What the analyzer checks
Select local files to inspect them in your browser. The report shows the detected encoding and confidence, BOM status, line-ending style, byte size, line, word, character, and longest-line counts, binary likelihood, replacement-character count, and a preview. Processing is read-only: files are not modified.
How to analyze files
- Click browse files, or drop one or more files onto the drop zone.
- Selecting more files appends them to the current batch. Use Clear all to remove the batch.
- Review the results in the analysis table.
Read the results and use filters
Encoding is identified from BOMs and byte patterns. Without a BOM, valid UTF-8 and ASCII can be reported with high confidence; a clear null-byte pattern can indicate BOM-less UTF-16 with heuristic confidence. Legacy-looking byte patterns remain Ambiguous / unidentified (possible legacy 8-bit encoding); the tool does not assign a specific legacy encoding.
Line endings are reported as LF, CRLF, CR, Mixed, None (single line), or N/A (empty). When available, counts include lines, words, characters, and longest line. Words are runs of non-whitespace characters. Character and longest-line values use JavaScript string length, so they count UTF-16 code units.
Use the filter checkboxes to narrow the table to likely-binary files, mixed line endings, invalid/replacement characters, or ambiguous encodings. Checked filters are combined. The count can show filtered results as N of M files.
Preview, search, sort, and export
Set Preview length (characters) from 20 to 2000 characters; the default is 200. Preview text updates automatically when the setting changes. Search by file name, click a table heading to sort, or export the full batch with CSV or JSON. Exports include all files in the batch, not only filtered rows. Previews are omitted for likely-binary and over-limit files.
Limits and interpretation
- The picker accepts arbitrary files, not only text files.
- Each file is limited to
200 MiB. A larger file shows Not analyzed (file too large), gets a too large flag, and has unavailable counts. - Likely-binary detection samples up to the first
8,000bytes. It flags a file when the control-byte ratio is greater than2%, excluding tab, LF, and CR. Recognized UTF-16/UTF-32 files are checked for decoded control characters instead of their normal encoding bytes. This is a heuristic, not definitive file-type detection. Likely-binary files have unavailable line, word, character, and longest-line counts. - The replacement metric counts
U+FFFDcharacters in the decoded result. Malformed UTF-16 surrogate sequences and incomplete UTF-16/UTF-32 units are represented by replacement characters; the tool does not claim comprehensive validation of every encoding.
Frequently asked questions
Does this tool convert encoding or fix line endings?
No. It reports encoding, BOM presence, line-ending style, and counts. It does not convert encoding, normalize line endings, or modify files.
Can it always tell me the exact encoding?
No. A BOM makes UTF-8, UTF-16, and UTF-32 identification reliable. Without one, valid UTF-8, ASCII, and some BOM-less UTF-16 patterns can still be identified. Legacy-looking data remains ambiguous and is not assigned a specific legacy encoding.
How does it handle binary files?
Files that match the control-byte heuristic are flagged as likely binary. Their text counts and previews may be unavailable.
How is this different from the Text Analyzer or cleanup tools?
Those tools work on pasted or edited text. This tool reports file-level encoding, BOM, line endings, size, and related metrics without changing file content.
Does it upload my files anywhere?
No. Files are read locally using the browser's File APIs, and the tool does not modify them.