Skip to content

[Feature] Recover ARC loading when canonical workbooks are zero-byte files #620

@Etschbeijer

Description

@Etschbeijer

Use case

An interrupted ARC entity creation can leave a zero-byte canonical workbook such as assays/<id>/isa.assay.xlsx. A later ARC.tryLoadAsync fails while reading that workbook with an error containing data length = 0, leaving the ARC unusable until the file is manually repaired or removed.

Requested behavior

Please provide an ARCtrl recovery path for zero-byte canonical workbooks, or document the intended recovery API. For a canonical entity workbook whose identifier can be derived from its path, regenerating the empty workbook scaffold allows the ARC to load again.

This applies to canonical assay, study, workflow, and run workbooks. It should not silently replace non-empty corrupt files.

Current workaround

Swate has a temporary ARC.LoadAsyncSwateZeroByteRepair extension. It first loads normally, and only when loading reports the zero-byte ZIP error does it find zero-byte canonical workbooks, recreate their empty scaffold contracts, and retry loading. The workaround can be removed once an ARCtrl release handles this case.

Metadata

Metadata

Assignees

Labels

Type

No type
No fields configured for issues without a type.

Projects

Status
No status

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions