* ci: add test sharding plumbing to server CI
Add infrastructure for upcoming test sharding without changing behavior:
- Add shard-index and shard-total inputs to server-test-template.yml
(defaults preserve existing single-runner behavior)
- Add timing cache restore step (activates only when shard-total > 1)
- Add merge-postgres-test-results job to server-ci.yml that:
- Merges JUnit XML reports from shard artifacts
- Saves timing data cache for future shard balancing
- Handles both single-artifact and multi-shard scenarios
- Add .gitignore entries for timing cache and shard work files
Co-authored-by: Claude <claude@anthropic.com>
* ci: shard server Postgres tests into 4 parallel runners
Extract sharding logic into standalone, tested scripts and enable
4-shard parallel test execution for server Postgres CI:
Scripts:
- server/scripts/shard-split.js: Node.js bin-packing solver that
assigns test packages to shards using timing data from previous runs.
Two-tier strategy: light packages (<2min) whole, heavy packages
(api4, app) split at individual test level.
- server/scripts/run-shard-tests.sh: Multi-run wrapper that calls
gotestsum directly for each package group with -run regex filters.
- server/scripts/shard-split.test.js: 8 test cases covering round-robin
fallback, timing-based balancing, heavy package splitting, JUnit XML
fallback, and enterprise package separation.
Workflow changes:
- server-test-template.yml: Add shard splitting step that discovers test
packages and runs the solver. Modified Run Tests step to use wrapper
script when sharding is active.
- server-ci.yml: Add 4-shard matrix to test-postgres-normal. Update
merge job artifact patterns for shard-specific names.
Performance: 7.2 min with timing cache vs 62.5 min baseline = 88%
wall-time improvement. First run without cache uses JUnit XML fallback
or round-robin, then populates the cache for subsequent runs.
Co-authored-by: Claude <claude@anthropic.com>
* fix: raise heavy package threshold to 5 min to preserve test isolation
sqlstore integrity tests scan the entire database and fail when other
packages' test data is present. At 182s, sqlstore was just over the
120s threshold and getting split at test level. Raising to 300s keeps
only api4 (~38 min) and app (~15 min) as heavy — where the real
sharding gains are — while sqlstore, elasticsearch, etc. stay whole
and maintain their test isolation guarantees.
Co-authored-by: Claude <claude@anthropic.com>
* ci: only save test timing cache on default branch
PR branches always restore from master's timing cache via restore-keys
prefix matching. Timing data is stable day-to-day so this eliminates
cache misses on first PR runs and reduces cache storage.
Co-authored-by: Claude <claude@anthropic.com>
* ci: skip FIPS tests on PRs (enterprise CI handles compile check)
Per review feedback: the enterprise CI already runs a FIPS compile
check on every PR. Running the full FIPS test suite on PRs is redundant
since it uses the identical test suite as non-FIPS — the only
FIPS-specific failure mode is a build failure from non-approved crypto
imports, which the enterprise compile check catches.
Full FIPS tests continue to run on every push to master.
Co-authored-by: Claude <claude@anthropic.com>
* fix: address review feedback on run-shard-tests.sh
- Remove set -e so all test runs execute even if earlier ones fail;
track failures and exit with error at the end (wiggin77)
- Remove unused top-level COVERAGE_FLAG variable (wiggin77)
- Fix RUN_IDX increment position so report, json, and coverage files
share the same index (wiggin77)
- Update workflow comment: heavy threshold is 5 min, not 2 min (wiggin77)
Co-authored-by: Claude <claude@anthropic.com>
* style: use node: prefix for built-in fs module in shard-split.js
Co-authored-by: Claude <claude@anthropic.com>
* fix: avoid interpolating file paths into generated shell script
Read shard package lists from files at runtime instead of interpolating
them into the generated script via printf. This prevents theoretical
shell metacharacter injection from directory names, as flagged by
DryRun Security.
Co-authored-by: Claude <claude@anthropic.com>
* fix(ci): rename merged artifact to match server-ci-report glob
The merged artifact was named postgres-server-test-logs-merged which
does not match the *-test-logs pattern in server-ci-report.yml,
causing Postgres test results to be missing from PR/commit reports.
Also pins junit-report-merger to exact version 7.0.0 for supply chain
safety.
Co-authored-by: Claude <claude@anthropic.com>
* fix(ci): pass RACE_MODE env into Docker container
RACE_MODE was set on the host runner but never included in the docker
run --env list. The light-package path worked because the heredoc
expanded on the host, but run-shard-tests.sh reads RACE_MODE at
runtime inside the container where it was unset. This caused heavy
packages (api4, app) to silently lose -race detection.
Co-authored-by: Claude <claude@anthropic.com>
* fix(ci): discover new tests in heavy packages not in timing cache
Tests not present in the timing cache (newly added or renamed) would
not appear in any shard -run regex, causing them to silently skip.
After building items from the cache, run go test -list to discover
current test names and assign any cache-missing tests to shards via
the normal bin-packing algorithm with a small default duration.
Co-authored-by: Claude <claude@anthropic.com>
* fix(ci): add missing line continuation backslash in docker run
The previous --env FIPS_ENABLED line was missing a trailing backslash
after adding --env RACE_MODE, causing docker run to see a truncated
command and fail with "requires at least 1 argument".
Co-authored-by: Claude <claude@anthropic.com>
* fix(ci): add setup-go step for shard test discovery
go test -list in shard-split.js runs on the host runner via execSync,
but Go is only available inside the Docker container. Without this
step, every invocation fails silently and new-test discovery is a
no-op. Adding actions/setup-go before the shard split step ensures
the Go toolchain is available on the host.
Co-authored-by: Claude <claude@anthropic.com>
---------
Co-authored-by: Claude <claude@anthropic.com>
* Fix docs-impact-review hitting max turns by adding filesystem tools
The claude-code-action workflow instructs Claude to search ./docs/source/
for existing documentation, but only grants gh and MCP tools. Claude
repeatedly attempts filesystem access, gets denied, and exhausts the
30-turn limit. Add find, grep, cat, and ls to --allowedTools so the
docs checkout can actually be searched.
* Harden docs-impact-review against prompt-injection exfiltration
- Drop find from allowedTools (unsafe due to -exec)
- Scope ls, cat, grep to ./docs/source* paths only
- Add persist-credentials: false to both checkout steps to
prevent token extraction from .git/config
* Use built-in Read/Glob/Grep tools instead of scoped Bash wrappers
Built-in tools cannot execute commands, chain via shell operators,
or abuse flags like -exec, making them inherently safer than any
Bash glob pattern for filesystem access.
* ci: cache prepackaged plugins in mmctl tests
Cache the prepackaged_plugins directory between mmctl test runs using
actions/cache. The cache key is derived from a SHA-256 hash of the
PLUGIN_PACKAGES lines in server/Makefile, so the cache automatically
invalidates whenever any plugin version is bumped — no manual key
updates needed.
Uses actions/cache output (cache-hit) to skip the download step
entirely on cache hits.
Saves ~1-2 minutes per mmctl test run by avoiding repeated plugin
downloads.
Co-authored-by: Claude <claude@anthropic.com>
* fix: align plugin cache key with actual downloaded packages
Address CodeRabbit review: cache key was derived from all 15 plugins
in the Makefile, but the download only fetched jira plugin. Now both
the hash and the download use the same MMCTL_PLUGIN_PACKAGES value,
defined once, preventing stale cache or unnecessary invalidation.
Co-authored-by: Claude <claude@anthropic.com>
* Revert "fix: align plugin cache key with actual downloaded packages"
This reverts commit 73303864bf.
* Revert "ci: cache prepackaged plugins in mmctl tests"
This reverts commit acd1ec757a.
* prepackage the existing test plugin to speed up tests
* fix: use public key for plugin signature verification in test
---------
Co-authored-by: Claude <claude@anthropic.com>
Co-authored-by: Jesse Hallam <jesse@mattermost.com>
* feat: add CI check for test analysis
* update per comments
* address comments
* on silence mode, with general prompt and webhook message for analysis and suggestions
* update test analysis and override only via comment not label
* address reviews
Add path exclusions for markdown files, NOTICE.txt, and CHANGELOG.md
to the pull_request trigger. This prevents the full 60-minute CI suite
from running when a PR only changes documentation.
Exclusions added:
- server/**/*.md
- server/NOTICE.txt
- server/CHANGELOG.md
Co-authored-by: Claude <claude@anthropic.com>
* Add -buildvcs=false to default GOFLAGS
This prevents Go from embedding VCS information into binaries, which
avoids false positives in container vulnerability scanners like Trivy
when using Go workspaces with enterprise dependencies.
Also updates mmctl-build target to use $(GO) and $(GOFLAGS) for
consistency with other build targets.
Made-with: Cursor
* Update comment wording
Trigger PR sync to test Enterprise CI
Made-with: Cursor
* Trigger CI to test Enterprise CI fix
Made-with: Cursor
* Test Enterprise CI
Made-with: Cursor
* replace buildvcs metadata in mmctl
* rm redundant -buildvcs=false in GitHub actions
* update mmctl-docs to $(GO)
* simplify getVersionInfo signature
* use GOOS/GOARCH convention
* export GOFLAGS for common use
* Clarify version.go var block comment
---------
Co-authored-by: Jesse Hallam <jesse@mattermost.com>
* Run docs-impact-review as regular CI instead of slash command
Switch the Documentation Impact Review workflow from a `/docs-review`
slash command trigger to running automatically on pull_request events.
The analysis posts a sticky PR comment and inline comments, and adds a
`docs/needed` label when documentation changes are needed.
Made-with: Cursor
* Address review feedback: harden docs-impact-review workflow
- Remove invalid `direct_prompt` input not recognized by the pinned
claude-code-action version (auto-detects mode from event type).
- Guard Claude step with secret availability check so fork and
Dependabot PRs skip gracefully instead of failing.
- Move label management out of the LLM into a deterministic
github-script step that parses the sticky comment, eliminating
prompt-injection and excessive-agency risks.
Made-with: Cursor
* Fix secret check: use job-level env var instead of secrets context in step if
The secrets context is not available in step-level if expressions.
Evaluate it as a job-level env var and reference that instead.
Made-with: Cursor
* Use paginated API call to fetch all PR comments
Prevents missing the bot comment on busy PRs with more than 30 comments.
Made-with: Cursor
* Address review: add issues:write permission, remove auto-label-removal
- Upgrade issues permission from read to write (required for addLabels API)
- Remove automatic docs/needed label removal to avoid stripping the label
prematurely when the analysis comment hasn't been posted yet
Made-with: Cursor
* Add CI check for broken mattermost.com links in webapp
Add a script that scans the webapp source files for links to mattermost.com
domains and tests each unique URL for 404s. This helps detect broken
documentation and marketing links early.
- New script: webapp/scripts/check-external-links.mjs
- New npm target: check-external-links
- New CI job in webapp-ci.yml to run on every commit
* Add --markdown flag for GitHub Actions job summary
* Fix job summary: use pipefail and suppress progress output
* Require mattermost.com links to use /pl/ permalink format
* Require all mattermost.com links (including subdomains) to use /pl/
* Allow exceptions for push servers and root domain
* Make non-permalink URLs warnings instead of errors
* Add User-Agent header and retry GET on 403
* Follow redirects when checking URLs
Check the final destination of redirects to catch broken links that
redirect to error pages. If a redirect response has the Cloudflare
cf-mitigated header, assume the URL is OK and stop following.
* Simplify link checker code
- Combine PUSH_SERVER_PATTERN and HPNS_PATTERN into single regex
- Simplify validatePermalink to return boolean (reason was unused)
- Consolidate Cloudflare header checks in processResponse
* replace broken links with valid ones
* updates
* test: increase parallel tests and removed smoke tests both in cypress and playwright
* add duration and retest info
* indicate overall, first-pass and re-run run and test durations
* Change moduleResolution to bundler
This makes TS follow the module resolution of newer versions of Node.js which
makes it use the `imports` and `exports` fields of the package.json while not
requiring file extensions in some cases which it does when set to node16 or
nodenext.
I'm changing this to make it so that VS Code can correctly import things from
our types package without adding `/src/` to the import path erroneously.
Hopefully it doesn't introduce any other issues.
* Change make clean to use package.json script
* Remove missing fields from SystemEmoji type
These were removed from emoji.json in
https://github.com/mattermost/mattermost-webapp/pull/9597, but we forgot to
remove the fields from the type definition. They weren't used anyway.
* MM-66867 Add initial version of shared package
This initial version includes the shared context, React Intl support (although
that's currently untested), linting, and testing support. It builds with
Parcel.
* Move isSystemEmoji into Types package
* MM-67318 Add Emoji component to shared package
To limit the number of changes to the web app, it still uses RenderEmoji which
wraps the new component for the time being. I'll likely replace RenderEmoji
with using it directly in a future PR, but I may leave it as-is if the changes
are too big because the API is different.
* Add postinstall script to build shared package
* Revert changes to moduleResolution and add typesVersions to shared package
I plan to still change moduleResolution to bundler since it's the new default
for TS projects, and since it lets TS use the exports field in package.json,
but it requires other changes to fix some minor issues in this repo which I
don't want to muddy this PR with.
Adding typesVersions lets TS resolve the components in the shared package like
it does with the types package while using the old value for moduleResolution.
Plugins still use the old value for moduleResolution, so this will let them use
the shared package with fewer updates changes as well.
* Fix Webpack not always watching other packages for changes
* Add shared package dependencies and build output to CI cache
* Update @parcel/watcher to fix segfaults
This package seems to be older than the rest of the newly added Parcel
dependencies because it's used by sass.
* Fix build script not doing that
* Go back to manually specifying postinstall order
I just learned that postinstall scripts run in parallel because I was running
into an issue where the client and types packages were building at the same
time, causing one of them to fail. They still run in parallel, so that may
still occasionally happen, but by specifying the order manually, we hopefully
avoid that happening like we seemed to do before.
* Further revert changes to postinstall script
The subpackages were also being built when installed
by a plugin
* Increment cache keys
* Fix typo
* Change the cache busting to look at shared/package.json
* Attempt to debug tests and caching
* Debugging...
* Add shared package to platform code coverage
* Remove caching of package builds and manually run postinstall during web app CI setup
* Debugging...
* Remove CI debugging logic
* Update package-lock.json
* Change Emoji component back to taking an emojiName prop
* Add .parcel-cache to .gitignore
* feat: run e2e full tests after successful smoke tests both in cypress and playwright
* fix lint check on jsdoc req in playwright test
* update smoke test filter
* update test filter for cypress tests
* update docker services, fix branch convention and rearrange secrets
* update e2e-test workflow docs
* reorganized
* fix lint
* fix playwright template
* fix results assertion
* add retest, e2e-test-verified, gh comments of failed tests, path filters, run e2e-tests check first and demote unstable tests
* run using master image for e2e-only changes, add ts/js actions for cypress and playwright calculations, add verified by label
---------
Co-authored-by: Mattermost Build <build@mattermost.com>
* chore: upgrade to node 24 and dependencies mainly with babel, webpack and jest
* fix components tests, make trial modal passed on all node 20-24
* fix cache for platform packages
* updated test
---------
Co-authored-by: Mattermost Build <build@mattermost.com>
* feat(webapp): migrate i18n extraction from mmjstool to @formatjs/cli
Replace custom mmjstool with industry-standard @formatjs/cli for i18n
message extraction. Adds support for localizeMessage as a recognized
extraction function and enables comprehensive ESLint rules for i18n.
Changes by area:
Dependencies (package.json):
- Add @formatjs/cli v6.7.4
- Remove @mattermost/mmjstool dependency
- Replace i18n-extract script with formatjs command
- Remove mmjstool-specific scripts (clean-empty, check-empty-src)
- Add i18n-extract:check for diff validation
Code (utils.tsx):
- Refactor localizeMessage to accept MessageDescriptor format
- Support {id, defaultMessage, description} parameters
- Maintain Redux store access (non-React context compatible)
- Add comprehensive JSDoc documentation
Linting (.eslintrc.json):
- Configure formatjs settings with additionalFunctionNames
- Enable 9 formatjs ESLint rules including enforce-id (error)
- Ensure all messages require explicit manual IDs
CI/CD (webapp-ci.yml):
- Simplify i18n check to formatjs extraction + diff
- Remove mmjstool and mobile-dir references
Context: Second attempt at PR #25830. Uses manual IDs only with
[id] placeholder pattern to enforce explicit ID requirements.
ESLint will catch missing IDs during development.
* chore(webapp): add temporary i18n reconciliation tooling
Add helper script to analyze conflicts between en.json and
defaultMessage values in code. This tooling will be reverted
after initial migration cleanup is complete.
Tooling added:
- scripts/i18n-reconcile-conflicts.js: Analyzes conflicts
- Detects spelling corrections
- Identifies placeholder mismatches
- Flags significant content changes
- Provides recommendations on which version to keep
- npm script: i18n-reconcile
Usage:
npm run i18n-reconcile
This will be reverted after reconciling the ~50 duplicate message
warnings and determining correct versions for conflicting messages.
* chore(webapp): upgrade eslint-plugin-formatjs and centralize formatjs deps
Move @formatjs/cli to root webapp package.json alongside other formatjs
dependencies for better monorepo organization. Upgrade eslint-plugin-formatjs
from 4.12.2 to 5.4.2 for latest features and bug fixes.
Changes:
- Upgrade eslint-plugin-formatjs: 4.12.2 → 5.4.2
- Move @formatjs/cli to webapp/package.json (from channels)
- Centralize formatjs tooling at monorepo root level
This provides:
- Better dependency management in monorepo
- Latest formatjs ESLint rules and features
- Consistent tooling across workspaces
* feat(webapp): make localizeMessage fully compatible with formatMessage API
Extend localizeMessage to accept a values parameter for placeholder
interpolation, making it fully compatible with react-intl's formatMessage API.
Remove the now-redundant localizeAndFormatMessage function.
Changes to localizeMessage:
- Add values parameter for {placeholder} interpolation
- Use getIntl() for proper ICU message formatting
- Support full formatMessage feature parity outside React contexts
- Update JSDoc with interpolation examples
- Mark as deprecated (prefer useIntl in React components)
Code cleanup:
- Remove localizeAndFormatMessage function (redundant)
- Migrate all localizeAndFormatMessage usages to localizeMessage
- Remove unused imports: getCurrentLocale, getTranslations
- Add getIntl import
Files migrated:
- notification_actions.tsx: notification.crt message
- app_command_parser_dependencies.ts: intlShim formatMessage
This consolidates i18n logic and provides a single, feature-complete
localization function for non-React contexts.
Example usage:
// Before (old function)
localizeAndFormatMessage(
{id: 'welcome', defaultMessage: 'Hello {name}'},
{name: 'John'}
)
// After (consolidated)
localizeMessage(
{id: 'welcome', defaultMessage: 'Hello {name}'},
{name: 'John'}
)
* refactor(i18n): remove intlShim abstraction, use IntlShape directly
Remove intlShim wrapper and use react-intl's IntlShape type throughout.
Key Changes:
- Remove intlShim constant from app_command_parser_dependencies.ts
- Update errorMessage() signature to accept IntlShape parameter
- Use IntlShape for intl properties in AppCommandParser and ParsedCommand
- Call getIntl() at function start in actions (command.ts, marketplace.ts)
- Pass IntlShape to AppCommandParser and doAppSubmit consistently
Files Modified:
- app_command_parser_dependencies.ts: Remove intlShim, update errorMessage
- app_command_parser.ts: Use IntlShape type
- command.ts, marketplace.ts: Call getIntl() early, pass to parser
- app_provider.tsx, command_provider.tsx: Use getIntl() in constructor
- apps.ts: Update doAppSubmit signature
Benefits: Better type safety, standard react-intl patterns, eliminates
unnecessary abstraction.
Context: Part of migration to @formatjs/cli tooling.
* refactor(webapp): remove deprecated t() function and refactor login validation
Remove the deprecated t() function that was used as a marker for the old
mmjstool extraction. With @formatjs/cli, this is no longer needed as
extraction happens directly from formatMessage calls.
Changes:
- Remove t() function definition from utils/i18n.tsx
- Remove t() import and marker calls from login.tsx
- Refactor login validation logic from multiple if statements to a clean
switch(true) pattern that evaluates boolean combinations directly
- Add helpful comments documenting the structure of login method combinations
- Add default case as safety fallback
The switch(true) pattern eliminates the need to build a key string and makes
the login method combination logic clearer and more maintainable.
Prompt: Remove deprecated t() translation marker function used by old mmjstool.
Refactor login validation from if blocks to switch(true) with boolean cases.
* refactor(i18n): share IntlProvider's intl instance with getIntl()
Make getIntl() return the same IntlShape instance that IntlProvider creates,
ensuring consistency between React components (using useIntl()) and non-React
contexts (using getIntl()).
Changes:
- Add intlInstance storage in utils/i18n.tsx
- Export setIntl() function to store the intl instance
- Modify getIntl() to return stored instance if available, with fallback to
createIntl() for tests and early initialization
- Add IntlCapture component in IntlProvider that uses useIntl() hook to
capture and store the intl instance via setIntl()
Benefits:
- Single source of truth: Both React and non-React code use the same IntlShape
- Consistent translations: No duplicate translation loading or potential desync
- Leverages IntlProvider's translation loading lifecycle automatically
- Maintains fallback for edge cases where IntlProvider hasn't mounted yet
Prompt: Refactor getIntl() to reuse IntlProvider's intl instance instead of
creating a separate instance, ensuring both React and non-React contexts use
the same translations and intl configuration.
* 1
* exclude test files
* reset en.json
* refactor(i18n): inline message ID constants for formatjs extraction
Replace dynamic ID references with literal strings to enable proper formatjs
extraction. Formatjs can only extract from static message IDs, not from
variables or computed IDs.
Changes:
- configuration_bar.tsx: Inline AnnouncementBarMessages constants to literal
string IDs ('announcement_bar.error.past_grace', 'announcement_bar.error.preview_mode')
- system_analytics.test.tsx: Replace variable IDs (totalPlaybooksID,
totalPlaybookRunsID) with literal strings ('total_playbooks', 'total_playbook_runs')
- password.tsx: Use full message descriptor from passwordErrors object instead
of dynamic ID with hardcoded defaultMessage. The errorId is still built
dynamically using template strings, but the actual message is looked up from
the passwordErrors defineMessages block where formatjs can extract it.
This eliminates all "[id]" placeholder entries from extraction output and
ensures all messages have proper extractable definitions.
Prompt: Fix dynamic message ID references that formatjs extraction can't parse.
Inline constants and use message descriptor lookups to enable clean extraction
without [id] placeholders.
* fix(i18n): add missing defaultMessage to drafts.tooltipText
The drafts.tooltipText FormattedMessage had an empty defaultMessage, causing
formatjs extraction to output an empty string for this message.
Added the proper defaultMessage with plural formatting for draft and scheduled
post counts.
Prompt: Fix empty defaultMessage in drafts tooltip causing extraction issues.
* feat(i18n): add --preserve-whitespace flag to formatjs extraction
Add the --preserve-whitespace flag to ensure formatjs extraction maintains
exact whitespace from defaultMessage values. This is important for messages
with intentional formatting like tabs, multiple spaces, or line breaks.
Prompt: Add --preserve-whitespace to i18n extraction to maintain exact formatting.
* feat(i18n): add custom formatjs formatter matching mmjstool ordering
Create custom formatter that replicates mmjstool's sorting behavior:
- Case-insensitive alphabetical sorting
- Underscore (_) sorts before dot (.) to match existing en.json ordering
- Simple key-value format (extracts defaultMessage)
The underscore-before-dot collation matches mmjstool's actual behavior and
reduces diff size by 50% compared to standard alphabetical sorting.
Changes:
- scripts/formatter.js: Custom compareMessages, format, and compile functions
- package.json: Use --format scripts/formatter.js instead of --format simple
Prompt: Create custom formatjs formatter matching mmjstool's exact sorting
behavior including underscore-before-dot collation to minimize migration diff.
* feat(i18n): enhance reconciliation script with git history metadata
Add git log lookups to show when each message was last modified in both
en.json and code files. This enables data-driven decisions about which
version to keep when resolving conflicts.
Enhancements:
- getEnJsonLastModified(): Find when an en.json entry last changed
- findCodeFile(): Locate source file for a message ID (handles id= and id:)
- getCodeLastModified(): Get last modification date for source files
- Caching for performance (minimizes git operations)
Output includes:
📅 en.json: [date] | [hash] | [commit message]
📅 code: [date] | [hash] | [commit message]
📂 file: [source file path]
Analysis shows en.json entries from 2023 while code actively updated
2024-2025, indicating code is more authoritative.
Prompt: Add git history to reconciliation report showing modification dates
for data-driven conflict resolution decisions.
* feat(i18n): add optional duplicate message ID check script
Add i18n-extract:check-duplicates script that fails if the same message ID
has different defaultMessages in multiple locations. This is kept as an
optional standalone script rather than auto-chained with i18n-extract to
allow the migration to proceed while duplicates are addressed separately.
Usage:
npm run i18n-extract:check-duplicates
The script runs formatjs extract to /dev/null and checks stderr for
"Duplicate message id" warnings, failing if any are found.
Prompt: Add optional duplicate ID check as standalone npm script for gradual
migration without blocking on existing duplicate message IDs.
* about.enterpriseEditionLearn
* admin.accesscontrol.title
* eslint-plugin-formatjs version to eslint 8 compat
* admin.channel_settings.channel_detail.access_control_policy_title
* admin.connectionSecurityTls
- now: admin.connectionSecurityTls.title
* admin.customProfileAttribDesc
now:
- admin.customProfileAttribDesc.ldap
- admin.customProfileAttribDesc.saml
* admin.gitlab.clientSecretExample
* package-lock
* admin.google.EnableMarkdownDesc
- fix err `]` in admin.google.EnableMarkdownDesc.openid
- now:
- admin.google.EnableMarkdownDesc.oauth
- admin.google.EnableMarkdownDesc.openid
* admin.image.amazonS3BucketExample
- now:
- admin.image.amazonS3BucketExample
- admin.image.amazonS3BucketExampleExport
* admin.license.trialCard.contactSales
* admin.log.AdvancedLoggingJSONDescription
- now:
- admin.log.AdvancedLoggingJSONDescription
- admin.log.AdvancedAuditLoggingJSONDescription
* admin.rate.noteDescription
- now
- admin.rate.noteDescription
- admin.info_banner.restart_required.desc (3)
* admin.system_properties.user_properties.title
- now:
- admin.system_properties.user_properties.title
- admin.accesscontrol.user_properties.link.label
* admin.system_users.filters.team.allTeams, admin.system_users.filters.team.noTeams
* admin.user_item.email_title
* announcement_bar.error.preview_mode
* channel_settings.error_purpose_length
* deactivate_member_modal.desc
- now:
- deactivate_member_modal.desc
- deactivate_member_modal.desc_with_confirmation
* deleteChannelModal.canViewArchivedChannelsWarning
* edit_channel_header_modal.error
* filtered_channels_list.search
- now:
- filtered_channels_list.search
- filtered_channels_list.search.label
* flag_post.flag
* generic_icons.collapse
* installed_outgoing_oauth_connections.header
* intro_messages.group_message
* intro_messages.notificationPreferences
- now:
- intro_messages.notificationPreferences
- intro_messages.notificationPreferences.label
* katex.error
* multiselect.placeholder
- now:
- multiselect.placeholder
- multiselect.placeholder.addMembers
* post_info.pin, post_info.unpin, rhs_root.mobile.flag
take en.json
* texteditor.rewrite.rewriting
* user_groups_modal.addPeople
split to user_groups_modal.addPeople.field_title
* user.settings.general.validImage
take en.json
* userSettings.adminMode.modal_header
take en.json
* webapp.mattermost.feature.start_call
split to user_profile.call.start
* admin.general.localization.enableExperimentalLocalesDescription
take en.json
* login.contact_admin.title, login.contact_admin.detail
take en.json
* fix: correct spelling of "complementary" in accessibility sections
* Revert "about.enterpriseEditionLearn"
This reverts commit 23ababe1ce.
* about.enterpriseEditionLearn, about.planNameLearn
* fix: improve clarity and consistency in activity log and command descriptions
* fix: update emoji upload limits and improve help text for emoji creation
* fix easy round 1
* fix: improve messaging for free trial notifications and sales contact
take en.json
* admin.billing.subscription.planDetails.features.limitedFileStorage
take en.json
* admin.billing.subscription.updatePaymentInfo
take en.json
* fix easy round 2
* admin.complianceExport.exportFormat.globalrelay
take en.json
* fix round 3
* fix round 4
* fix round 5
* fix round 6
* fix closing tag, newlines, and popout title
* fix linting, +
* update snapshots
* test(webapp): update tests to match formatjs migration message changes
Update test assertions to reflect message text changes from the
mmjstool → @formatjs/cli migration. All changes align with the
corresponding i18n message updates in the codebase.
Changes:
- limits.test.tsx: Update capitalization (Message History → Message history,
File Storage → File storage)
- emoji_picker.test.tsx: Update label (Recent → Recently Used)
- command.test.js: Add period to error message
- channel_activity_warning_modal.test.tsx: Update warning text assertion
- channel_settings_archive_tab.test.tsx: Update archive warning text
- feature_discovery.test.tsx: Update agreement name (Software and Services
License Agreement → Software Evaluation Agreement)
- flag_post_modal.test.tsx: Update placeholder text (Select a reason for
flagging → Select a reason)
- channel_intro_message.test.tsx: Remove trailing space from assertion
- elasticsearch_settings.tsx: Fix searchableStrings placeholder names
(documentationLink → link) to match updated message format
All 983 test suites passing (12 tests fixed, 9088 tests total).
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* chore(i18n): improve cross-platform compatibility and cleanup temporary tooling
Simplify i18n check scripts to reuse npm run i18n-extract, improving
cross-platform compatibility and reducing duplication.
Changes:
- Refactor i18n-extract:check to reuse base extraction command
- Uses project-local temp file instead of /tmp (Windows compatible)
- Simplified from duplicated formatjs command to npm run reuse
- Add --throws flag to fail fast on extraction errors
- Add helpful error message: "To update: npm run i18n-extract"
- Add i18n-extract:check to main check script
- Now runs alongside ESLint and Stylelint checks
- Ensures en.json stays in sync with code changes
- Remove i18n-extract:check-duplicates script
- Redundant - duplicates are caught by i18n-extract automatically
- Avoided grep dependency (not cross-platform)
- Remove temporary reconciliation tooling
- Deleted scripts/i18n-reconcile-conflicts.js (served its purpose)
- Removed i18n-reconcile npm script
- Add .i18n-check.tmp.json to .gitignore
i18n-extract:check now works on Windows, macOS, and Linux.
Requires only diff command (available via Git on Windows).
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* ci(webapp): simplify i18n check using npm script
Replace custom i18n check in CI with npm run i18n-extract:check for
consistency with local development workflow.
Changes:
- Update check-i18n job to use npm run i18n-extract:check
- Remove manual cp/extract/diff steps (3 lines → 1 line)
- Consistent behavior between CI and local development
- Cross-platform compatible (no /tmp dependency)
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* feat(i18n): add i18n extraction targets to Makefile and workspace scripts
Add Make targets and npm workspace scripts for i18n message extraction
to provide consistent tooling access across the project.
Changes:
- Add make i18n-extract target to webapp/Makefile
- Extracts i18n messages from code to en.json
- Add make i18n-extract-check target to webapp/Makefile
- Checks if en.json is in sync with code
- Add i18n-extract workspace script to webapp/package.json
- Runs extraction across all workspaces
- Add i18n-extract:check workspace script to webapp/package.json
- Runs sync check across all workspaces
Usage:
make i18n-extract # Extract messages
make i18n-extract-check # Check if en.json needs update
npm run i18n-extract # From webapp root
npm run i18n-extract:check
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* feat(i18n): add cross-platform cleanup utility for temp files
Add Node.js cleanup script to handle temporary file removal in a
cross-platform manner, replacing platform-specific rm commands.
Changes:
- Add webapp/scripts/cleanup.js
- Cross-platform file deletion utility
- Silently handles missing files
- Works on Windows/Mac/Linux
- Update i18n-extract:check to use cleanup script
- Removes .i18n-check.tmp.json after diff
- Cleans up on both success and failure paths
Usage:
node scripts/cleanup.js <file1> [file2] [...]
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* chore(i18n): remove 'defineMessage' from additionalFunctionNames in ESLint config
* chore(i18n): update i18n-extract check script and remove unused cleanup utility
context: WSL is already required on windows
* chore(i18n): remove ESLint rule disabling for formatjs placeholders in admin definition
* chore(i18n): admin_definition
- re-add defineMessage to eslint additionalfunctionnames
- mark additional individual placeholder lines in admin_definition
* fix(i18n): correct typo in enterprise edition link message
* chore(i18n): fix escape sequences
* revert emoji.ts changes
* fix snapshots
* fix admin.google.EnableMarkdownDesc
- keep en.json but fix linkAPI and <strong>Create</strong>
* restore newlines, and final corrections check
* update snapshots
* fix: i18n
* chore(e2e): admin.complianceExport.exportJobStartTime.title
* chore(e2e): emoji_picker.travel-places
* chore(e2e): user.settings.general.incorrectPassword
* chore(e2e): signup_user_completed.create
* chore(e2e): fix Invite People
* chore(fix): admin console email heading
* chore(e2e): fix edit_channel_header_modal.placeholder
* chore(e2e): fix Create account
* chore(e2e): fix Create account
* chore(e2e): correct success message text in password update alert
* fix: admin guide label
* chore(e2e): update role name in invite people combobox
* chore(e2e): more "Create account"
* chore(e2e): fix authentication_spec
* lint
---------
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Mattermost Build <build@mattermost.com>
* add shard and log heap usage in jest
* separate each type of tests
* fix code coverage and indicate heap usage to all tests
* exercise each 10 times, change channels shard to 4
* revert to single run, fix npm cache and merge/report code coverage with nyc
* s/Server CI (PR|Master)/Server CI/
* whitespace changes
* fixup! s/Server CI (PR|Master)/Server CI/
* Check workflow event type instead of name for PR reporting
Address code review feedback by checking github.event.workflow_run.event
instead of github.event.workflow_run.name since Server CI now runs for
both PRs and master branches.
---------
Co-authored-by: Mattermost Build <build@mattermost.com>
* add GH action to save mmctl e2e test report to zephyr
* test on pr
* bundle dependencies and set conditonal run on local and GH
* ensure test keys are saved
* improve github summary
* add test, organize types
* update dependencies
* only run on master and release branch
* add option to run fips build
* modify gh comments
* organize gh and commit status
* sanitize inputs, handle group concurrency
* indicate fix in status context and fix comparison
* handle fips comparison
* updated per comments
* reverted missed lines
* fix failed cypress test by migrating to playwright
---------
Co-authored-by: Mattermost Build <build@mattermost.com>
* only build for mattermostdevelopment/mattermost-build-server-fips
* calculate go version from Docker image, not .go-version
* allow manual dispach of build-server-image
* Update mmctl-test-template.yml to conditionally execute docker login for mattermost repository owner
* Update build-server-image.yml and server-test-template.yml to conditionally execute docker login for mattermost repository owner
* Update workflow files to conditionally execute docker login for pull requests from mattermost/mattermost
* Enhance CI workflows to conditionally execute jobs for pull requests from mattermost/mattermost
* Remove conditional check for docker login in CI workflows to streamline execution across all pull requests
* Add conditional check for docker login in CI workflows for pull requests from mattermost/mattermost
* Refactor CI workflows to gracefully skip tests for external pull requests, enhancing clarity and execution flow
* revert checks
* skip fips testing for forks
* Fix fork detection in FIPS tests to allow push events
The previous condition `github.repository_owner == 'mattermost'` was
incorrect as it would always be false for forks. Updated to properly
detect forks using `github.event.pull_request.head.repo.full_name == github.repository`
and allow FIPS tests to run on push events.
---------
Co-authored-by: Jesse Hallam <jesse@mattermost.com>
* pin to ubuntu-24.04
* always use FIPS compatible Postgres settings
* use sha256 for remote cluster IDs
* use sha256 for client config hash
* rework S3 backend to be FIPS compatible
* skip setup-node during build, since already in container
* support FIPS builds
* Dockerfile for FIPS image, using glibc-openssl-fips
* workaround entrypoint inconsistencies
* authenticate to DockerHub
* fix FIPS_ENABLED, add test-mmctl-fips
* decouple check-mattermost-vet from test/build steps
* fixup! decouple check-mattermost-vet from test/build steps
* only build-linux-amd64 for fips
* rm entrypoint workaround
* tweak comment grammar
* rm unused Dockerfile.fips (for now)
* ignore gpg import errors, since would fail later anyway
* for fips, only make package-linux-amd64
* set FIPS_ENABLED for build step
* Add a FIPS-specific list of prepackaged plugins
Note that the names are still temporary, since they are not uploaded to
S3 yet. We may need to tweak them when that happens.
* s/golangci-lint/check-style/
This ensures we run all the `check-style` checks: previously,
`modernize` was missing.
* pin go-vet to @v2, remove annoying comment
* add -fips to linux-amd64.tz.gz package
* rm unused setup-chainctl
* use BUILD_TYPE_NAME instead
* mv fips build to enterprise-only
* fixup! use BUILD_TYPE_NAME instead
* temporarily pre-package no plugins for FIPS
* split package-cleanup
* undo package-cleanup, just skip ARM, also test
* skip arm for FIPS in second target too
* fmt Makefile
* Revert "rm unused Dockerfile.fips (for now)"
This reverts commit 601e37e0ff.
* reintroduce Dockerfile.fips and align with existing Dockerfile
* s/IMAGE/BUILD_IMAGE/
* bump the glibc-openssl-fips version
* rm redundant comment
* fix FIPS checks
* set PLUGIN_PACKAGES empty until prepackaged plugins ready
* upgrade glibc-openssl-fips, use non-dev version for final stage
* another BUILD_IMAGE case
* Prepackage the FIPS versions of plugins
* relocate FIPS_ENABLED initialization before use
* s/Config File MD5/Config File Hash/
* Update the FIPS plugin names and encode the + sign
* add /var/tmp for local socket manipulation
---------
Co-authored-by: Alejandro García Montoro <alejandro.garciamontoro@gmail.com>
Co-authored-by: Mattermost Build <build@mattermost.com>
* consolidate go version computation
* use Go from mattermost-build-server
* work around "error obtaining VCS status: exit status 128" in a container"
* no longer require sudo
* fix config-reset for GOFLAGS
* only test-postgres-binary on master/release
* add check-prereqs-enterprise to test-mmctl* to ensure go.work setup
Replace empty string fallback with github.run_id to prevent workflow
evaluation errors when concurrency group evaluates to empty string on
non-pull-request events.
* Consolidate server and webapp CI workflows into single templates
- Merge server-ci-master.yml and server-ci-pr.yml into server-ci-template.yml
- Merge webapp-ci-master.yml and webapp-ci-pr.yml into webapp-ci-template.yml
- Add conditional concurrency control for PRs only
- Maintain all existing behavior while eliminating duplication
* narrow paths to trigger (server|webapp)-ci
* Consolidate server and webapp CI workflows into single templates
- Merge server-ci-master.yml and server-ci-pr.yml into server-ci-template.yml
- Merge webapp-ci-master.yml and webapp-ci-pr.yml into webapp-ci-template.yml
- Add conditional concurrency control for PRs only
- Maintain all existing behavior while eliminating duplication
* Fix concurrency control to preserve master build behavior
Address feedback about master builds being queued. Make concurrency
group conditional so master builds have no concurrency control at all,
maintaining identical behavior to before consolidation.
- Master builds: No concurrency (multiple builds can run simultaneously)
- PR builds: Concurrency with cancellation (one build per PR)
Setting `id-token: write` allows the workflow to request and use a JSON Web Token (JWT) from GitHub's OIDC provider. This JWT can then be exchanged with an external service -- in this case, ChainGuard -- to obtain short-lived credentials for authenticating and performing actions on that service.
We need to identify ourselves to Chainguard for pulling images during
build steps. Note that the identity value is not a secret, at least in
so far as we're already using it this way over at
https://github.com/mattermost/mattermost/pull/33549/files.
Amy BlaisandMattermost Build
Pavel ZemanandClaude
Daniel Schalla
sabril
Eva SarafianouandJesse Hallam
Jesse Hallam
dependabot[bot]anddependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Harrison Healey
Carlos Garcia
Pablo Vélez
sabril
Ibrahim Serdar AcikgozandMattermost Build
Lorenzo
Angelos KyratzakosandJesse Hallam
