itstool/trunk/0001-Fix-the-crash-from-912099.patch

   1 From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
   2 From: Tanguy Ortolo <tanguy+debian@ortolo.eu>
   3 Date: Fri, 7 Dec 2018 00:00:00 +0000
   4 Subject: [PATCH] Fix the crash from #912099
   5
   6 ITS Tool 2.0.4 crashes when building some documentation, as reported in #912099.
   7 This comes from translations with invalid XML markup, which ITS Tool fails to
   8 merge (which is not abnormal), and to report these issues, needlessly encodes
   9 the original msgstr from unicode to bytes, causing it to be recoded using the
  10 default ascii codec, which fails when the msgstr contains anything out of ascii.
  11
  12 This patch removes the useless decoding, avoiding the failing subsequent
  13 recoding. It also explicitly encodes the output strings to be able to print them
  14 in all cases, even when the output encoding cannot be detected.
  15
  16 Bug: https://github.com/itstool/itstool/issues/25
  17 Bug-Debian: https://bugs.debian.org/912099
  18 Forwarded: https://github.com/itstool/itstool/issues/25
  19 ---
  20  itstool.in | 21 +++++++++++++++++----
  21  1 file changed, 17 insertions(+), 4 deletions(-)
  22
  23 diff --git a/itstool.in b/itstool.in
  24 index c21ad4bfeb86..f34673581c88 100755
  25 --- a/itstool.in
  26 +++ b/itstool.in
  27 @@ -44,9 +44,22 @@ if PY3:
  28          else:
  29              return str(s)
  30      ustr_type = str
  31 +    def pr_str(s):
  32 +        """Return a string that can be safely print()ed"""
  33 +        # Since print works on both bytes and unicode, just return the argument
  34 +        return s
  35  else:
  36      string_types = basestring,
  37      ustr = ustr_type = unicode
  38 +    def pr_str(s):
  39 +        """Return a string that can be safely print()ed"""
  40 +        if isinstance(s, str):
  41 +            # Since print works on str, just return the argument
  42 +            return s
  43 +        else:
  44 +            # print may not work on unicode if the output encoding cannot be
  45 +            # detected, so just encode with UTF-8
  46 +            return unicode.encode(s, 'utf-8')
  47
  48  NS_ITS = 'http://www.w3.org/2005/11/its'
  49  NS_ITST = 'http://itstool.org/extensions/'
  50 @@ -1077,36 +1090,36 @@ class Document (object):
  51              if strict:
  52                  raise
  53              else:
  54 -                sys.stderr.write('Warning: Could not merge %stranslation for msgid:\n%s\n' % (
  55 +                sys.stderr.write(pr_str('Warning: Could not merge %stranslation for msgid:\n%s\n' % (
  56                          (lang + ' ') if lang is not None else '',
  57 -                        msgstr.encode('utf-8')))
  58 +                        msgstr)))
  59                  self._xml_err = ''
  60                  return node
  61          def scan_node(node):
  62              children = [child for child in xml_child_iter(node)]
  63              for child in children:
  64                  if child.type != 'element':
  65                      continue
  66                  if child.ns() is not None and child.ns().content == NS_BLANK:
  67                      ph_node = msg.get_placeholder(child.name).node
  68                      if self.has_child_elements(ph_node):
  69                          self.merge_translations(translations, None, ph_node, strict=strict)
  70                          newnode = ph_node.copyNode(1)
  71                          newnode.setTreeDoc(self._doc)
  72                          child.replaceNode(newnode)
  73                      else:
  74                          repl = self.get_translated(ph_node, translations, strict=strict, lang=lang)
  75                          child.replaceNode(repl)
  76                  scan_node(child)
  77          try:
  78              scan_node(trnode)
  79          except:
  80              if strict:
  81                  raise
  82              else:
  83 -                sys.stderr.write('Warning: Could not merge %stranslation for msgid:\n%s\n' % (
  84 +                sys.stderr.write(pr_str('Warning: Could not merge %stranslation for msgid:\n%s\n' % (
  85                      (lang + ' ') if lang is not None else '',
  86 -                    msgstr.encode('utf-8')))
  87 +                    msgstr)))
  88                  self._xml_err = ''
  89                  ctxt.doc().freeDoc()
  90                  return node