml2-python, which is deprecated upstream.
It's not clear when upstream will integrate it so use the porposed
patch for now.

Remove patches which conflict with the above, with the hope that
moving to lxml might have made the obsolete, or the issues are no
longer there.

If there are any regressions, please speak up and we'll have a look.
This commit is contained in:
Christoph Reiter 2025-09-29 06:37:49 +02:00 committed by GitHub
parent 1e0954a432
commit bf0cde41de
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
4 changed files with 1505 additions and 102 deletions

1493
itstool/57.patch Normal file

File diff suppressed because it is too large Load Diff

View File

@ -2,24 +2,27 @@
pkgname=itstool
pkgver=2.0.7
pkgrel=1
pkgrel=2
pkgdesc="XML to PO and back again"
arch=('i686' 'x86_64')
url="https://itstool.org/"
license=("GPL3")
depends=("python" "libxml2" "libxml2-python")
msys2_repository_url="https://github.com/itstool/itstool"
license=("spdx:GPL-3.0-or-later")
depends=("python" "python-lxml")
makedepends=('autotools' 'gcc')
source=(https://files.itstool.org/itstool/${pkgname}-${pkgver}.tar.bz2
fix-segfault.patch
itstool-2.0.5-fix-crash-wrong-encoding.patch)
https://github.com/itstool/itstool/commit/32c7d07664dc37765100285d1202d488cd6a27e8.patch
57.patch)
sha256sums=('6b9a7cd29a12bb95598f5750e8763cee78836a1a207f85b74d8b3275b27e87ca'
'ccdf85ae9bdfd0259728a79b6b9806d8b9bcb27ed1f974ac49c551587cb4e6c6'
'cb57e3694ab3d7c62b063629b2e9edc6327260c0797d0f33c8dc97fe37c40ebb')
'4e64a2e884f9d4cbc493732fcbde9f1d5bed534f9a66330bbcc1cbeb54808c1e'
'ad89a36fd36b985daeec87003c9cc8c98abee3f2de84644aa10241b2d22098cf')
prepare() {
cd "${srcdir}/${pkgname}-${pkgver}"
patch -p1 -i ${srcdir}/itstool-2.0.5-fix-crash-wrong-encoding.patch
patch -p1 -i ${srcdir}/fix-segfault.patch
patch -p1 -i ${srcdir}/32c7d07664dc37765100285d1202d488cd6a27e8.patch
# https://github.com/itstool/itstool/pull/57 - squashed without tests
patch -p1 -i ${srcdir}/57.patch
autoreconf -fi
}

View File

@ -1,25 +0,0 @@
From 1549b6d12eb2f35e5c7f1b1856c21768e92ba794 Mon Sep 17 00:00:00 2001
From: Guido Trentalancia <guido@trentalancia.com>
Date: Wed, 1 Nov 2017 18:23:44 +0100
Subject: [PATCH] Fix a segmentation fault bug introduced with version 2.0.4.
https://github.com/itstool/itstool/issues/17
This fix seems a lot easier than the previous reverted commit.
---
itstool.in | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/itstool.in b/itstool.in
index c1d0585..e492e95 100755
--- a/itstool.in
+++ b/itstool.in
@@ -1048,7 +1048,7 @@ class Document (object):
else:
ctxt.replaceEntities(1)
ctxt.parseDocument()
- trnode = ctxt.doc().getRootElement()
+ trnode = ctxt.doc().getRootElement().copyNode(1)
try:
self._check_errors()
except libxml2.parserError:

View File

@ -1,68 +0,0 @@
Description: Fix the crash from #912099
ITS Tool 2.0.4 crashes when building some documentation, as reported in
#912099. This comes from translations with invalid XML markup, which ITS Tool
fails to merge (which is not abnormal), and to report these issues, needlessly
encodes the original msgstr from unicode to bytes, causing it to be recoded
using the default ascii codec, which fails when the msgstr contains anything
out of ascii.
.
This patch removes the useless decoding, avoiding the failing subsequent
recoding. It also explicitly encodes the output strings to be able to print
them in all cases, even when the output encoding cannot be detected.
Bug: https://github.com/itstool/itstool/issues/25
Bug-Debian: https://bugs.debian.org/912099
Forwarded: https://github.com/itstool/itstool/issues/25
Author: Tanguy Ortolo <tanguy+debian@ortolo.eu>
Last-Update: 2018-12-071
Index: itstool/itstool.in
===================================================================
--- itstool.orig/itstool.in 2018-12-10 18:31:23.762143539 +0100
+++ itstool/itstool.in 2018-12-10 18:38:03.496777117 +0100
@@ -44,9 +44,22 @@
else:
return str(s)
ustr_type = str
+ def pr_str(s):
+ """Return a string that can be safely print()ed"""
+ # Since print works on both bytes and unicode, just return the argument
+ return s
else:
string_types = basestring,
ustr = ustr_type = unicode
+ def pr_str(s):
+ """Return a string that can be safely print()ed"""
+ if isinstance(s, str):
+ # Since print works on str, just return the argument
+ return s
+ else:
+ # print may not work on unicode if the output encoding cannot be
+ # detected, so just encode with UTF-8
+ return unicode.encode(s, 'utf-8')
NS_ITS = 'http://www.w3.org/2005/11/its'
NS_ITST = 'http://itstool.org/extensions/'
@@ -1060,9 +1073,9 @@
if strict:
raise
else:
- sys.stderr.write('Warning: Could not merge %stranslation for msgid:\n%s\n' % (
+ sys.stderr.write(pr_str('Warning: Could not merge %stranslation for msgid:\n%s\n' % (
(lang + ' ') if lang is not None else '',
- msgstr.encode('utf-8')))
+ msgstr)))
self._xml_err = ''
return node
def scan_node(node):
@@ -1087,9 +1100,9 @@
if strict:
raise
else:
- sys.stderr.write('Warning: Could not merge %stranslation for msgid:\n%s\n' % (
+ sys.stderr.write(pr_str('Warning: Could not merge %stranslation for msgid:\n%s\n' % (
(lang + ' ') if lang is not None else '',
- msgstr.encode('utf-8')))
+ msgstr)))
self._xml_err = ''
ctxt.doc().freeDoc()
return node