How to use this kit
About this AppKatz Kit
AppKatz Kits · Import Template Mapper · v1.0. Original implementation authored for AppKatz, 2026. The AppKatz logo, wordmark, tool identity, source attribution, recipe identifier, and human-readable review reports carry the AppKatz identity. Mapped CSV data stays free of branding rows and added columns.
These visible touches deter casual copying; client-side code remains inspectable and can be modified. There is no activation check, tracking, anti-debugging, or restriction on normal keyboard and copy behavior. No third-party code or assets are included.
- Choose the delimiter and load a UTF-8 source CSV. Load your destination’s header row or sample CSV.
- Map each target to a source field, constant, or intentional blank. “Match exact names” is an optional starting point.
- Add transformations and rules. Choose MDY or DMY explicitly for slash dates. Reorder columns if needed.
- Validate, review exceptions, and download the accepted file and rejected-row report.
- Save the recipe for next time. Load both CSVs before importing a recipe.
This checks the rules you select. It does not guarantee acceptance by a third-party system. Check that system’s current import requirements.
Examples & rule behavior
Identifier: 00123 stays 00123. No numeric coercion is performed.
Date: 03/04/2026 becomes 2026-03-04 with MDY, or 2026-04-03 with DMY. ISO requires YYYY-MM-DD; slash formats allow one or two digits, with a four-digit year.
Example data: five synthetic contacts show quoted commas, Unicode, a multiline note, an invalid date, a missing email, duplicate IDs, an enum mismatch, and a formula-like note. One row passes the supplied rules.
How rules decide what passes
Required rejects empty or whitespace-only final values. Optional date and enum rules skip only an exactly empty value. Enums are case-sensitive, one exact value per line; spaces are significant. Uniqueness is case-sensitive and checks final non-empty values across all source records. Every occurrence of a duplicate is rejected, including matches on otherwise invalid rows. No automatic deduplication or repair occurs.
CSV width mismatches reject the affected source record; malformed quoting rejects the entire file. Truly blank records are kept, not silently skipped. Source record numbers count CSV records, so embedded newlines do not change the numbering.
Limits, privacy & troubleshooting
Hard limits: 10 MiB per CSV, 50,000 data records, 100 columns, 1,000,000 cells including headers, and 100,000 characters per field; 200 per header. Transformed values are capped at 20 Mi characters total; validation reasons at 5 Mi characters. A recipe is limited to 1 MiB. Processing runs in a worker; Cancel discards the in-progress operation.
Verified on Windows with Chrome 154: a source with 50,000 rows × 19 columns (8.55 MB; 950,019 cells including headers), cancellable processing, downloads, and 390 px layout. Engine boundary tests also covered exactly 10 MiB, 100 columns, 1,000,000 cells, and 100,000-character fields. Other browsers and physical mobile devices are not yet qualified.
Verified local processing: source inspection and a Chrome network audit of loading, mapping, validation, recipes and downloads observed no outgoing HTTP(S) or WebSocket requests. This file has no external assets, analytics, ad scripts, or storage API; its content security policy blocks outgoing connections. This verification covers this standalone file, not browser extensions or a future hosting page.
File and recipe troubleshooting
One giant column usually means the wrong delimiter: choose it and reload the file. Duplicate or blank headers must be fixed upstream; header spelling and spaces are preserved. Invalid UTF-8, UTF-16 and NUL bytes are rejected; re-save as UTF-8 CSV.
Recipes require exact, ordered source and target headers. Renamed, missing, added or reordered columns block the import; the current setup is retained. Remap and save a new recipe after reviewing the new schema. Recipes include constants and rules, which may contain sensitive text; store and share them accordingly.
The kit keeps data in memory only. Reloading or clearing the workspace removes the current setup; downloaded files remain on your device. A failed or cancelled import keeps the last successful setup. Choosing a different export policy may alter values or expose formula risk. Apostrophe escaping is a precaution, not a guarantee across every spreadsheet program.
Use a current desktop browser with JavaScript, Web Workers, TextDecoder and Blob downloads. Locked-down browsers may block local workers or downloads. Mobile layout is supported, but large-file performance on mobile has not been qualified. This version does not read XLSX, infer delimiters, merge files, evaluate formulas, or validate destination-specific business rules.
