Repository navigation
pybabel extract crashes when it encounters a byte string #1190
Description
Activity
Agreed. We should warn about that and leave it non-extracted.
Reacted by Tomas R.I'll put it on my todo list for this weekend
- added a commit that references this issue
on Jul 29, 2026 Root cause and fix:
_parse_python_string()is annotated as-> str | None, but when it encounters a byte-string literal (e.g.b'foo'), it evaluates to anast.Constantwhose.valueis abytesobject and returns bytes unchanged. The caller inextract_pythonappends this tobuf(a list of string fragments), then calls''.join(buf)— which fails becausestr.join()requires all items to bestr:# extract.py — _parse_python_string if isinstance(body, ast.Constant): return body.value # returns b'foo' for byte-string literals # extract_python buf.append(val) # buf = [b'foo'] messages.append(''.join(buf)) # TypeError: expected str, got bytes
Fix: Add an
isinstance(body.value, str)guard in_parse_python_stringso that byte-string literals are silently skipped (returningNone), matching the behaviour for all other non-extractable expressions:if isinstance(body, ast.Constant): - return body.value + if isinstance(body.value, str): + return body.value
Byte strings cannot be used as
gettextmessages anyway, so ignoring them is the correct behaviour. The crash is avoided and callers calling_(b'foo')will simply have no entry extracted, with no error.I have a PR ready.
- added a commit that references this issue
on Oct 3, 2026
(Happens on current master)
Sample file:
now run
pybabel extract test.py -o messages.potand you'll get this error:It seems like we're trying to call
.join()on a byte string. I suggest just fixing the crash but keeping byte strings disallowed - they won't work withgettextanyway because gettext uses normal strings internally.