Detect incorrectly delimited strings

Strings should be delimited on both ends by double quotes, but this
is currently not being been detected, and content is simply being
incorrectly trimmed.

This commit adds a check for each string to verify it starts and ends
with a double quote character, issuing a warning/error if that's not the
case (and fixing it as appropriate).

A few new test cases have been added to check that the lack of double
quotes to delimit strings issues errors as expected.

Signed-off-by: Rodrigo Tobar <[email protected]>

Author: rtobar

babel/messages/pofile.py

@@ -278,6 +278,17 @@ def _process_keyword_line(self, lineno, line, obsolete=False) -> None:
 
         self.obsolete = obsolete
 
+        def _normalized_string(s):
+            # whole lines are already stripped on both ends
+            s = s.lstrip()
+            if s[0] != '"':
+                self._invalid_pofile(line, lineno, "String must be delimited on the left by double quotes")
+                s = '"' + s
+            if s[-1] != '"':
+                self._invalid_pofile(line, lineno, "String must be delimited on the left by double quotes")
+                s += '"'
+            return _NormalizedString(s)
+
         # The line that has the msgid is stored as the offset of the msg
         # should this be the msgctxt if it has one?
         if keyword == 'msgid':
@@ -286,20 +297,20 @@ def _process_keyword_line(self, lineno, line, obsolete=False) -> None:
         if keyword in ['msgid', 'msgid_plural']:
             self.in_msgctxt = False
             self.in_msgid = True
-            self.messages.append(_NormalizedString(arg))
+            self.messages.append(_normalized_string(arg))
 
         elif keyword == 'msgstr':
             self.in_msgid = False
             self.in_msgstr = True
             if arg.startswith('['):
                 idx, msg = arg[1:].split(']', 1)
-                self.translations.append([int(idx), _NormalizedString(msg)])
+                self.translations.append([int(idx), _normalized_string(msg)])
             else:
-                self.translations.append([0, _NormalizedString(arg)])
+                self.translations.append([0, _normalized_string(arg)])
 
         elif keyword == 'msgctxt':
             self.in_msgctxt = True
-            self.context = _NormalizedString(arg)
+            self.context = _normalized_string(arg)
 
     def _process_string_continuation_line(self, line, lineno) -> None:
         if self.in_msgid:
@@ -311,6 +322,10 @@ def _process_string_continuation_line(self, line, lineno) -> None:
         else:
             self._invalid_pofile(line, lineno, "Got line starting with \" but not in msgid, msgstr or msgctxt")
             return
+        assert line[0] == '"'
+        if line[-1] != '"':
+            self._invalid_pofile(line, lineno, "String must be delimited by double quotes")
+            line += '"'
         s.append(line)
 
     def _process_comment(self, line) -> None:

tests/messages/test_pofile.py

@@ -529,6 +529,22 @@ def test_abort_invalid_po_file(self):
             Pour toute question, veuillez communiquer avec Fulano a [email protected]
             "
             '''
+        invalid_po_3 = '''
+            msgid *A"
+            msgstr "B"
+        '''
+
+        invalid_po_4 = '''
+            msgid "A"
+            msgstr "B*
+        '''
+
+        invalid_po_5 = '''
+            msgid "A"
+            msgstr ""
+            "*
+        '''
+
         # Catalog not created, throws Unicode Error
         buf = StringIO(invalid_po)
         output = pofile.read_po(buf, locale='fr', abort_invalid=False)
@@ -544,6 +560,12 @@ def test_abort_invalid_po_file(self):
         output = pofile.read_po(buf, locale='fr', abort_invalid=False)
         assert isinstance(output, Catalog)
 
+        # Catalog not created, aborted with PoFileError
+        for incorrectly_delimited_po_string in (invalid_po_3, invalid_po_4, invalid_po_5):
+            buf = StringIO(incorrectly_delimited_po_string)
+            with pytest.raises(pofile.PoFileError):
+                pofile.read_po(buf, locale='es', abort_invalid=True)
+
     def test_invalid_pofile_with_abort_flag(self):
         parser = pofile.PoFileParser(None, abort_invalid=True)
         lineno = 10