MAT Tagger CleanerPrivate by design · Files stay on your device← All toolkits

DAILWIND / CORPUS PREPARATION

MAT Tagger Cleaner

Prepare and organize your text corpora for linguistic research.

01 · Select Files

Add books or folders. Keep independent collections organized.

02 · Prepare Files

Text is checked and prepared automatically on your device.

03 · Download

Get your prepared corpus, source mapping and report in one ZIP.

Add your text files to get started.
0 / 0Documents processed
0Total prepared words
0Documents prepared
Specialist checks & earlier cleaning options

Optional tools for existing annotations and deliberate text repairs. These checks are separate from automatic preparation. Their downloads keep the earlier workflow format.

DAILWIND / MAT PREPARATION

Prepare the text.
Protect the evidence.

A focused preparation desk for MAT workflows. Check the input, keep annotations intact and separate files that need attention before counting.

Add .txt files or try the samples. This utility prepares and checks files; it does not linguistically tag them.

Technical preflight notes.

The examined MAT counter divides total word length by its counted-token baseline at count_only.pl line 100. A zero baseline can trigger that error. This preflight checks that condition, but the exact cause in an individual file and every downstream MAT calculation cannot be guaranteed.

01 · Plain text goes to the tagger.

Use MAT Tag or Tag and Analyze. A cleaned text file is not tagged output. Changing its filename does not add annotations.

02 · Final MAT output goes to counting.

Use the count-input/ folder only after tagged preflight. Empty, zero-baseline, intermediate and malformed inputs stay in review/.

02 / REVIEW & SAVE

Every file gets a clear next step.

No results yet.

Original preview

Prepared preview
Preview: first 12,000 characters. UTF-8 without BOM downloads contain full prepared content. Originals are unchanged.
ALSO FOR YOUR COMPUTER
The same workflow. Offline.
Dailwind MAT Cleaner for Python includes a desktop interface and command-line batch mode, folder selection, strict encoding checks, separate output folders and CSV/JSON audits. Python 3.10+ with Tkinter is required; no pip dependencies or Internet processing.
Download Python desktop app ↗
Use the right output folder.
  1. ready-for-tagging/ contains prepared plain text. Tag it in MAT first.
  2. count-input/ contains only files passing tagged syntax / baseline checks. Select this folder in MAT counting.
  3. review/ contains ambiguous or blocked copies. Never mix them into counting.
  4. Keep audit files outside MAT input folders. Check your MAT installation, Java/Stanford setup and original tagging output if counting still fails.
This is an independently written Dailwind utility, not an official MAT product. It preserves numbers, case and function words. It does not invent missing words, assign POS/features, run dimension scores or verify tagger accuracy. Examined MAT counter ↗ · Official MAT documentation ↗
Your prepared file is ready.
Download file

 
About preparation, privacy & research integrity
Original wording, order, case, punctuation and deliberate repetitions are preserved. Automatic changes standardize text encoding, line endings, page-break whitespace and nonbreaking spaces. Suspected page markers, scanning notes and OCR placeholders are noted in the report and kept in the text.
Work runs locally in a Web Worker. Temporary local disk storage is used where supported; otherwise a bounded memory fallback is used. Capacity depends on your browser and free storage. Clear the session after downloading to release temporary files.
Books are automatically organized into numbered text parts. The manifest maps each part to its original file. The book folders and MAT input contain identical prepared copies: analyze only one copy. No tagging or MAT analysis is performed here. Read the technical preparation guide ↗