Local Text File Search

How to use Local Text File Search

Search your selected local UTF-8 text files for a literal string and download exact line, column and source-offset evidence. Search a small file collection directly in the browser without sending its contents to an indexing service. Keep per-file zero-match counts and original source witnesses, and choose an explicit case policy instead of running a supplied regular expression.

1. Provide local utf8 text files, single-line literal search string.
2. Choose the supported mode and run the tool. Review its assumptions and any reported errors.
3. Copy the result or download a separate output; reset to clear the current form.

Good to know

• Select 1–20nonempty UTF-8 files, at most 200,000 bytes each and4MiB aggregate. Names are exact distinct single-line strings up to200 bytes; no filesystem paths are traversed or files renamed. All file-size/name bounds precede decoding.
• UTF-8 decoding is fatal and preserves BOM. One file may contain a retained initial BOM; it occupies a source position. At most30,000 logical lines across files, including a final empty line after a terminating newline. CRLF, CR and LF are recognized without rewriting source offsets.
• Query is a nonempty well-formed single-line literal up to2,000 UTF-8 bytes. Exact compares original code units; ASCII case-insensitive folds A–Z only. No regex, full Unicode folding, normalization, word semantics or matches across newlines.
• Matches are nonoverlapping. Each includes1-based line/UTF16-column and0-based whole-file UTF16 offset, not UTF-8 bytes or grapheme columns. At most5,000 matches; overflow refuses the complete export rather than silently truncating results.
• Each match keeps its exact source spelling and first200UTF16 units of that line, clipped before a surrogate split. A match can lie beyond its truncated snippet. The report can expose names, query and source text; it is not anonymization.
• JSON preserves literal witnesses; CSV uses formula-prefix protection on source cells. Report text is at most 500,000 characters. No upload, automatic persistence, background index or user code execution occurs.
• A same-origin module Worker loads its selected code, including the YAML writer for Dependabot, before ready within 10 seconds. A separate2-second computation deadline follows. Source is never evaluated, fetched, submitted, uploaded or automatically saved.
• Structured text output is limited to 500,000 characters. Reduce the supplied source when that separate output budget is exceeded.
• Inputs are processed on this device and are not automatically saved or uploaded.