Personal Archive Search

What it is for

Years of notes end up in many small files: a renovation log per room, meeting notes per month, recipes, letters. When you need "where did I write down the worktop delivery date?", opening file after file is slow, and a desktop search often just says which file matches. This tool searches up to 100 of your own text documents together and lists every match with the document, line, column and the full line as an excerpt, so you can cite or open the exact passage.

It searches plain text only: no PDF text extraction, no OCR for scans and no regular expressions. In the browser interface the search runs locally on your device. If you use the API or the form without JavaScript, the supplied text is processed on the server for that request.

Worked example

The default searches one file, notes.md, for "cabinet" and finds 1 match on Line 2 · column 10: "The blue cabinet is 600 mm wide."

A small renovation archive with two files:

• 2025-kitchen.md: "Ordered the oak worktop." and "Worktop delivery: 14 March."
• 2026-bathroom.md: "Tiles chosen." and "The worktop in the utility room needs oiling."

Search phrase "worktop", Literal phrase, Whole words only on, case ignored. Result: "3 matches across 2 documents": line 1 column 17 and line 2 column 1 in the kitchen file, line 2 column 5 in the bathroom file. Switch on Match case and set File name contains to "2025", and only line 1 of the kitchen file remains, because "Worktop" with a capital W no longer counts.

The match modes behave differently. "worktop oiling" as All words per line finds 2 matches, both on the bathroom line, the only line containing both words. As Any word it finds 4. With Whole words only off, "work" also matches inside "worktop" and returns 3 hits. Ignore accents lets "cafe muller" find "Café Müller" in a note.

How to read the result

• Matches counts every occurrence; Matching files counts documents with at least one hit; Documents and Characters describe what was searched.
• Each hit shows the file name and its position in the list, Line … · column … and the line itself, with Open passage to jump into the full text.
• Filter displayed matches and Matching document narrow a long hit list without searching again.
• All stored matches and CSV export gives a table of every hit for a spreadsheet.

Tips and common mistakes

• Columns count characters from 1, so column 1 is the first character of the line.
• All words per line works line by line. If the words are on neighbouring lines, use Any word and look at nearby hits.
• Turn on Whole words only for short words; "oak" otherwise also finds "soak".
• Give files distinct names. Duplicates are told apart by their position, but readable names make the result easier to use.
• Use Add text files to load many .txt or .md files at once instead of pasting them row by row.
• Only the first 500 hits are listed; refine the phrase or add a file filter if you hit that limit.

FAQ

Can I search PDFs or Word files?

Not directly. Save or export them as plain text first; the tool does not extract text from PDFs or images.

Are my files uploaded?

In the enhanced browser interface the search runs locally. The API and the no-JavaScript form send the text to the server to process the request.

How large can my archive be?

Up to 100 documents with two million characters in total.

Why does "Café" not match "cafe"?

Accent-insensitive search is off by default. Turn on Ignore accents.