summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
...
* When content filtering, we may not always want to just discard thebwarsaw2002-10-071-0/+8
| | | | | | | | | message. This is especially the case when debugging your filtering rules. To support this we now have a "filter action" which can be Discard, Reject, Forward to Admin, Preserve on disk. Here, set the default action to discard. Also set the flag that allows owners to preserve filtered messages on disk to true.
* Officially add the archive-url-in-footer item to the wish list.bwarsaw2002-10-071-0/+3
|
* send_i18n_digests(): Chuq politely reminds me that in MM2.0, digestsbwarsaw2002-10-071-0/+2
| | | | | had their Reply-To header pointing at the posting address, and that there was no good reason not to continue this tradition <wink>.
* _setValue(): Tighten the rules for accepting filter_mime_types andbwarsaw2002-10-071-2/+16
| | | | | pass_mime_types entries. They have to look more like a content maintype or maintype/subtype.
* process(): Typo detected by Michael Meltzer; mtype will never be equalbwarsaw2002-10-071-1/+1
| | | | to multipart/alternative (we meant ctype).
* Regenerating the Russian .mo file.bwarsaw2002-10-073-2294/+1823
| | | | Updated Dutch catalog from Danny Terweij.
* Updated catalog files, including mailman.pot, for MM2.1b4 release.bwarsaw2002-10-0729-9155/+11384
|
* Japanese updates from Tokio Kikuchi.bwarsaw2002-10-068-1024/+999
|
* main(): Catch BadListNameError's that can get raised if the proposedbwarsaw2002-10-061-2/+4
| | | | list name is invalid.
* process_request(): Catch BadListNameError's that can get raised if thebwarsaw2002-10-061-2/+6
| | | | proposed list name is invalid.
* ValidateEmail(): Include the string argument in the MMBadEmailErrorbwarsaw2002-10-061-3/+3
| | | | exception.
* Create(): Test the list name to make sure that it will yeild a validbwarsaw2002-10-061-0/+10
| | | | | | | posting address. If not, reject the proposed new list. We use DEFAULT_EMAIL_HOST because at this point we don't know what the host_name of the list is (and in fact the attribute isn't set), but it doesn't matter anyway.
* BadListNameError: New exception for when the list name is invalid.bwarsaw2002-10-061-0/+1
|
* Prepare for beta4bwarsaw2002-10-051-0/+2
|
* process(): Watch for RFC 2047 encoded From headers and decode thebwarsaw2002-10-051-0/+10
| | | | | realname either to a byte string if it contains just ascii characters, or to a unicode.
* subscription_confirm(): Patch by Tokio Kikuchi, we need to pass thebwarsaw2002-10-051-1/+1
| | | | language into the canonstr() call to properly display the fullname.
* checkpointmss2002-10-052-98/+135
|
* _dispose(): Possibly forward the message in two situations: if thebwarsaw2002-10-051-16/+36
| | | | | | | | | message did not bounce match, or if no member addresses could be extracted from it. maybe_forward(): MIME Forward the given message to the list owner + list moderator, if bounce_unrecognized_goes_to_list_owner is true. Otherwise it discards the message. Write a log entry in either case.
* verpdeliver(): I think I've (finally!) nailed the full name encodingbwarsaw2002-10-041-16/+18
| | | | | | | | problem reported a few weeks ago. We do here largely what we do in CookHeaders.py for encoding non-ascii Subject prefixes. First, we convert the string to Unicode (possibly throwing out characters if necessary), then we use the Header class to properly RFC 2047 encode the name.
* uncanonstr(): We always want this to return a byte string, but itbwarsaw2002-10-041-5/+17
| | | | | | | | | | | | needs to be "safe" for the given charset, which will be the charset for the list's language, and the charset the page will be rendered in. If it's a Unicode and can be encoded in the desired charset, fine, just return the encoded byte string. If it's already a byte string encoded in the charset, again that's fine, just return the string. Otherwise, html-ify the string and return it as a byte string.
* membership_options(): Slight code re-arrangement so we're always surebwarsaw2002-10-041-2/+2
| | | | | that the member's name gets encoded safely for the list's language's charset, which is the charset the page is going to be rendered in.
* Updated docstringbwarsaw2002-10-041-32/+31
|
* Better support for "funny" characters in subject prefixes.bwarsaw2002-10-021-42/+32
| | | | | | | | | | | | | | | encode_p(): Removed. _isunicode(): Helper for unicode type testing. prefix_subject(): Better algorithm for handling unicode subject prefixes. Now we always set the Subject header to a Header instance if we're adding the prefix. Always convert the prefix and all previous Subject header bits to unicode (possibly with character replacements) so that the Header class does the right thing, i.e. does the us-ascii, charset hint, utf-8 encoding. Because of this, if the list's charset is us-ascii, we'll substitute iso-8859-1 for a slightly wider character coverage.
* Better support for "funny" characters in subject prefixes.bwarsaw2002-10-021-0/+10
| | | | | | | | | | | | CookHeaders.py will now be aware that subject_prefix can be a unicode string. _setValue(): Capture changes to subject_prefix and canonstr() it, meaning convert html references to unicode characters. getValue(): Capture access of subject_prefix and uncanonstr() it, meaning convert any non-list-charset characters to their html refs for error free display.
* is_administrivia(): Change slightly the way the administrivia searchbwarsaw2002-10-021-3/+9
| | | | | | | | | | text is calculated. Because msg['subject'] could be a Header instance, we need to str-ify it before appending it to the body lines. uncanonstr(): Coerce the return value to a byte string when html ref substitution is done. Note that this code probably needs to be converted to a type test of s (unicode vs. str).
* Update to email package version 2.4bwarsaw2002-10-013-1/+1
|
* run_main(): When running the cgi, redirect stdout to a cStringIO, andbwarsaw2002-10-011-0/+7
| | | | | | | only if the cgi script runs to completion successfully do we write the entire output to the real stdout (connected to the web server). This way, if any exception occurs during the rendering of the page (e.g. in "print doc.Format()") we won't mess up the diagnostics page.
* The usual assortment of updatesbwarsaw2002-09-289-84/+226
|
* Comment out the Chinese codec lookup and message.bwarsaw2002-09-282-49/+25
|
* hope this sounds bettermss2002-09-261-2/+2
|
* checkpointmss2002-09-262-1153/+610
|
* reduce the amount of messages for 'make check'mss2002-09-251-1/+1
|
* got rid of (encodings)mss2002-09-251-2/+1
|
* translatedmss2002-09-252-19/+16
|
* print a nothing-to-do messagemss2002-09-251-0/+8
|
* got rid of a leftovermss2002-09-251-5/+0
|
* translatedmss2002-09-252-13/+10
|
* translatedmss2002-09-252-7/+6
|
* translatedmss2002-09-252-12/+10
|
* oopsmss2002-09-241-67/+0
|
* almost donemss2002-09-242-34/+6
|
* almost donemss2002-09-242-49/+49
|
* finishedmss2002-09-242-11/+12
|
* seems to have it finishedmss2002-09-242-19/+16
|
* minor fixmss2002-09-241-1/+1
|
* punctuation fixmss2002-09-241-1/+1
|
* *** empty log message ***mss2002-09-242-9/+12
|
* finally translatedmss2002-09-242-10/+8
|
* bringing up-to-date with the english variantmss2002-09-243-13/+13
|
* reflect the progress (updated the untranslated part and translated a bit) of ↵mss2002-09-242-27/+49
| | | | options.html
generated by cgit v1.3.1 (git 2.54.0) at 2026-08-06 15:21:40 +0000