Allow extractors to provide message flags

babel/messages/catalog.py

@@ -27,18 +27,6 @@
 __all__ = ['Message', 'Catalog', 'TranslationError']
 
 
-PYTHON_FORMAT = re.compile(r'''
-    \%
-        (?:\(([\w]*)\))?
-        (
-            [-#0\ +]?(?:\*|[\d]+)?
-            (?:\.(?:\*|[\d]+))?
-            [hlL]?
-        )
-        ([diouxXeEfFgGcrs%])
-''', re.VERBOSE)
-
-
 def _parse_datetime_header(value):
     match = re.match(r'^(?P<datetime>.*?)(?P<tzoffset>[+-]\d{4})?$', value)
 
@@ -96,10 +84,6 @@ def __init__(self, id, string=u'', locations=(), flags=(), auto_comments=(),
         self.string = string
         self.locations = list(distinct(locations))
         self.flags = set(flags)
-        if id and self.python_format:
-            self.flags.add('python-format')
-        else:
-            self.flags.discard('python-format')
         self.auto_comments = list(distinct(auto_comments))
         self.user_comments = list(distinct(user_comments))
         if isinstance(previous_id, str):
@@ -196,15 +180,14 @@ def python_format(self):
         """Whether the message contains Python-style parameters.
 
         >>> Message('foo %(name)s bar').python_format
+        False
+        >>> Message('foo %(name)s bar', flags=['python-format']).python_format
         True
-        >>> Message(('foo %(name)s', 'foo %(name)s')).python_format
+        >>> Message(('foo %(name)s', 'foo %(name)s'), flags=['python-format']).python_format
         True
 
         :type:  `bool`"""
-        ids = self.id
-        if not isinstance(ids, (list, tuple)):
-            ids = [ids]
-        return any(PYTHON_FORMAT.search(id) for id in ids)
+        return 'python-format' in self.flags
 
 
 class TranslationError(Exception):

babel/messages/checkers.py

@@ -11,7 +11,8 @@
     :license: BSD, see LICENSE for more details.
 """
 
-from babel.messages.catalog import TranslationError, PYTHON_FORMAT
+from babel.messages.catalog import TranslationError
+from babel.util import PYTHON_FORMAT
 
 
 #: list of format chars that are compatible to each other

babel/messages/extract.py

@@ -22,7 +22,7 @@
 import sys
 from tokenize import generate_tokens, COMMENT, NAME, OP, STRING
 
-from babel.util import parse_encoding, parse_future_flags, pathmatch
+from babel.util import parse_encoding, parse_future_flags, pathmatch, has_python_format
 from textwrap import dedent
 
 
@@ -79,7 +79,7 @@ def extract_from_dir(
     """Extract messages from any source files found in the given directory.
 
     This function generates tuples of the form ``(filename, lineno, message,
-    comments, context)``.
+    comments, context, flags)``.
 
     Which extraction method is used per file is determined by the `method_map`
     parameter, which maps extended glob patterns to extraction method names.
@@ -185,7 +185,7 @@ def check_and_call_extract_file(filepath, method_map, options_map,
     So, given an absolute path to a file `filepath`, we want to check using
     just the relative path from `dirpath` to `filepath`.
 
-    Yields 5-tuples (filename, lineno, messages, comments, context).
+    Yields 6-tuples (filename, lineno, messages, comments, context, flags).
 
     :param filepath: An absolute path to a file that exists.
     :param method_map: a list of ``(pattern, method)`` tuples that maps of
@@ -205,8 +205,8 @@ def check_and_call_extract_file(filepath, method_map, options_map,
     :param strip_comment_tags: a flag that if set to `True` causes all comment
                                tags to be removed from the collected comments.
     :param dirpath: the path to the directory to extract messages from.
-    :return: iterable of 5-tuples (filename, lineno, messages, comments, context)
-    :rtype: Iterable[tuple[str, int, str|tuple[str], list[str], str|None]
+    :return: iterable of 6-tuples (filename, lineno, messages, comments, context)
+    :rtype: Iterable[tuple[str, int, str|tuple[str], list[str], str|None, set[str]]
     """
     # filename is the relative path from dirpath to the actual file
     filename = relpath(filepath, dirpath)
@@ -228,7 +228,10 @@ def check_and_call_extract_file(filepath, method_map, options_map,
             options=options,
             strip_comment_tags=strip_comment_tags
         ):
-            yield (filename, ) + message_tuple
+            if len(message_tuple) == 4:
+                yield (filename, *message_tuple, set())
+            else:
+                yield (filename, *message_tuple)
 
         break
 
@@ -237,7 +240,7 @@ def extract_from_file(method, filename, keywords=DEFAULT_KEYWORDS,
                       comment_tags=(), options=None, strip_comment_tags=False):
     """Extract messages from a specific file.
 
-    This function returns a list of tuples of the form ``(lineno, message, comments, context)``.
+    This function returns a list of tuples of the form ``(lineno, message, comments, context, flags)``.
 
     :param filename: the path to the file to extract messages from
     :param method: a string specifying the extraction method (.e.g. "python")
@@ -250,8 +253,8 @@ def extract_from_file(method, filename, keywords=DEFAULT_KEYWORDS,
     :param strip_comment_tags: a flag that if set to `True` causes all comment
                                tags to be removed from the collected comments.
     :param options: a dictionary of additional options (optional)
-    :returns: list of tuples of the form ``(lineno, message, comments, context)``
-    :rtype: list[tuple[int, str|tuple[str], list[str], str|None]
+    :returns: list of tuples of the form ``(lineno, message, comments, context, flags)``
+    :rtype: list[tuple[int, str|tuple[str], list[str], str|None, set[str]]
     """
     if method == 'ignore':
         return []
@@ -279,7 +282,7 @@ def extract(method, fileobj, keywords=DEFAULT_KEYWORDS, comment_tags=(),
     >>> from io import BytesIO
     >>> for message in extract('python', BytesIO(source)):
     ...     print(message)
-    (3, u'Hello, world!', [], None)
+    (3, u'Hello, world!', [], None, set())
 
     :param method: an extraction method (a callable), or
                    a string specifying the extraction method (.e.g. "python");
@@ -299,8 +302,8 @@ def extract(method, fileobj, keywords=DEFAULT_KEYWORDS, comment_tags=(),
     :param strip_comment_tags: a flag that if set to `True` causes all comment
                                tags to be removed from the collected comments.
     :raise ValueError: if the extraction method is not registered
-    :returns: iterable of tuples of the form ``(lineno, message, comments, context)``
-    :rtype: Iterable[tuple[int, str|tuple[str], list[str], str|None]
+    :returns: iterable of tuples of the form ``(lineno, message, comments, context, flags)``
+    :rtype: Iterable[tuple[int, str|tuple[str], list[str], str|None, set[str]]
     """
     func = None
     if callable(method):
@@ -339,7 +342,7 @@ def extract(method, fileobj, keywords=DEFAULT_KEYWORDS, comment_tags=(),
     results = func(fileobj, keywords.keys(), comment_tags,
                    options=options or {})
 
-    for lineno, funcname, messages, comments in results:
+    for lineno, funcname, messages, comments, *rest in results:
         if funcname:
             spec = keywords[funcname] or (1,)
         else:
@@ -349,6 +352,11 @@ def extract(method, fileobj, keywords=DEFAULT_KEYWORDS, comment_tags=(),
         if not messages:
             continue
 
+        if rest:
+            flags = rest[0]
+        else:
+            flags = set()
+
         # Validate the messages against the keyword's specification
         context = None
         msgs = []
@@ -390,7 +398,7 @@ def extract(method, fileobj, keywords=DEFAULT_KEYWORDS, comment_tags=(),
 
         if strip_comment_tags:
             _strip_comment_tags(comments, comment_tags)
-        yield lineno, messages, comments, context
+        yield lineno, messages, comments, context, flags
 
 
 def extract_nothing(fileobj, keywords, comment_tags, options):
@@ -404,7 +412,7 @@ def extract_python(fileobj, keywords, comment_tags, options):
     """Extract messages from Python source code.
 
     It returns an iterator yielding tuples in the following form ``(lineno,
-    funcname, message, comments)``.
+    funcname, message, comments, flags)``.
 
     :param fileobj: the seekable, file-like object the messages should be
                     extracted from
@@ -422,6 +430,7 @@ def extract_python(fileobj, keywords, comment_tags, options):
     translator_comments = []
     in_def = in_translator_comments = False
     comment_tag = None
+    flags = set()
 
     encoding = parse_encoding(fileobj) or options.get('encoding', 'UTF-8')
     future_flags = parse_future_flags(fileobj, encoding)
@@ -468,6 +477,9 @@ def extract_python(fileobj, keywords, comment_tags, options):
                 else:
                     messages.append(None)
 
+                if has_python_format(message for message in messages if message):
+                    flags.add("python-format")
+
                 if len(messages) > 1:
                     messages = tuple(messages)
                 else:
@@ -479,12 +491,13 @@ def extract_python(fileobj, keywords, comment_tags, options):
                     translator_comments = []
 
                 yield (message_lineno, funcname, messages,
-                       [comment[1] for comment in translator_comments])
+                       [comment[1] for comment in translator_comments], flags)
 
                 funcname = lineno = message_lineno = None
                 call_stack = -1
                 messages = []
                 translator_comments = []
+                flags = set()
                 in_translator_comments = False
                 if nested:
                     funcname = value
@@ -610,7 +623,7 @@ def extract_javascript(fileobj, keywords, comment_tags, options):
 
                 if messages is not None:
                     yield (message_lineno, funcname, messages,
-                           [comment[1] for comment in translator_comments])
+                           [comment[1] for comment in translator_comments], set())
 
                 funcname = message_lineno = last_argument = None
                 concatenate_next = False

babel/messages/frontend.py

@@ -502,14 +502,14 @@ def callback(filename, method, options):
                         strip_comment_tags=self.strip_comments,
                         directory_filter=self.directory_filter,
                     )
-                for filename, lineno, message, comments, context in extracted:
+                for filename, lineno, message, comments, context, flags in extracted:
                     if os.path.isfile(path):
                         filepath = filename  # already normalized
                     else:
                         filepath = os.path.normpath(os.path.join(path, filename))
 
                     catalog.add(message, None, [(filepath, lineno)],
-                                auto_comments=comments, context=context)
+                                auto_comments=comments, context=context, flags=flags)
 
             self.log.info('writing PO template file to %s', self.output_file)
             write_po(outfile, catalog, width=self.width,

babel/messages/pofile.py

@@ -475,7 +475,7 @@ def write_po(fileobj, catalog, width=76, no_location=False, omit_header=False,
 
     >>> catalog = Catalog()
     >>> catalog.add(u'foo %(name)s', locations=[('main.py', 1)],
-    ...             flags=('fuzzy',))
+    ...             flags=('fuzzy', 'python-format'))
     <Message...>
     >>> catalog.add((u'bar', u'baz'), locations=[('main.py', 3)])
     <Message...>

babel/util.py

@@ -263,3 +263,21 @@ def dst(self, dt):
 
 def _cmp(a, b):
     return (a > b) - (a < b)
+
+
+PYTHON_FORMAT = re.compile(r'''
+    \%
+        (?:\(([\w]*)\))?
+        (
+            [-#0\ +]?(?:\*|[\d]+)?
+            (?:\.(?:\*|[\d]+))?
+            [hlL]?
+        )
+        ([diouxXeEfFgGcrs%])
+''', re.VERBOSE)
+
+
+def has_python_format(ids):
+    if isinstance(ids, str):
+        ids = [ids]
+    return any(PYTHON_FORMAT.search(id) for id in ids)

tests/messages/data/project/i18n/fi_BUGGY/LC_MESSAGES/messages.po

@@ -1,5 +1,6 @@
 msgid ""
 msgstr ""
 
+#, python-format
 msgid "bar %(sign)s"
 msgstr "tanko %(merkki)s"

tests/messages/test_catalog.py

@@ -23,23 +23,6 @@
 
 class MessageTestCase(unittest.TestCase):
 
-    def test_python_format(self):
-        assert catalog.PYTHON_FORMAT.search('foo %d bar')
-        assert catalog.PYTHON_FORMAT.search('foo %s bar')
-        assert catalog.PYTHON_FORMAT.search('foo %r bar')
-        assert catalog.PYTHON_FORMAT.search('foo %(name).1f')
-        assert catalog.PYTHON_FORMAT.search('foo %(name)3.3f')
-        assert catalog.PYTHON_FORMAT.search('foo %(name)3f')
-        assert catalog.PYTHON_FORMAT.search('foo %(name)06d')
-        assert catalog.PYTHON_FORMAT.search('foo %(name)Li')
-        assert catalog.PYTHON_FORMAT.search('foo %(name)#d')
-        assert catalog.PYTHON_FORMAT.search('foo %(name)-4.4hs')
-        assert catalog.PYTHON_FORMAT.search('foo %(name)*.3f')
-        assert catalog.PYTHON_FORMAT.search('foo %(name).*f')
-        assert catalog.PYTHON_FORMAT.search('foo %(name)3.*f')
-        assert catalog.PYTHON_FORMAT.search('foo %(name)*.*f')
-        assert catalog.PYTHON_FORMAT.search('foo %()s')
-
     def test_translator_comments(self):
         mess = catalog.Message('foo', user_comments=['Comment About `foo`'])
         self.assertEqual(mess.user_comments, ['Comment About `foo`'])
@@ -322,8 +305,10 @@ def test_message_pluralizable():
 
 
 def test_message_python_format():
-    assert catalog.Message('foo %(name)s bar').python_format
-    assert catalog.Message(('foo %(name)s', 'foo %(name)s')).python_format
+    assert catalog.Message('foo %(name)s bar', flags={'python-format'}).python_format
+    assert catalog.Message(('foo %(name)s', 'foo %(name)s'), flags={'python-format'}).python_format
+    assert not catalog.Message('foo %(name)s bar').python_format
+    assert not catalog.Message(('foo %(name)s', 'foo %(name)s')).python_format
 
 
 def test_catalog():