OCDocker.Workbench.CampaignProgress module

Structured per-sample progress for vs_campaign jobs, parsed from the job’s own stdout — no Snakemake internals, no extra process. Supports both execution engines (see OCDocker.Workbench.Jobs): the "snakemake" engine’s own console output (job start/finish/error blocks, X of Y steps lines) and the "shell" engine’s [sample i/N] markers.

Text-format parsing is inherently tied to the exact wording a given tool version prints; a future Snakemake major version could change it. When nothing recognizable is found, this degrades to engine: "unknown" with an empty samples mapping — never an exception.

OCDocker.Workbench.CampaignProgress.parse_campaign_progress(log_text, *, expected_samples=None)[source]

Parse structured per-sample progress from a vs_campaign job’s log.

Tries the Snakemake console-output format first, then the shell engine’s [sample i/N] markers, and degrades to engine: "unknown" with an empty samples mapping for anything else — never raises.

Parameters:

  • log_text (str) – A tracked job’s stdout (and, for the error-block detection in the Snakemake format, ideally its stderr too — callers may concatenate both; see OCDocker.Workbench.Jobs.JobManager.logs()).

  • expected_samples (Sequence[str] or None) – Sample names known ahead of parsing (e.g. from the launched manifest); samples not yet mentioned in the log are reported as "pending" instead of being omitted.

Returns:

{"engine": "snakemake" | "shell" | "unknown", "overall": {...} | None, "samples": {name: {"status": ...}}}.

Return type:

dict[str, Any]