* fix: Explicitly set `executable_path` for Selenium
So it finds `geckodriver` again.
* Minimize diff
* fix: use existing executable_name
Co-authored-by: Lars Eggert <lars@eggert.org>
---------
Co-authored-by: Robert Sparks <rjsparks@nostrum.com>
* feat: nginx + robots.txt
* feat: minimal /health/ endpoint
* ci: startupProbe for datatracker pod
* ci: probe auth pod; set timeoutSeconds
* feat: add CSP and other headers to nginx
* fix: typo in nginx.conf
* feat: split auth/dt nginx confs
* test: test health endpoint
* ci: auth service on port 80
We'll remove http-old (8080) in the future.
* ci: rename auth container/nginx cfg
* refactor: always use console log handler
* refactor: json for k8s, plain otherwise
* chore: remove syslog from wsgi.py
* chore: remove debug.log()
* chore: drop syslog from settings.py
* refactor: use log.log() in person.utils
* refactor: fetch_meeting_attendance->log.log()
* chore: gunicorn logs as JSON (wip)
* feat: better json log formatting
* refactor: improve log config
* feat: gunicorn access log fields
* fix: remove type hints
The gunicorn logger plays tricks with the
LogRecord args parameter to let it have string
keys instead of being a simple tuple.
The mypy tests rightly flag this. Rather
than fighting the typing, just remove the
hints and leave a comment warning not to
use the gunicorn-specific formatter with
other loggers.
* feat: generate_wg_summary_files_task()
* refactor: wg summaries from filesys for view
* refactor: use new helper for charter views
* refactor: use FileResponse
* refactor: don't use FileResponse
FileResponse generates a StreamingHttpResponse
which brings with it differences I don't fully
understand, so let's stay with HttpResponse
* test: update view tests
* test: test_generate_wg_summary_files_task()
* chore: create PeriodicTask
N.B. that this makes it hourly instead of daily
* feat: send_apikey_usage_emails_task
* chore: update test to use task instead of cmd
* chore: add PeriodicTask
* chore: remove old command + empty management dir
* chore: remove now-empty bin/weekly
* refactor: only consider keys that might have events
---------
Co-authored-by: Robert Sparks <rjsparks@nostrum.com>
* feat: get tool versions without VersionInfo model
* chore: remove update_external_command_info call
* feat: get tool version without VersionInfo
* chore: Remove VersionInfo model
* chore: Migration to remove VersionInfo
* fix: handle errors better; ignore stderr
* fix: type annotation
* fix: log to stdout/stderr in json format
* chore: remove UTILS_LOGGER_LEVELS
This is not used (there _is_ a setting for the
django.security logger in settings_local.py on
production, but it is redundant with the
settings.LOGGING configuration and is not doing
anything).
* chore: revert to debug_console django logging
* chore: log.log to syslog via datatracker logger
* chore: remove unused imports
---------
Co-authored-by: Robert Sparks <rjsparks@nostrum.com>
* feat: use gunicorn
* fix: let gunicorn emit logs to stdout/stderr
* fix: log to stdout/stderr in json format
* fix: run collectstatic for the local copy of the statics
* wip: identify whats needed to obviate ghostlinkd
* fix: hardlink new charter files to ftp directory
* fix: hardlink new charter files to ftp directory (continued)
* chore: bring settings comment up to date
* chore: add archive and ftp dirs to setup of various environments
* fix: test charter submits write to ftp dir
* chore: remove debug
* fix: test charter approval writes to ftp dir
* fix: link review revisions into ftp dir
* fix: link to all archive and ftp on submission post
* chore: clean comments, move action to github issue
* fix: link idindex files to all archive and ftp
* chore: deflake
* chore: remove TODO comment
* fix: use settings
* chore: rename new setting
* feat: Show bluesheets using Attended tables (#7094)
* feat: Show bluesheets using Attended tables (#6898)
* feat: Allow users to add themselves to session attendance (#6454)
* chore: Correct copyright year
* fix: Address review comments
* fix: Don't try to generate empty bluesheets
* refactor: Complete rewrite of bluesheet.html
* refactor: Fill in a few gaps, close a few holes
- Rename the live "bluesheet" to "attendance", add some explanatory text.
- Add attendance links in materials view and pre-finalized proceedings view.
- Don't allow users to add themselves after the corrections cutoff date.
* fix: Report file-save errors to caller
* fix: Address review comments
* fix: typo
* refactor: if instead of except; refactor gently
* refactor: Rearrange logic a little, add comment
* style: Black
* refactor: auto_now_add->default to allow override
* refactor: jsonschema to validate API payload
* feat: Handle new API data format
Not yet tested except that it falls back when the old
format is used.
* test: Split test into deprecated/new version
Have not yet touched the new version
* style: Black
* test: Test new add_session_attendees API
* fix: Fix bug uncovered by test
* refactor: Refactor affiliation lookup a bit
* fix: Order bluesheet by Attended.time
* refactor: Move helpers from views.py to utils.py
* test: Test that finalize calls generate_bluesheets
* test: test_bluesheet_data()
* fix: Clean up merge
* fix: Remove debug statement
* chore: comments
* refactor: Renumber migrations
---------
Co-authored-by: Paul Selkirk <paul@painless-security.com>
* chore: Remove unused import
* style: Black
* feat: Stub session update notify API
* feat: Add order & rev to slides JSON
* style: Black
* feat: Stub actual Meetecho slide deck mgmt API
* refactor: Limit reordering to type="slides"
* chore: Remove repository from meetecho API
(API changed on their end)
* feat: update Meetecho on slide reorder
* refactor: drop pytz from meetecho.py
* chore: Remove more repository refs
* refactor: Eliminate more pytz
* test: Test add_slide_deck api
* fix: Allow 202 status code / absent Content-Type
* test: Test delete_slide_deck api
* test: Test update_slide_decks api
* refactor: sessionpresentation_set -> presentations
* test: Test send_update()
* fix: Debug send_update()
* test: ajax_reorder_slides calls Meetecho API
* test: Test SldesManager.add()
* feat: Implement SlidesManager.add()
* test: Test that ajax_add_slides... calls API
* feat: Call Meetecho API when slides added to session
* test: Test SlidesManager.delete()
* feat: Implement SlidesManager.delete()
* test: ajax_remove_slides... calls Meetecho API
* feat: Call Meetecho API when slides removed
* chore: Update docstring
* feat: rudimentary debug mode for Meetecho API
* test: remove_sessionpresentation() calls Meetecho API
* feat: Call Meetecho API from remove_sessionpresentation()
* test: upload_slides() calls Meetecho API
* style: Black
* fix: Refactor/debug upload_session_slides
Avoids double-save of a SessionPresentation for the session
being updated and updates other sessions when apply_to_all
is set (previously it only created ones that did not exist,
so rev would never be updated).
* test: Fix test bug
* feat: Call Meetecho API when uploading session slides
* fix: Only replace slides actually linked to session
* fix: Delint
Removed some type checking rather than debugging it
* fix: Send get_versionless_href() as url for slides
* test: TZ-aware timestamps, please
* chore: Add comments
* feat: Call Meetecho API in edit_sessionpresentation
* feat: Call Meetecho API in remove_sessionpresentation
* feat: Call Meetecho API from add_sessionpresentation
* fix: Set order in add_sessionpresentation
* fix: Restrict API calls to "slides" docs
* feat: Call Meetecho API on title changes
* test: Check meetecho API calls in test_revise()
* fix: better Meetecho API "order" management
* fix: no PUT if there are no slides after DELETE
* feat: Catch exceptions from SlidesManager
Don't let errors in the MeetEcho slides API interfere with
the ability to modify slides for a session.
* feat: Limit which sessions we send notifications for
* fix: handle absence of request_timeout in api config
* test: always send slide notifications in tests
* fix: save slides before sending notification (#7172)
* fix: save slides before sending notification
* style: fix indentation
It's not a bug, it's a flourish!
---------
Co-authored-by: Jennifer Richards <jennifer@staff.ietf.org>
Co-authored-by: Paul Selkirk <paul@painless-security.com>
* chore: remove unused setting
* feat: initial import of iesg minutes
* fix: let the meetings view show older iesg meetings
* feat: iesg narrative minutes
* feat: import bof coordination call minutes
* wip: import commands for iesg appeals and statements
* feat: import iesg statements.
* feat: import iesg artifacts
* feat: many fewer n+1 queries for the group meetings view
* fix: restore chain of elifs in views_doc
* fix: use self.stdout.write vs print in mgmt commands
* fix: use replace instead of astimezone when appropriate
* chore: refactor new migrations into one
* fix: transcode some old files into utf8
* fix: repair overzealous replace
* chore: black
* fix: address minro review comments
* fix: actually capture transcoding work
* fix: handle multiple iesg statements on the same day
* fix: better titles
* feat: pill badge replaced statements
* fix: consolodate source repos to one
* feat: liberal markdown for secretariat controlled content
* fix: handle (and clean) html narrative minutes
* feat: scrub harder
* fix: simplify and improve a scrubber
* chore: reorder migrations
* feat: DraftAliasGenerator class
Encapsulates logic from generate_draft_aliases.py
* refactor: Avoid circular imports
* feat: Add draft_aliases API endpoint
* feat: Add @requires_api_token decorator
Stolen from feat/rpc-api
* feat: Add token auth to draft_aliases endpoint
* feat: draft-aliases-from-json.py script
Parses output from the draft_aliases API call
* chore: Remove unused cruft
* refactor: Avoid shadowing "draft" name
* fix: Suppress empty lists from DraftAliasGenerator
* refactor: Use a GET instead of POST
* feat: GroupAliasGenerator class
* feat: group aliases API view
* fix: Handle domains array correctly
* fix: Suppress empty group aliases
* refactor: Generalize aliases-from-json.py script
* refactor: Same output fmt for draft and group alias apis
* feat: Sort addresses for stability
* fix: Add "anything" virtual alias
* test: Test requires_api_token decorator
* feat: Harden is_valid_token against misconfig
* test: Test is_valid_token
* test: Test draft_aliases view
* test: Test group_aliases view
* test: Test DraftAliasGenerator
* fix: ise group is type "ise" in test data
* test: Fix logic in testManagementCommand
The test was incorrect - and fails when fixed. :-(
* test: Test GroupAliasGenerator
Test currently fails
* fix: Suppress empty -ads alias
* test: Fix group acronym copy/paste error
I *think* this must be what had been intended. The
code does not look like it ever dealt with GroupHistory,
so I'm pretty sure it wasn't meant to have the same
acronym used by two different Groups at different
times.
* test: Check draft .notify alias generation
* test: Cover get_draft_notify_emails()
* refactor: Change import style for clarity
* feat: Add iana_changes_updates_task()
* chore: Squelch lint warning
My linter does not like variables defined outside
of __init__()
* feat: Add PeriodicTask for iana_changes_updates_task
* refactor: tasks instead of scripts on sync.views.notify()
* test: Test iana_changes_updates_task
* refactor: rename task for consistency
* feat: Add iana_protocols_update_task
* feat: Add PeriodicTask for iana protocols sync
* refactor: Use protocol sync task instead of script in view
* refactor: itertools.batched() not available until py312
* test: test iana_protocols_update_task
* feat: Add idindex_update_task()
Calls idindex generation functions and does the file
update dance to put them in place.
* chore: Add comments to bin/hourly
* fix: annotate types and fix bug
* feat: Create PeriodicTask for idindex_update_task
* refactor: Move helpers into a class
More testable this way
* refactor: Make TempFileManager a context mgr
* test: Test idindex_update_task
* test: Test TempFileManager
* fix: Fix bug in TestFileManager
yay testing
* feat: Add expire_ids_task()
* feat: Create PeriodicTask for expire_ids_task
* test: Test expire_ids_task
* test: Test request timeout in iana_protocols_update_task
* refactor: do not re-raise timeout exception
Not sure this is the right thing to do, but it's the
same as rfc_editor_index_update_task
* feat: Add notify_expirations_task
* feat: Add "weekly" celery beat crontab
* refactor: Reorder crontab fields
This matches the crontab file field order
* feat: Add PeriodicTask for notify_expirations
* test: Test notify_expirations_task
* test: Add annotation to satisfy mypy
* refactor: change references from User to Person (#5821)
* refactor: Change CommunityList reference from User to Person
* refactor: Convert more user references to person
* refactor: Change augment_docs_and_user_with_user_info to person
* refactor: Change Nomination and Feedback references from User to Person
* refactor: Change a few test case function signatures to be more pythonic
* refactor: Harmonize how profile and photo views look up email_or_name
* refactor: Rework community views to operate on Person instead of User (#5859)
* test: Update tests to try all of the person's emails and aliases
* fix: Recode a test case to avoid an exception if there's Unicode in the URL
This only happens using the form-filling and submission feature of
WebTest, which is only used in this one test case, so just it rip out.
* test: Add duplicate-person tests
* fix: If there are multiple matching users, prefer the logged-in one.
* chore: We no longer use WebTest, so don't include it.
* fix: Address review comments
* fix: case-insensitive person name or email matching (#6096)
* chore: Renumber migrations
* fix: Update merged code so tests pass (#6887)
* fix: Use refactored method
* fix: Don't assume user has person
* fix: Use new view param name
* chore: Drop community lists w/o person; cleanup (#6896)
* fix: Don't assume user has person
* fix: user->person in update_community_list_index.py
* feat: Remove CommunityLists without Person
* refactor: Speed up nomcom migrations
---------
Co-authored-by: Paul Selkirk <paul@painless-security.com>
Co-authored-by: Jennifer Richards <jennifer@staff.ietf.org>
* test: Test rfc_editor_index_update_task
* chore: Add docstring to test
* fix: Reuse stats instead of fetching twice
* test: Test fetch_meeting_attendance_task
* chore: Remove outdated tasks
* Revert "chore: Add docstring to test"
This reverts commit 0395410d665c0d310248fd151386f013357c5d13.
* Revert "test: Test rfc_editor_index_update_task"
This reverts commit 4dd9dd131723497db3d2aa76166169fd32c42fdd.
* test: Test rfc_editor_index_update_task
This time without reformatting the entire file...
* chore: Remove accidentally committed fragment
* test: Annotate function to satisfy mypy
* chore: Remove unused imports
* refactor: Factor out helper from fetch_meeting_attendance.py
* feat: Define fetch_meeting_attendance_task task
Equivalent to the fetch_meeting_attendance management command
* chore: Disable fetch_meeting_attendance in bin/daily
* feat: Log errors in fetch_meeting_attendance_task
* feat: Enable a result backend for celery
Ignore results by default, but enable the backend so we
can manage tasks
* feat: Define daily task in ietf.utils.tasks
* refactor: Make bin/send-review-reminders into a task
* refactor: Make bin/send-scheduled-mail into a task
* chore: Update copyright years
* refactor: Make bin/rfc-editor-index-updates into a task
* refactor: Accept date type in rfc index update fn
* chore: Update comment
* fix: Annotate param as Optional
* fix: Revert treating skip_older_than_date as str
Misunderstood the comment, "fixed" a non-bug. Oops.
* feat: mgmt command to create periodic tasks
* feat: add summary of tasks to mgmt cmd
* style: black
* fix: Remove debug statements
* feat: Enable/disable tasks
* chore: Disable periodic tasks by default
* chore: Revert changes to daily and every15m
* fix: Call intended function
* chore: Add task descriptions
* Switch to using geckodriver
* Switch to selenium 4
* Undo
* Remove comment
* Fixes
* Restore non-standard line endings to minimize diff
* Undo
* Remove comment
* test: Fix test_upcoming_view_time_zone_selection
The inputs are hidden and managed by JS, so click
the visible elements instead.
* test: Clumsy fix to test_upcoming_materials_modal
Waiting for the button to be clickable does not
work because the modal is still fading in, so does
not actually close. Would be better to check for it
responding, but I haven't found the right way to do
that yet.
* test: Fix test_add_author_forms
Sending '\n' does not seem to work as it did before,
so click the option instead.
Also reverted some fixme hacks that seem to be obe.
* ci: Update base.Dockerfile
* test: add resource limits to dev/tests/debug.sh env
* ci: add upload of geckodriver.log on failure
* ci: run tests as user 1001
* ci: run app-create-dirs as sudo
* ci: set home folder to root to run tests
---------
Co-authored-by: Jennifer Richards <jennifer@staff.ietf.org>
Co-authored-by: Nicolas Giard <github@ngpixel.com>
Co-authored-by: Robert Sparks <rjsparks@nostrum.com>
* Fix#3911 by adding a markdown filter for template
* fix: move the new markdown filter to utils
* fix: linkify after markdown
---------
Co-authored-by: Robert Sparks <rjsparks@nostrum.com>
Co-authored-by: Jennifer Richards <jennifer@staff.ietf.org>
