-
Notifications
You must be signed in to change notification settings - Fork 0
DAT Audit
The DAT Audit window checks your files against one or more DAT catalogues — community-maintained lists that map a file's content hash to its canonical name — and reports what you have, what's misnamed, what's unrecognised, and what's missing. It's the verification companion to the Scraper: the scraper fetches metadata, the audit verifies your files.
DAT catalogues exist for many curated media sets (reference archives, audio
collections, video libraries, and more). If you have a .dat / .xml catalogue
for your collection, the audit tells you at a glance how complete and how
correctly-named it is.
Open it from File → DAT Audit…, or from a collection's settings page (Audit collection…) to start pre-aimed at that collection. The window has a sidebar with three sections — Audit, DAT Library, and Download — matching the rest of the app's look.
-
Add DAT files — click Add DAT… and pick one or more
.dat/.xmlcatalogues. Hover an entry to see what the catalogue calls itself and how many entries it carries. A catalogue shipped as a single.datzipped inside a.zipis read transparently — point Kartend straight at the.zip(an archive tool such asunzip/7zmust be on your PATH). - Add scan folders — click Add folder… and pick the folder(s) holding the files to check.
- Click Run audit. Hashing runs in the background with a live progress bar; click Cancel to stop.
When it finishes, the summary line shows the totals and the table lists every file and every catalogue entry with a status:
| Status | What it means | What to do |
|---|---|---|
| Have | Correct file, correct name | Nothing |
| Wrong name | Right content, wrong filename | Rename to the canonical name |
| Wrong content | Right name, wrong/old content | Re-acquire a good copy |
| Duplicate | A second copy of something you already have | Optional cleanup |
| Unknown | Not in any catalogue | Identify or set aside |
| Corrupt | The file could not be read or extracted | Re-acquire it |
| Missing | In the catalogue, not in your folder | Acquire it |
Use the Show filter to focus on one status (e.g. just Missing, or just Wrong name).
A profile saves an audit setup — DAT files, scan folders, and the options below — so a re-check is one click. Manage them with the Profile row at the top of the window (New / Edit / Duplicate / Rename / Delete). The profile editor also holds:
- 1G1R region priority — when a catalogue lists the same title for several regions, count only your preferred region toward completeness.
-
Ignore globs — filename patterns (one per line, e.g.
*.txt) excluded from the scan. - Fix output mode — see Fixing problems below.
A profile can be linked to a collection (the Linked collection picker in the profile editor). For a linked profile, the collection's own settings are the single source of truth: the DAT list and scan folder are taken live from the collection's configuration page, and the editors in the audit window lock with a hint saying so. Edit the DAT list on the collection's settings page; the audit picks the change up automatically.
After a completed audit of a linked (saved) profile, the results persist: the collection's settings page shows Last audited with the present / missing counts, without re-running anything.
Instead of attaching catalogues one by one, point Kartend at a DAT library — a folder where you keep your catalogue files. Set it on the DAT Library page (Scan & review proposals…).
- At startup, Kartend quietly checks the library for new or updated catalogues. When one looks like it belongs to one of your collections (its header name resembles the collection's name), a status-bar hint appears.
- The Library review lists each proposed match with a per-row picker. Each
picker offers, in addition to the candidate collections:
- (No collection — keep in library) — leave the catalogue in the library folder, unattached. It isn't dismissed, so a later scan can propose it again.
- Add to new collection… — create a brand-new collection (you choose the name and content folder) with this catalogue already attached.
- Apply selected carries out the chosen action for the selected rows; Dismiss selected stops Kartend from asking about that catalogue again — until the file itself is updated, which counts as a new revision worth asking about.
- Nothing is ever attached automatically, and associating is optional — you can import catalogues into the library without attaching them to anything.
The Download page fetches catalogue packs from an online service and imports them straight into your DAT library. Pick a Source at the top:
- No-Intro — enter the service's system id (or paste a daily-download URL) and click Load; Kartend shows the available pack date and the sets you can include, each with its real name. Tick the sets you want.
- Redump — pick a system by name from the dropdown (the list loads the first time you choose this source; Refresh list reloads it).
Then click Download & import: the pack downloads in the background (progress bar + Cancel), and every catalogue inside it is extracted into your DAT library. When the import finishes, the library review opens so you can attach the new catalogues to your collections — or leave them in the library, or spin up a new collection for one. Importing never forces you to associate anything.
If you haven't set a library folder yet, Kartend asks for one the first time you download. Downloads run one at a time and report clearly if the service is unreachable or returns something unexpected.
Not every cataloguer offers an in-app download. For those — and for packs you
already have — the DAT Library page has Import DAT zip… and Import DAT
folder…: point Kartend at a downloaded DAT archive, a folder of DATs, or a
single .dat, and it copies them into the library and runs the match review,
exactly like a download. Any Logiqx-style catalogue works.
Folders are organised in different ways — all files directly in one folder, spread across subfolders, or one archive per item. The audit can detect which shape a scan folder has: click Detect structure, or let the automatic probe run when you open the audit for a collection. The result appears as a banner suggestion; Apply confirms it, and only a confirmed layout changes how the audit scans:
- Flat — subfolders are skipped (they're noise, e.g. artwork or notes).
- Nested — everything is scanned recursively (the default behaviour).
- Archive-per-item — the audit opens each archive and checks the files inside it against the catalogue. Without this, an archived set would report nearly everything as Unknown and Missing, because the archive's outer bytes never match the catalogue's content hashes.
- Subfolder-per-item — each item lives in its own subfolder as a multi-file set (e.g. multi-part masters or disc-image pairs). Scanned recursively like Nested; the distinct label just reflects the shape the probe found.
The first audit of a large archive set extracts and hashes every member, which takes a while — the progress bar shows real counts, and the results are cached so the next audit only re-reads archives that changed.
Fix… opens the fix dialog over the current results:
- Rename wrong-named files to their canonical names, in place.
- Relocate present entries into a clean managed-output folder (copies; originals untouched).
- Quarantine unknown files into a side folder (moved, never deleted).
Every action is previewed first and undoable afterwards. Files inside archives are reported but not modified — fixing rewrites loose files only.
When a fix renames files in a collection-linked audit, Kartend offers to re-scrape that collection afterwards: a canonical filename matches metadata far more reliably than a wrong one, so renamed items are exactly the ones worth re-fetching. It's a prompt, not automatic — say no to skip it.
- Export CSV — the full table as a spreadsheet, one row per entry with a Status column you can sort/filter.
- Export fixdat — a DAT file containing only the missing entries. Hand it to any DAT tool to fetch exactly the gaps.
- Export miss list — a plain-text list of the missing titles.
- Auditing matches on content hash, so a correctly-downloaded but wrongly-named file is still recognised (and flagged Wrong name).
- A first audit hashes every file, which can take a while for large media; the progress bar shows how far along it is, and unchanged files are skipped on re-audits.
- The DAT files themselves are never reported, even if they sit inside a scan folder.