Repository navigation
Comparing changes
Open a pull request
base repository: fhgffy/client_python
base: master
head repository: prometheus/client_python
compare: master
- 9 commits
- 15 files changed
- 10 contributors
Commits on Oct 14, 2026
-
docs: fix duplicated word in exemplars guide (prometheus#1212)
Signed-off-by: rohit singh <145866377+Rohit-Singh-01@users.noreply.github.com>
Configuration menu - View commit details
-
Copy full SHA for bd1b31d - Browse repository at this point
Copy the full SHA bd1b31dView commit details -
Fall back to the text format for a non-numeric Accept version (promet…
…heus#1213) choose_encoder compared the version parameter of the Accept header as a tuple of mixed ints and strings, so a version such as 1.0.0-rc.1, a quoted "1.0.0" or latest raised TypeError and the scrape answered 500. Treat such a version as unsupported, which falls through to the next media range and finally to text/plain version 0.0.4. Signed-off-by: Christopher Krah <github@christopher-krah.com>
Configuration menu - View commit details
-
Copy full SHA for 2aac0e0 - Browse repository at this point
Copy the full SHA 2aac0e0View commit details -
Avoid parent lock for existing metric labelsets (prometheus#1216)
* Avoid parent lock for existing metric labelsets Signed-off-by: skulitom <artemskulimovskiy@gmail.com> * Comment the lock-free label lookup Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Signed-off-by: skulitom <artemskulimovskiy@gmail.com> --------- Signed-off-by: skulitom <artemskulimovskiy@gmail.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Configuration menu - View commit details
-
Copy full SHA for dc3fc54 - Browse repository at this point
Copy the full SHA dc3fc54View commit details -
Fix gzip content encoding quality negotiation (prometheus#1220)
Honor q=0 when clients explicitly reject gzip encoding. Add WSGI regression cases for gzip quality parameters. Signed-off-by: mhatipog <metin.hatipogullari@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for acb7b14 - Browse repository at this point
Copy the full SHA acb7b14View commit details -
Make mostrecent gauge tests independent of clock resolution (promethe…
…us#1219) Signed-off-by: nischita44 <nischitasadananda@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 0eef1a2 - Browse repository at this point
Copy the full SHA 0eef1a2View commit details -
openmetrics: escape sample metric names consistently with HELP/TYPE (p…
…rometheus#1204) In the OpenMetrics exposition, the # HELP/# TYPE/# UNIT lines escape the metric name with escape_metric_name() (which treats ':' as a valid metric name character), but the sample line escaped it with the label-name rune set, turning ':' into '_'. A metric such as 'sglang:token_usage' was then emitted as 'sglang:token_usage' in the metadata lines but 'sglang_token_usage' in the sample line, producing output that strict OpenMetrics parsers reject as an orphaned metadata set plus an untyped metric. Use escape_metric_name() for the sample name too, so the same helper drives HELP, TYPE, UNIT and the sample line for every escaping scheme. Fixes prometheus#1177 Signed-off-by: Md. Amdadul Bari Imad <amdadulbari@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for ded318a - Browse repository at this point
Copy the full SHA ded318aView commit details
Commits on Oct 15, 2026
-
Skip mark_process_dead when the multiprocess path is unset (prometheu…
…s#1205) django_q and similar callers invoke this even when PROMETHEUS_MULTIPROC_DIR is missing. Return instead of passing None into os.path.join. Signed-off-by: Hashim1999164 <64767361+Hashim1999164@users.noreply.github.com>
Configuration menu - View commit details
-
Copy full SHA for 6c5cb2e - Browse repository at this point
Copy the full SHA 6c5cb2eView commit details -
Allow make_asgi_app to be added as a Starlette/FastAPI route (prometh…
…eus#1218) * Allow make_asgi_app to be added as a Starlette/FastAPI route make_asgi_app() returned a plain function. Starlette (and therefore FastAPI) treats a plain function passed to add_route() as a request -> response endpoint rather than an ASGI app, so app.add_route("/metrics", make_asgi_app()) failed with a TypeError. The only working option was app.mount("/metrics", ...), which is a path prefix: metrics are served at /metrics/ and requests for /metrics get a 307 redirect. Return a callable ASGI application object instead. Starlette passes non-function callables through to Route as raw ASGI apps, so add_route("/metrics", make_asgi_app()) now serves /metrics directly. The returned object is called with the same (scope, receive, send) signature, so mounting it or using it as a standalone ASGI app keeps working. Update the FastAPI docs to use add_route and explain the mount behaviour. Fixes prometheus#1016 Signed-off-by: Bhaskar Gurram <gurrambhaskar.ai@gmail.com> * docs: note that add_route needs a release newer than 0.26.0 Signed-off-by: Bhaskar Gurram <gurrambhaskar.ai@gmail.com> * tox: install starlette in the py3.9 and pypy3.9 envs Runs the Starlette add_route and mount tests in CI instead of skipping them, alongside the other optional framework dependencies. Signed-off-by: Bhaskar Gurram <gurrambhaskar.ai@gmail.com> --------- Signed-off-by: Bhaskar Gurram <gurrambhaskar.ai@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 512a8d4 - Browse repository at this point
Copy the full SHA 512a8d4View commit details -
Reject duplicate label names before metric registration (prometheus#1217
) * Reject duplicate label names before metric registration Signed-off-by: fhgffy <102001626+fhgffy@users.noreply.github.com> * Improve duplicate label name diagnostics Remove date prefixes and include each duplicated label name in the validation error. Cover multiple duplicates, repeated names and Unicode labels. Signed-off-by: fhgffy <102001626+fhgffy@users.noreply.github.com> --------- Signed-off-by: fhgffy <102001626+fhgffy@users.noreply.github.com>
Configuration menu - View commit details
-
Copy full SHA for ff9a2d3 - Browse repository at this point
Copy the full SHA ff9a2d3View commit details
This comparison is taking too long to generate.
Unfortunately it looks like we can’t render this comparison for you right now. It might be too big, or there might be something weird with your repository.
You can try running this command locally to see the comparison on your machine:
git diff master...master