How a scan runs
- The browser decodes the source as UTF-8 and records a review when replacement characters appear.
- Papa Parse reads comma, tab, or semicolon data with values kept as strings, preserving identifiers such as
001. - MergeLint checks header names, data-row width, and image-field conventions.
- The offline app can compare image paths with filenames and relative paths from a folder you explicitly select.
- Findings are ordered as blockers, reviews, and notes, then by row, field, and stable code.
Finding origins
| Origin | Meaning | Examples |
|---|---|---|
| Adobe | A source convention documented for Data Merge. | Named fields, image-field headers, source paths. |
| File fact | A directly observed property of the selected data or paths. | Duplicate header, uneven row width, case mismatch. |
| Advisory | A conservative prompt for human review, not an Adobe requirement. | Every named field appears to be an image field. |
Rule boundary
The implementation checks blank and duplicate headers, edge whitespace, inconsistent row widths, no-data sources, UTF-8 replacement characters, image-path whitespace, exact and case-only path matches, ambiguous basenames, missing selected assets, and whether asset existence was outside the scan scope.
It does not open an InDesign document or make claims about fonts, overset, layout, merged output, PDF settings, color, bleed, or print readiness.
Primary references
- Adobe InDesign User Guide: Data merge — source formats, field names, image fields, quoted text, and paths.
- Papa Parse documentation — parser behavior and CSV serialization used in local scans and report export.
Community discussions informed the decision to test this problem, but they do not define product rules or prove willingness to pay.
Reproducibility
The same bytes, selected asset paths, delimiter choice, and MergeLint version produce the same ordered findings. The report includes its version-relevant metadata and scan counts. Timestamps differ by run.