Skip to content

Sync with template (14b63e0) - #291

Merged
t0mdavid-m merged 25 commits into
developfrom
claude/project-thread-l702ez
Nov 13, 2026
Merged

t0mdavid-m merged 25 commits into
developfrom
claude/project-thread-l702ez

Conversation

@t0mdavid-m

Copy link
Copy Markdown
Member

Requested by Tom · project thread

Merges OpenMS/streamlit-template 830628c..14b63e0 (24 commits) into develop. Merge with a merge commit, not a squash.

What users get

Conflicts and how they were resolved

  • .github/workflows/build-and-test.yml, content/quickstart.py: app-owned, FLASHApp's kept.
  • .github/workflows/build-desktop-app.yaml: the template's comment, FLASHApp's TOPP_TOOLS.
  • content/raw_data_viewer.py, docs/deployment.md, src/view.py: stay deleted in FLASHApp.

Generated by Claude Code

claude and others added 25 commits November 10, 2026 06:43
… left

A run over 200 files listed every file on the Run page. The list now shows
up to 12: the failed and running ones, the next waiting and the latest
finished, and counts the rest in one line. The progress bar estimates the
time left from the files already finished. After a run, the first five
failed files keep their own error and the others are listed in one
expander; a long list of samples has the full list in an expander.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
It jumped too much between files to be useful for now.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
A run with several failed files listed each with its own error box. Now each
tool that failed gets one line with the count, a table of the files and
their errors, and one Tool output section with a file picker.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
The per-sample lines with their marks are for following a run. Afterwards
the table of failed files and the totals say what happened.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
…m (#577)

* Read Thermo .raw files with OpenMS's built-in reader on every platform

FileConverter defaults to ThermoRawFileParser on Windows and to the
built-in reader on macOS and Linux, and the Docker image had only
ThermoRawFileParser on mono. The two write slightly different mzML
(extra low peaks, a second precursor entry on supplemental-activation
scans), so the same .raw file gave different results per platform.

- convert_to_mzml() passes -RawToMzML:reader inprocess whenever the
  installed FileConverter offers it (thermo_reader_args()).
- The full Docker images build OpenMS WITH_THERMO_RAW=ON with the .NET
  SDK OpenMS pins in tools/ci/install_dotnet_sdk_linux.sh, and check
  that the reader and its managed assemblies are there.
- The Windows desktop app ships the .NET runtime too; ThermoRawFileParser
  is staged only for an OpenMS without the built-in reader.
- The desktop smoke test and the Apptainer test convert a real .raw file
  with the built-in reader.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGen6esZDGqZn6bzrBBG

* stage-dotnet.sh: strip the CR of jq's output on Windows

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGen6esZDGqZn6bzrBBG

* Tests: conversion race tests stub the Thermo reader lookup

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGen6esZDGqZn6bzrBBG

* Record the Thermo reader with a converted mzML; fall back to the default reader

- A converted .raw file's mzML is reused only when it was made by the
  reader convert_to_mzml now uses (.<name>.mzML.reader next to it), so
  mzML of an earlier version's reader is converted again, not mixed in.
- When the built-in reader fails (no .NET 8 runtime, e.g. a source
  install on Windows), the conversion is tried again with FileConverter's
  default reader, with a warning in the log.
- The desktop smoke test checks that the FileConverter .ini and the
  managed assemblies agree on the built-in reader.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGen6esZDGqZn6bzrBBG

* Retry a failed .raw conversion with the tool's own default reader

The retry names FileConverter's default reader, and a value in
custom_params now replaces a stored one on the command line, so a reader
saved in the workflow's parameters cannot take its place. A file converted
after such a fallback is kept on later runs instead of being converted
again while .NET is missing.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGen6esZDGqZn6bzrBBG

---------

Co-authored-by: Claude <noreply@anthropic.com>
A failed run groups its failed files per tool, also for steps run with
allow_failed; a finished run still names the files left out.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGepShSG5cNRTqoCXFo1
Both are applied once per cluster by its admin and are no longer part of
the template's deployment, so no app's Argo CD project needs a
cluster-scoped kind. CI keeps checking its storage stand-in against the
production class's name, reclaim policy and mount options.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemTJKJs8ZhnhDz43cQ
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemTJKJs8ZhnhDz43cQ
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemTJKJs8ZhnhDz43cQ
Run page: list a long run's failed and running files first
* Draw the first experiment's selector above the view bar

experiment_selector() draws an experiment's dataset selector with its
buttons where the page calls it, and render_experiments(first_selector=False)
then shows the dataset picked there without drawing the selector again.
A Viewer can so read top to bottom: what is shown and what can be done
with it, how it is shown, the components.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemhZJqShow7AQz5c3q

* Screenshots for the pull request

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemhZJqShow7AQz5c3q

* Remove pull request screenshots

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemhZJqShow7AQz5c3q

* View MS data: one panel style for every plot, full-width peak map

Plotly charts get the same rounded panel as Insight components and
transparent figure backgrounds, so no grey box sits inside a white page.
The peak map takes the whole row unless the 3D plot is shown.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemhZJqShow7AQz5c3q

* Screenshots for PR description

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemhZJqShow7AQz5c3q

* Remove PR screenshots

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemhZJqShow7AQz5c3q

* Docs: pass label and buttons to experiment_selector by keyword

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemhZJqShow7AQz5c3q

* View bar: a changed ready-made view reaches the stored layout

The layout stored for the current view is written again when its rows no
longer match the view, so an update that changes a ready-made view shows
on the pages that already display it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemhZJqShow7AQz5c3q

---------

Co-authored-by: Claude <noreply@anthropic.com>
…581)

The installer's license page now names the same copyright holder as LICENSE.


Claude-Session: https://claude.ai/code/session_01FkEGenTCt1fet8pkp5v8T2

Co-authored-by: Claude <noreply@anthropic.com>
…ter-config

Leave the openms namespace and the storage tier to the cluster
Conflicts resolved by ownership:
- .github/workflows/build-and-test.yml, content/quickstart.py: app-owned, kept FLASHApp's
- .github/workflows/build-desktop-app.yaml: template comment, FLASHApp's TOPP_TOOLS
- content/raw_data_viewer.py, docs/deployment.md, src/view.py: stay deleted in FLASHApp

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkEGemrJRTcypYoksP1PDd
@coderabbitai

coderabbitai Bot commented Nov 13, 2026

Copy link
Copy Markdown
Contributor

Important

  • Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

Run configuration

  • Source mode: Auto
  • Repository: Unavailable
  • Branch: Unavailable
  • File: No configuration file from this repository applied
  • Commit: Unavailable
  • Layers: Organization UI
  • Review profile: CHILL
  • Plan: Advanced
  • Run ID: b54c71cf-5e29-4397-aed4-4770b176099c
  • Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

Share

Copy link
Copy Markdown
Member Author

The Linux desktop e2e failure (ValueError: mode must be 'lines', 'markers' or 'lines+markers', got 'bars' from src/render/go_view.py → SeriesPlot) comes from develop, not from this sync. #282 draws the GO and comparison plots with SeriesPlot(mode='bars'), which needs t0mdavid-m/OpenMS-Insight#101 (merged). That change is not in a release yet: PyPI's newest version is 0.7.5, which requirements.txt pins and which only accepts lines/markers. This sync doesn't touch those files or the openms-insight pin. To fix it, release openms-insight with #101 and #104, then bump the pin.


Generated by Claude Code

@t0mdavid-m
t0mdavid-m merged commit d85d5d4 into develop Nov 13, 2026
20 of 23 checks passed
@t0mdavid-m
t0mdavid-m deleted the claude/project-thread-l702ez branch November 13, 2026 23:30
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants