Mailman Core merge requestshttps://staging.gitlab.com/mailman/mailman/-/merge_requests2017-10-03T14:29:02Zhttps://staging.gitlab.com/mailman/mailman/-/merge_requests/323Added (\+.*)? to -bounces and -confirm regexp maps for Postfix.2017-10-03T14:29:02Zusername-removed-107958Added (\+.*)? to -bounces and -confirm regexp maps for Postfix.Fixes #401
recipient_delimiter = + is not sufficient for regexp tables.Fixes #401
recipient_delimiter = + is not sufficient for regexp tables.3.2https://staging.gitlab.com/mailman/mailman/-/merge_requests/321There is no smtp-failure log anymore2017-09-28T15:21:27Zusername-removed-56772There is no smtp-failure log anymore3.2username-removed-56772username-removed-56772https://staging.gitlab.com/mailman/mailman/-/merge_requests/320Make sphinx-build ignore .pc directory2017-10-01T00:53:58Zusername-removed-116260Make sphinx-build ignore .pc directoryThe quilt patch management tool might create the '.pc' directory during
the package build process. In case that '.rst' files are patched by
quilt, they end up in '.pc' as well. A sphinxdoc recusively scans the
sources for '.rst' files, t...The quilt patch management tool might create the '.pc' directory during
the package build process. In case that '.rst' files are patched by
quilt, they end up in '.pc' as well. A sphinxdoc recusively scans the
sources for '.rst' files, this causes unwanted behaviour. This patch
adds '.pc' to the list of excluded patterns.
Fixes: #4033.2https://staging.gitlab.com/mailman/mailman/-/merge_requests/319Consider /etc/mailman3/mailman.cfg as config file2017-09-28T16:37:55Zusername-removed-116260Consider /etc/mailman3/mailman.cfg as config file* Debian (and probably other distributions in the future) installs the
'mailman.cfg' config file to '/etc/mailman3/'.
* Patching the mailman sources to consider '/etc/mailman3/mailman.cfg'
as config file.
Fixes: #399* Debian (and probably other distributions in the future) installs the
'mailman.cfg' config file to '/etc/mailman3/'.
* Patching the mailman sources to consider '/etc/mailman3/mailman.cfg'
as config file.
Fixes: #3993.2https://staging.gitlab.com/mailman/mailman/-/merge_requests/317Fix argument swap in documentation for qmail2017-09-28T16:40:17Zusername-removed-1609655Fix argument swap in documentation for qmailI tripped over this when I setup mailman with qmail. The text explains it right, but the example code is wrong.I tripped over this when I setup mailman with qmail. The text explains it right, but the example code is wrong.3.2https://staging.gitlab.com/mailman/mailman/-/merge_requests/316Add an admonition about requiring a UTF-8 locale2017-09-07T02:26:55Zusername-removed-56772Add an admonition about requiring a UTF-8 locale3.2username-removed-56772username-removed-56772https://staging.gitlab.com/mailman/mailman/-/merge_requests/315Allow a list's acceptable aliases to be cleared via REST2017-09-04T14:17:19Zusername-removed-56772Allow a list's acceptable aliases to be cleared via REST3.2username-removed-56772username-removed-56772https://staging.gitlab.com/mailman/mailman/-/merge_requests/311Fix runner started/exiting log messages2017-09-02T17:42:30Zusername-removed-56772Fix runner started/exiting log messagesCloses #393
@msapiro I suck for not writing a test.Closes #393
@msapiro I suck for not writing a test.3.2username-removed-56772username-removed-56772https://staging.gitlab.com/mailman/mailman/-/merge_requests/308PR288 vetted2017-08-29T14:07:59Zusername-removed-56772PR288 vettedThis is based on !288 and vets the changes for API, functionality, testing, and code organization.This is based on !288 and vets the changes for API, functionality, testing, and code organization.3.2username-removed-56772username-removed-56772https://staging.gitlab.com/mailman/mailman/-/merge_requests/307Pluggable components vetted2017-08-19T23:03:08Zusername-removed-56772Pluggable components vettedThis is based on mailman/mailman!288 and vets the changes for API, functionality, testing, and code organization.This is based on mailman/mailman!288 and vets the changes for API, functionality, testing, and code organization.3.2username-removed-56772username-removed-56772https://staging.gitlab.com/mailman/mailman/-/merge_requests/306Print the list of available commands with `mailman --help`2017-08-26T18:40:37Zusername-removed-262Print the list of available commands with `mailman --help`Overriding the `format_options` method of `click.MultiCommand` skips printing the
list of available commands. This commit just brings that option back so that
help text is now more informative.Overriding the `format_options` method of `click.MultiCommand` skips printing the
list of available commands. This commit just brings that option back so that
help text is now more informative.3.2username-removed-56772username-removed-56772https://staging.gitlab.com/mailman/mailman/-/merge_requests/305Set a timeout on the lock acquisition2017-07-27T23:10:44Zusername-removed-56772Set a timeout on the lock acquisitionOccasionally, status.rst hangs. This is an attempt to at least avoid such
hangs by timing out on the lock acquisition. It's the only thing in this test
that I can see would cause such a hang, but the problem is so intermittent
that it'...Occasionally, status.rst hangs. This is an attempt to at least avoid such
hangs by timing out on the lock acquisition. It's the only thing in this test
that I can see would cause such a hang, but the problem is so intermittent
that it's very difficult to debug (and never happens in local tests).
If this *is* the problem, it means the master lock was acquired before we got
here, so the test should have failed in first command() call. But we would
never see that because doctests don't have a trace functionality afaict.
It's all I can think of though.3.2username-removed-56772username-removed-56772https://staging.gitlab.com/mailman/mailman/-/merge_requests/304Rename metadata key for clarity2017-08-04T01:13:06Zusername-removed-56772Rename metadata key for clarity@msapiro This is what I was thinking. Note I did not change the dmarc metadata key, but all the tests still pass. Will this break dmarc mitigation?@msapiro This is what I was thinking. Note I did not change the dmarc metadata key, but all the tests still pass. Will this break dmarc mitigation?3.2username-removed-56772username-removed-56772https://staging.gitlab.com/mailman/mailman/-/merge_requests/303Fix for #383: email encoding errors can crash the REST API2017-08-30T02:52:29Zusername-removed-35210Fix for #383: email encoding errors can crash the REST APISee #383.See #383.3.2username-removed-56772username-removed-56772https://staging.gitlab.com/mailman/mailman/-/merge_requests/302Python 3.4 no more!2017-07-14T02:52:24Zusername-removed-56772Python 3.4 no more!Closes #373Closes #3733.2username-removed-56772username-removed-56772https://staging.gitlab.com/mailman/mailman/-/merge_requests/301Ensure that various rules set a reason.2017-08-11T00:31:00Zusername-removed-107958Ensure that various rules set a reason.Fixes #368
Several rules do not set a `moderation_reason` resulting in holds and rejects with reason "N/A". This MR adds `moderation_reasons` and where applicable `moderation_sender` and `moderation_action` to the various rules than ca...Fixes #368
Several rules do not set a `moderation_reason` resulting in holds and rejects with reason "N/A". This MR adds `moderation_reasons` and where applicable `moderation_sender` and `moderation_action` to the various rules than can cause a message to be held, rejected or discarded. It also adds tests for `moderation_reasons` being set.
Some of these attributes are not currently used in all these cases, but I don't think it hurts to add them in case we want to implement increased logging in the future.
Note: I have unchecked `Remove source branch when merge request is accepted.` in case that helps with possible backporting.3.2username-removed-107958username-removed-107958https://staging.gitlab.com/mailman/mailman/-/merge_requests/300Instantiate components only in add_components, not at lower levels.2017-08-30T19:41:59Zusername-removed-1178112Instantiate components only in add_components, not at lower levels.The `scan_module` and `find_components` functions are also useful when they don't instantiate the components they load, and instantiation is left to `add_components`.
This will for example be useful for https://gitlab.com/mailman/mailma...The `scan_module` and `find_components` functions are also useful when they don't instantiate the components they load, and instantiation is left to `add_components`.
This will for example be useful for https://gitlab.com/mailman/mailman/merge_requests/299.3.2https://staging.gitlab.com/mailman/mailman/-/merge_requests/299WIP: Pluggable workflows2017-08-30T19:39:52Zusername-removed-1178112WIP: Pluggable workflowsThis MR will refactor the current Subscription and Unsubscription workflows into pluggable components. Which will allow to easily insert additional steps into the [un]subscription process for both core and plugins.
This also means split...This MR will refactor the current Subscription and Unsubscription workflows into pluggable components. Which will allow to easily insert additional steps into the [un]subscription process for both core and plugins.
This also means splitting the current [Unn]SubscriptionWorkflow classes into multiple blocks that can be chained, somewhat like chains work. So that they can be reused by other workflows.
Let's say a subscription workflow, introduced by a plugin, that checks the address for being present in some company/organization/other database so that only customers/members/... are allowed to subscribe to a list with that policy.3.2https://staging.gitlab.com/mailman/mailman/-/merge_requests/297Added a rule to discard messages with no valid sender address.2017-07-01T02:40:31Zusername-removed-107958Added a rule to discard messages with no valid sender address.Fixes #369
Unless `config.mailman.sender_headers` is very wrong, a message with no valid sender addresses is clearly bogus and should be discarded and logged which this change will do.Fixes #369
Unless `config.mailman.sender_headers` is very wrong, a message with no valid sender addresses is clearly bogus and should be discarded and logged which this change will do.3.2https://staging.gitlab.com/mailman/mailman/-/merge_requests/295Fixed an AttributeError in subject prefixing.2017-07-01T02:35:48Zusername-removed-107958Fixed an AttributeError in subject prefixing.Fixes #359Fixes #3593.2