GeezerVison is showing github.com — Document IMAP4.append() message type and optional arguments · Issue #149962 · python/cpython
Document IMAP4.append() message type and optional arguments #149962DescriptionThe IMAP4.append() documentation currently says only:
Append message to named mailbox.
It now documents the flags argument, but the remaining argument behavior is still easier to discover from Lib/imaplib.py than from the library docs.
In the implementation:
mailbox is treated as optional in practice: a false value is replaced with 'INBOX'.
flags and date_time may be None, in which case they are omitted from the command.
message is not optional. It is passed to MapCRLF.sub(CRLF, message), where MapCRLF is a bytes regular expression, so message must be a bytes-like object. Passing a str raises TypeError before the command is sent.
The method docstring already hints at this by saying "All args except 'message' can be None", but that detail is not in Doc/library/imaplib.rst.
This is related to the broader documentation issue in #68215, but this issue is intentionally narrower: clarify the public documentation for IMAP4.append() only.
Suggested documentation scope:
- State that message must be a bytes-like object.
- State which arguments may be
None.
- Clarify that
mailbox=None (or another false value) uses INBOX.
- Clarify that
date_time=None omits the internal date argument.
This should stay separate from the existing CR/LF-normalization behavior tracked in #49680.
Linked PRs
Activitygvanrossum commented on May 17, 2026 AFAIK it is okay to submit doc PRs without an Issue now. savagemechanic commented on May 17, 2026 ContributorAuthor More actions
AFAIK it is okay to submit doc PRs without an Issue now.
Understood 👊 Thanks Guido picnixz commented on May 19, 2026 While it is ok not to create issues most of the time, it is also good for tracking discussions if any (like, if the docs are changed or reorganized, having an issue is good; if it is only for improving, amending slightly, or fixing inconsistencies or incompleteness, no issue is needed).
(In this case, there was no need for an issue but since one was created, we can keep it) savagemechanic commented on May 20, 2026 ContributorAuthor More actions Thank you for taking your time to break that down for me :-)
MetadataMetadataStatusShow more project fields
DevelopmentNo branches or pull requestsIssue actions
|