gh-148518 fix index error in local part attribute (#148522)

As part of fixing bpo-27931 code was introduced to get_bare_quoted_string
that added an empty Terminal if the quoted string was empty.  This isn't
the best answer in terms of the parse tree; we really want the token
list to be empty in that case.  But having it be empty resulted in
local_part raising the index error.  We find that same problem if we
try to parse an address consisting of a single dquote.  By fixing
local_part to not raise on an empty token list, we can have the
bare_quoted_string code correctly return an empty token list for
the empty string cases (two dquotes or a single dquote as the
entire addrespec, at the end of a line).
This commit is contained in:
R. David Murray
2026-04-30 17:41:43 -04:00
committed by GitHub
parent 5110738779
commit bdbb55c403
3 changed files with 46 additions and 6 deletions
+3 -4
View File
@@ -639,11 +639,11 @@ class LocalPart(TokenList):
for tok in self[0] + [DOT]:
if tok.token_type == 'cfws':
continue
if (last_is_tl and tok.token_type == 'dot' and
if (last_is_tl and tok.token_type == 'dot' and last and
last[-1].token_type == 'cfws'):
res[-1] = TokenList(last[:-1])
is_tl = isinstance(tok, TokenList)
if (is_tl and last.token_type == 'dot' and
if (is_tl and last.token_type == 'dot' and tok and
tok[0].token_type == 'cfws'):
res.append(TokenList(tok[1:]))
else:
@@ -1249,8 +1249,7 @@ def get_bare_quoted_string(value):
bare_quoted_string = BareQuotedString()
value = value[1:]
if value and value[0] == '"':
token, value = get_qcontent(value)
bare_quoted_string.append(token)
return bare_quoted_string, value[1:]
while value and value[0] != '"':
if value[0] in WSP:
token, value = get_fws(value)
+39 -2
View File
@@ -1271,12 +1271,12 @@ class TestAddressHeader(TestHeaderBase):
'example.com',
None),
}
# XXX: Need many more examples, and in particular some with names in
# trailing comments, which aren't currently handled. comments in
# general are not handled yet.
}
def example_as_address(self, source, defects, decoded, display_name,
addr_spec, username, domain, comment):
h = self.make_header('sender', source)
@@ -1294,6 +1294,43 @@ class TestAddressHeader(TestHeaderBase):
# XXX: we have no comment support yet.
#self.assertEqual(a.comment, comment)
example_broken_header_params = {
'just_dquote':
('"',
[errors.InvalidHeaderDefect]*2,
'<>',
'',
'<>',
'',
'',
),
}
def example_broken_header_as_address(
self,
source,
defects,
decoded,
display_name,
addr_spec,
username,
domain,
):
h = self.make_header('sender', source)
self.assertEqual(h, decoded)
self.assertDefectsEqual(h.defects, defects)
a = h.address
self.assertEqual(str(a), decoded)
self.assertEqual(len(h.groups), 1)
self.assertEqual([a], list(h.groups[0].addresses))
self.assertEqual([a], list(h.addresses))
self.assertEqual(a.display_name, display_name)
self.assertEqual(a.addr_spec, addr_spec)
self.assertEqual(a.username, username)
self.assertEqual(a.domain, domain)
def example_as_group(self, source, defects, decoded, display_name,
addr_spec, username, domain, comment):
source = 'foo: {};'.format(source)
@@ -0,0 +1,4 @@
If an email containing an address header that ended in an open double quote
was parsed with a non-``compat32`` policy, accessing the ``username`` attribute
of the mailbox accessed through that header object would result in an
``IndexError``. It now correctly returns an empty string as the result.