spacr.qt.ingest_preview¶

Extraction-preview bridge.

Turns a described dataset — either a container file (nd2 / czi / lif / multi-page tiff / npz) inspected by spacr.qt.multi_format, or a folder-structured layout recognised by spacr.qt.folder_metadata — into a flat list of the individual image “planes” it would expand to, without reading any pixel data.

Each plane is a plain dict row:

{"original": <source path or series>,
 "plate":    "plate1",
 "well":     "plate1_A01",
 "field":    1,
 "channel":  1,
 "time":     1,
 "canonical": "plate1_A01_T0001F001L01C01.tif"}

The canonical names match spacr.io.convert_to_yokogawa() (the pipeline’s own container extractor) so the preview the user edits is the layout the extraction will actually produce. The rows feed the editable metadata table (spacr.qt.widgets.metadata_table) and can be written to a filename_map.csv via rows_to_mappings() + spacr.qt.folder_metadata.save_filename_map().

Functions¶

`mapping_to_row`(→ Dict[str, Any])	Convert a `spacr.qt.folder_metadata.NameMapping` to a row dict.
`plan_container_extraction`(→ List[Dict[str, Any]])	Enumerate the planes a container file would expand into.
`plan_folder_extraction`(→ List[Dict[str, Any]])	Enumerate the planes a folder-structured dataset would map to.
`rows_to_mappings`(→ List[Any])	Convert edited table rows back into `NameMapping` objects ready
`summarize_rows`(→ str)	One-line count summary of a preview (wells / fields / channels).

Module Contents¶

spacr.qt.ingest_preview.mapping_to_row(m: Any) → Dict[str, Any][source]¶

Convert a spacr.qt.folder_metadata.NameMapping to a row dict.

Parameters:: m – mapping object, read by attribute; a missing attribute falls back to "" (paths, well), "plate1" or 1 (field, channel, time). The row’s keys are ROW_COLUMNS.

spacr.qt.ingest_preview.plan_container_extraction(desc: Any, plate: str = 'plate1', well: str = 'A01') → List[Dict[str, Any]][source]¶

Enumerate the planes a container file would expand into.

Mirrors spacr.io.convert_to_yokogawa(): a single container file is assigned one well, and its fields / channels / timepoints become the F / C / T indices of the generated TIFFs. Z-slices are max-projected (MIP) by the converter, so they are not enumerated here.

Parameters:

desc – a DatasetDescription (needs n_fields, n_channels, n_timepoints and path).
plate – plate id for the canonical name.
well – bare well id (A01); combined with plate into the Yokogawa well token plate1_A01.

Returns:

one row dict per (time, field, channel) plane.

spacr.qt.ingest_preview.plan_folder_extraction(root: Any, plate: str = 'plate1', limit: int | None = 200, files: Iterable[pathlib.Path] | None = None, template: Any = _UNSET) → List[Dict[str, Any]][source]¶

Enumerate the planes a folder-structured dataset would map to.

Uses spacr.qt.folder_metadata.detect_folder_metadata() to decide which fields the folder tree already provides, then spacr.qt.folder_metadata.assign_missing_fields() to mint the rest (stable, sorted order). Every image file becomes one row.

Parameters:

root – dropped folder.
plate – plate id used in the canonical names.
limit – cap on the number of rows returned (the table only needs a representative preview). None for no cap.
files – image paths to plan from, instead of walking root. May be a generator — it is consumed here. A caller that has already walked the tree (see spacr.qt.folder_metadata.iter_image_files()) passes it in so the tree is not walked a second time.
template – an already-detected FolderTemplate, or None for “detection ran and found nothing”. Omit to detect here — which walks the tree again, so a caller that already has one should pass it.

Returns:

one row dict per source image, or [] if nothing matched.

spacr.qt.ingest_preview.rows_to_mappings(rows: Sequence[Dict[str, Any]]) → List[Any][source]¶

Convert edited table rows back into NameMapping objects ready for spacr.qt.folder_metadata.save_filename_map().

Parameters:: rows – row dicts keyed by ROW_COLUMNS; missing text keys become "" (plate becomes "plate1") and a missing or empty field, channel or time becomes 1.

spacr.qt.ingest_preview.summarize_rows(rows: Sequence[Dict[str, Any]]) → str[source]¶

One-line count summary of a preview (wells / fields / channels).

Parameters:: rows – preview row dicts; distinct well, field, channel and time values are counted, and timepoints are only mentioned when there is more than one.