schema-crosswalk
See docs/reference/README.md for the MUST/SHOULD/MAY convention, and
docs/reference/shared.md for rules schema-crosswalk shares with other tools.
Matching and renaming¶
schema-crosswalkMUST map a source-column -> target-schema crosswalk exactly as standaloneschema-mapdoes (seedocs/reference/schema_map.md): the same matching passes, noise-column exclusion, and output ordering rules apply unchanged,schema-crosswalkcallsschema-map's own matching stage directly rather than owning separate logic.schema-crosswalkMUST then apply that freshly-generated crosswalk exactly as standaloneschema-refactordoes (seedocs/reference/schema_refactor.md): the same coverage-validation, renaming, and dropping rules apply unchanged,schema-crosswalkcallsschema-refactor's own validation and rename stages directly.
Outputs¶
schema-crosswalkMUST always produce both the crosswalk CSV (same shape as standaloneschema-map's output) and the mapped output file (same shape as standaloneschema-refactor's output).schema-crosswalkperforms no topology hard gate at all; neither underlying stage touches geometry.
Configuration (api.schema_crosswalk.crosswalk() / CLI)¶
schema-crosswalkMUST process exactly one input file per call.- The mapped-output path MUST default to the input path with a
_mappedsuffix. The crosswalk path MUST default to the input path with a_crosswalkstem suffix and a.csvextension. schema-crosswalkMUST raiseFileExistsErrorif either output path already exists and overwriting wasn't requested.levelMUST behave as inschema-map.step, if given, MUST be one ofinputs,schema-map,apply,outputs; any other value MUST raiseValueError.- To iterate on a
schema-crosswalk-generated crosswalk (hand-edit it, then re-apply), re-run standaloneschema-refactoron the written crosswalk file; re-runningschema-crosswalkalways maps fresh, discarding any hand edits.