autowisp.error_render module

Class Inheritance Diagram

Human-readable projections of a persisted Error record.

The CLI and BUI never format an exception directly – they render the persisted Error row (and, on demand, its detail sidecar) through the functions here, so the two front-ends cannot drift. Rendering is a projection of the stored record, never of a live exception.

autowisp.error_render._describe_artifact(error_row, db_session)

Return a short description of the artifact the error is about.

Resolves the row’s artifact FK to a path; falls back to the bare id if the artifact row is gone, and to None when no artifact is linked.

Parameters:

• error_row (Error) – The error row.

• db_session – Active database session for resolving the FK.

Returns:

e.g. "image '/data/raw/x.fits'".

Return type:

str or None

autowisp.error_render._run_provenance(error_row, db_session)

Return {host, process_id, code_version} for the error’s run.

Empty dict when the error has no pipeline run (standalone CLI/BUI) or the run row is gone.

autowisp.error_render.error_count(db_session=None)

Return the number of open (unresolved) errors.

This is what the error badge shows – resolved errors are kept as history but no longer counted.

Parameters:

db_session – Optional active session; one is opened if omitted.

Returns:

The open-error count (0 if none).

Return type:

int

autowisp.error_render.error_counts_by_step(db_session=None)

Return {step_name: count} for open errors that name a step.

Powers the per-step markers on the progress grid; resolved errors and errors with no step (pipeline/BUI) are excluded.

Parameters:

db_session – Optional active session; one is opened if omitted.

Returns:

Mapping of step name to its open-error count.

Return type:

dict

autowisp.error_render.error_detail(error_row, db_session=None, *, developer=False)

Return the full human view of an error row as a dict.

Lazily loads the sidecar. With developer=False the result holds the user-facing fields (summary, message, artifact, and remediation if the error provided one). With developer=True it adds the technical fields: exception class, full message, traceback, details, subprocess_id, the run’s host/PID/code_version, and the related-file list. A missing sidecar degrades gracefully (the sidecar-backed fields are simply absent / empty).

Parameters:

• error_row (Error) – The error row to render.

• db_session – Optional active session; one is opened if omitted.

• developer (bool) – Include the technical fields.

Returns:

The structured detail view.

Return type:

dict

autowisp.error_render.error_list_rows(db_session=None, *, pipeline_run_id=None, step_name=None)

Return the rows for a list view, newest first, from inline columns.

Reads only the queryable columns – never a sidecar – so a list view stays cheap regardless of how many errors there are.

Parameters:

• db_session – Optional active session; one is opened if omitted.

• pipeline_run_id (int or None) – If given, restrict to that run.

• step_name (str or None) – If given, restrict to that step.

Returns:

One dict per error with id, created,

component, step_name, artifact, user_message, and a one-line summary, ordered newest first.

Return type:

list[dict]

autowisp.error_render.error_summary(error_row, db_session=None)

Return a one-line human summary of an error row.

Format: [component:step] <artifact>: <user_message> (the artifact clause is omitted when none is linked).

Parameters:

• error_row (Error) – The error row to summarize.

• db_session – Optional active session; one is opened if omitted (only needed to resolve the artifact FK).

Returns:

The one-line summary.

Return type:

str

autowisp.error_render.format_detail_text(detail)

Render an error_detail() dict as plain text for a terminal.

Keeps all formatting in this module (the front-ends never format fields themselves). Only the keys present in detail are shown, so the same function serves both the user and developer views.

Parameters:

detail (dict) – The result of error_detail().

Returns:

A multi-line, human-readable rendering.

Return type:

str

autowisp.error_render.open_error_count_for_steps(step_names, db_session=None)

Return how many open errors would gate launching step_names.

Counts open errors that bear on running the pipeline:

• every open pipeline error (an orchestration/config failure is run-level, so it gates any launch until resolved), and

• open step errors for the steps about to run (all steps for a full run, i.e. an empty step_names).

Open BUI errors are excluded – they are web-interface issues, not a reason to hold back processing. Used by the start-processing gate.

Parameters:

• step_names (iterable) – Step names about to be run; empty means a full run (every step).

• db_session – Optional active session; one is opened if omitted.

Returns:

The number of open errors relevant to the launch.

Return type:

int