summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
...
* process(): Dan Mick points out we probably shouldn't add separators ifbwarsaw2002-08-151-2/+2
| | | | there's no header (or footer).
* process(): Might as well add a link to the member's preferences page.bwarsaw2002-08-151-1/+2
|
* Might as well add a link to the member's preferences page.bwarsaw2002-08-151-0/+1
|
* process(): If we're adding the header and footer by stringbwarsaw2002-08-151-1/+7
| | | | concatenation, make sure there's a newline separating each part.
* process(): Correct typo: "multipart" -> "multipart/alternative".bwarsaw2002-08-151-5/+1
| | | | | reset_payload(): Don't hack on the trailing newline here. Probably do it in Decorate.py instead.
* GetConfigInfo(): Add a note that you probably will want to addbwarsaw2002-08-151-1/+5
| | | | | | `multipart' to the pass filter. We still won't set DEFAULT_PASS_MIME_TYPES = ['multipart'] because we want to leave the pass filter empty by default.
* NewVars(): pass_mime_types needs to get its value from the mm_cfgbwarsaw2002-08-151-1/+1
| | | | default.
* Added new peoplepioppo2002-08-151-1/+8
|
* InitVars(): Added pass_mime_types.bwarsaw2002-08-151-9/+10
| | | | Also, whitespace normalization.
* DEFAULT_PASS_MIME_TYPES: Default value for pass_mime_types.bwarsaw2002-08-151-5/+8
| | | | Also, whitespace normalization.
* DATA_FILE_VERSION: Bump to pick up pass_mime_types.bwarsaw2002-08-152-8/+8
| | | | | | NewVars(): Add pass_mime_types if missing. Also, whitespace normalization and PyChecker cleanup.
* process(): Add support for a whitelisting feature contributed by Danbwarsaw2002-08-151-32/+58
| | | | | | | | | | | | | | | Mick, hacked by Barry (so any bugs are my fault). The idea is that after the filter_mime_types pass is made, if there are any pass_mime_types defined, only parts that match these types are allowed through. reset_payload(): New function, factorization. filter_parts(): Add passtypes to the argument list and do the passtypes filtering. Also, use the new .get_content_type() / .get_content_maintype() interface everywhere.
* Whitespace normalization.bwarsaw2002-08-151-3/+3
|
* GetConfigInfo(): Rewrite large portions of this text to give morebwarsaw2002-08-151-18/+51
| | | | | | | | | | | | | background on the content filtering process, make descriptions more consistent, and hopefully more useful. Also, added a description of the pass_mime_types variable. This is a whitelisting feature contributed by Dan Mick, hacked by Barry (so any bugs are my fault). The idea is that after the filter_mime_types pass is made, if there are any pass_mime_types defined, only parts that match these types are allowed through. _setValue(), getValue(): Add support for pass_mime_types.
* handleForm(): A `return' should have been a `continue'. This fixesbwarsaw2002-08-141-1/+1
| | | | value setting when submitting an option's details page.
* options_help(): When displaying a variable detail, use the humanbwarsaw2002-08-141-1/+2
| | | | readable category name in the "return to <category> page" text.
* Stop runaway restarts by imposing a maximum restart value (currentlybwarsaw2002-08-141-4/+14
| | | | | | 10). If a qrunner exits with something other than SIGINT, its restart counter gets incremented and if that is greater than MAX_RESTARTS, that qrunner won't be restarted.
* Catalog updates.wilane2002-08-142-29/+66
|
* and also the .mo is updated toodanohnesorg2002-08-131-0/+0
|
* Translated strings about encodings of ascii characters etc.danohnesorg2002-08-131-50/+41
|
* VERSION: Set to 2.1b3+bwarsaw2002-08-131-1/+1
|
* Rework the directory layout for attachments. There were two problemsbwarsaw2002-08-131-40/+63
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | with the old approach, as pointed out by Michael Meltzer: - it was broken <wink> - it was still susceptible to inode overload The fixes are to put attachments one more directory level down, where the path is archives/private/<listname>/attachments/<YYYYMMDD>/<msgidhash> where YYYYMMDD is the date of the message encoded as 8 digits, and the msgidhash is the first 2 and last 2 octets of the sha hash of the outer message's Message-ID. Any name collisions inside that directory (e.g. a message that contains two images with the same filename), are resolved by finding a unique filebase extension. Specific changes include: calculate_attachments_dir(): Factor out the calculate of the attachments directory so that e.g. Michael can use it in his MimeDel.py hacks. process(): Be sure to pass the appropriate arguments to save_attachment() based on its new signature. makedirs(): Factor out the creation and mod settings of the attachment dir, and all subdirs. save_attachment(): Pass the attachments directory in the arguments. Fix calculation of the url and the saving of the attachment data.
* Expand on the note about Python packages. It seems like every Linuxbwarsaw2002-08-121-4/+11
| | | | | distro needs the Python development package to work correctly, although I'm unsure why that is.
* Japanese updates from Tokio Kikuchibwarsaw2002-08-127-252/+327
|
* Japanese template from Tokio Kikuchi.bwarsaw2002-08-120-0/+0
|
* intermediatebwarsaw2002-08-121-0/+10
|
* Oops, forgot thispioppo2002-08-101-0/+0
|
* Aligned to 2.1b3pioppo2002-08-101-158/+193
|
* The usual updatesbwarsaw2002-08-106-57/+177
|
* More namesbwarsaw2002-08-101-0/+1
|
* Updated catalogs for 2.1b3.bwarsaw2002-08-1027-4197/+6447
|
* What's new for 2.1 beta 3.bwarsaw2002-08-101-0/+69
|
* Fixed some string markup for i18n.bwarsaw2002-08-101-3/+3
|
* A translation checker written by Simone Piunno.bwarsaw2002-08-100-0/+0
| | | | | I forgot to check this in. Translators may want to use this to find errors in their templates and catalogs.
* intermediatebwarsaw2002-08-101-0/+389
|
* process(): Mentor Cana reports a crash when the subject header is anbwarsaw2002-08-101-10/+4
| | | | | | | | | | | email.Header.Header instance; it won't have a .startswith() method. The real problem is that the attempt to send back the original, non-prefixed header was way too naive. Fortunately, the msgdata will contain the original subject if it was munged, so just use that. Question: should we try to send back the decoded header?
* prefix_subject(): Elaborate a comment.bwarsaw2002-08-091-1/+3
|
* encode_p(): New helper function.bwarsaw2002-08-091-3/+29
| | | | | | prefix_subject(): Only encode the subject prefix if it's necessary because it contains non-ASCII characters, or the list owner has requested to do so (via encode_ascii_prefixes).
* GetConfigInfo(): Add description of encode_ascii_prefixes.bwarsaw2002-08-091-0/+25
|
* NewVars(): Add encode_ascii_prefixes if missing.bwarsaw2002-08-091-0/+6
|
* DATA_FILE_VERSION: Bump to pick up encode_ascii_prefixes.bwarsaw2002-08-091-1/+1
|
* InitVars(): Set the new config variable encode_ascii_prefixes whichbwarsaw2002-08-091-0/+6
| | | | | controls whether the Subject: prefix gets encoded when it's ASCII and the list's preferred language uses a non-ASCII character set.
* Always adding new features to the list. ;)bwarsaw2002-08-091-1/+11
|
* Debian users need to install the python2.2-dev package if they'rebwarsaw2002-08-091-0/+7
| | | | using Python 2.2.
* More updatesbwarsaw2002-08-091-3/+6
|
* Get ready for 2.1b3 release.bwarsaw2002-08-091-2/+2
|
* subscription_prompt(): Don't provide the choice of digests orbwarsaw2002-08-091-3/+6
| | | | | non-digests unless the subscriber actually has a choice! Closes SF bug # 563091 (not the u/i part).
* Added a -f / --force switch to force notifications to members who havebwarsaw2002-08-091-11/+18
| | | | already been notified within the bounce_you_are_disabled_warnings_interval.
* __oneloop(): Renamed to _oneloop() -- i.e. single leading underscore,bwarsaw2002-08-091-6/+6
| | | | so that it can be overridden in derived classes.
* OWNER_PIPELINE: Moved to Defaults.py.inbwarsaw2002-08-091-8/+1
|
generated by cgit v1.3.1 (git 2.54.0) at 2026-08-09 19:11:34 +0000