get_obs_local_part fails to handle empty local part

@warsaw

BPO	42484
Nosy	@warsaw, @bitdancer, @ZackerySpytz, @dicksonch
PRs	gh-86650: get_obs_local_part() fails to handle empty local part #24669

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = None
created_at = <Date 2020-11-27.12:36:59.942>
labels = ['type-bug', '3.8', 'expert-email', '3.9']
title = 'get_obs_local_part fails to handle empty local part'
updated_at = <Date 2021-02-28.16:44:21.314>
user = 'https://github.com/dicksonch'

bugs.python.org fields:

activity = <Date 2021-02-28.16:44:21.314>
actor = 'ZackerySpytz'
assignee = 'none'
closed = False
closed_date = None
closer = None
components = ['email']
creation = <Date 2020-11-27.12:36:59.942>
creator = 'dxn126'
dependencies = []
files = []
hgrepos = []
issue_num = 42484
keywords = ['patch']
message_count = 2.0
messages = ['381947', '382153']
nosy_count = 4.0
nosy_names = ['barry', 'r.david.murray', 'ZackerySpytz', 'dxn126']
pr_nums = ['24669']
priority = 'normal'
resolution = None
stage = 'patch review'
status = 'open'
superseder = None
type = 'behavior'
url = 'https://bugs.python.org/issue42484'
versions = ['Python 3.8', 'Python 3.9']

Linked PRs

parse_message_id in the email module crashes with bogus message-id

Having a Message-ID '<[>' gives me an IndexError: list index out of range

This happens when

creating an EmailMessage with the said Message-ID
msg = EmailMessage()
msg['Message-ID'] = '<[>'
accessing the bogus Message-ID through
msg.items()
or
msg.get('Message-ID')

this doesn't happen with python 3.6 or 3.7 when MessageIDHeader didn't exist

3.8/Lib/email/headerregistry.py line 542

_default_header_map = {
    ....
    'message-id': MessageIDHeader,
    }

Traceback (most recent call last):
  File "/usr/lib/python3.8/email/_header_value_parser.py", line 2069, in get_msg_id
    token, value = get_dot_atom_text(value)
  File "/usr/lib/python3.8/email/_header_value_parser.py", line 1334, in get_dot_atom_text
    raise errors.HeaderParseError("expected atom at a start of "
email.errors.HeaderParseError: expected atom at a start of dot-atom-text but found '[>'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "main.py", line 4, in <module>
    msg['Message-ID'] = '<[>'
  File "/usr/lib/python3.8/email/message.py", line 409, in __setitem__
    self._headers.append(self.policy.header_store_parse(name, val))
  File "/usr/lib/python3.8/email/policy.py", line 148, in header_store_parse
    return (name, self.header_factory(name, value))
  File "/usr/lib/python3.8/email/headerregistry.py", line 607, in __call__
    return self[name](name, value)
  File "/usr/lib/python3.8/email/headerregistry.py", line 202, in __new__
    cls.parse(value, kwds)
  File "/usr/lib/python3.8/email/headerregistry.py", line 535, in parse
    kwds['parse_tree'] = parse_tree = cls.value_parser(value)
  File "/usr/lib/python3.8/email/_header_value_parser.py", line 2126, in parse_message_id
    token, value = get_msg_id(value)
  File "/usr/lib/python3.8/email/_header_value_parser.py", line 2073, in get_msg_id
    token, value = get_obs_local_part(value)
  File "/usr/lib/python3.8/email/_header_value_parser.py", line 1516, in get_obs_local_part
    if (obs_local_part[0].token_type == 'dot' or
IndexError: list index out of range

as you can see in the traceback
get_msg_id() calls get_obs_local_part()
and in get_obs_local_part(), you have this

def get_obs_local_part(value):

    obs_local_part = ObsLocalPart()

    while value and (value[0]=='\\' or value[0] not in PHRASE_ENDS):
        ...
    if (obs_local_part[0].token_type == 'dot':
        ...

if value does not satisfy the condition in the while loop,
this gives an IndexError as obs_local_part is empty
(the value in my example is '[>' from the message id '<[>')

shouldn't we have a proper Error or default back to no parsing if parsing fails?
There's no way of bypassing the parser and getting the Message-ID and
I can't even handle the error with a try catch

Yep, you've found another in a category of bugs that have shown up in the parser: places where there is a missing check for there being any value at all before checking character [0].

In this case, the fix should be to add

    if not obs_local_part:
        return obs_local_part, value

just before the if that is blowing up.

I am currently experiencing this issue, has this been looked at yet? Is there a workaround?

Duplicate of #105802

pythongh-86650: Fix IndexError when parse emails with invalid Message-ID

gh-86650: Fix IndexError when parse emails with invalid Message-ID (G…

pythongh-86650: Fix IndexError when parse emails with invalid Message…

[3.12] gh-86650: Fix IndexError when parse emails with invalid Messag…

pythongh-86650: Fix IndexError when parse emails with invalid Message…

Navigation Menu

Repository navigation

get_obs_local_part fails to handle empty local part #86650

Linked PRs

Metadata

Assignees

Labels

Projects

Milestone

Relationships

Development

get_obs_local_part fails to handle empty local part #86650

Description

Linked PRs

Activity

dicksonch commented on Nov 27, 2020

bitdancer commented on Nov 30, 2020

lesleyxyz commented on Feb 27, 2023

fsc-eriker commented on Feb 9, 2024

Metadata

Metadata

Assignees

Labels

Projects

Milestone

Relationships

Development

Issue actions