How to analyze file encoding and line endings
Select local files and read their bytes in File Encoding & Line-Ending Analyzer. Each row shows the file name, size, Encoding, BOM, Line endings, Lines, Words, Chars, Longest line, Flags, and Preview. The page keeps the files unchanged and makes no request that uploads their contents.
Check file encoding and line endings
- Click browse files, or drop files onto the drop area.
- Select another batch to append those files to the current results.
- Set Preview length (characters) before reviewing the text preview.
- Read the table, apply filters, then click Clear all to remove the current batch.
How encoding detection works without a BOM
A BOM identifies UTF-8, UTF-16 LE or BE, and UTF-32 LE or BE with High (BOM) confidence. Without a BOM, valid UTF-8 and ASCII receive High confidence. A null-byte pattern can identify UTF-16 LE or BE with Medium (heuristic) confidence. Other byte sequences receive Ambiguous / unidentified (possible legacy 8-bit encoding).
The Line endings value is LF, CRLF, CR, Mixed, None (single line), or N/A (empty). Words are whitespace-separated runs. Chars and Longest line use JavaScript string length in UTF-16 code units.
Hello
MartaFor this ASCII file with LF between the two names, the row shows:
Encoding: ASCII
Line endings: LF
Lines: 2
Words: 2
Chars: 11
Longest line: 5Find mixed line endings and invalid characters
Check Mixed line endings only to keep files whose decoded text contains more than one of LF, CRLF, and CR. Invalid/replacement characters only finds decoded U+FFFD characters. These include replacements produced by malformed UTF-8, unmatched UTF-16 surrogates, and incomplete UTF-16 or UTF-32 units. The count describes decoded text; it does not validate every encoding.
Likely binary only samples the first 8,000 bytes for narrow encodings and the first 8,000 decoded characters for UTF-16 and UTF-32. A control-byte ratio above 2% flags the file, with tab, LF, and CR excluded. Likely binary files can still show Encoding and Line endings, while text counts and Preview are unavailable.
Search, sort, preview, and export file results
Search files matches file names. Click a sortable heading, including Encoding, Line endings, or Longest line, to reorder the table. Reset filters clears the search and checkbox selections. Checked filters apply together.
Preview length (characters) accepts whole numbers from 20 to 2,000 and starts at 200. A valid change rebuilds the previews. Download CSV and Download JSON include every file in the batch, including rows hidden by search or filters. Charts use the full batch too.
What happens with empty, binary, or large files
The picker accepts arbitrary files. An empty file receives N/A (empty) for Line endings and zero for its text counts. A file larger than 200 MiB is recorded as Not analyzed (file too large), marked Too large, and has no text counts. Likely binary files can still show Encoding and Line endings.
Where file encoding checks come up
When a log displays incorrectly, inspect its Encoding, BOM, replacement count, and Preview before changing the reader configuration.
For a batch of source files, check Mixed line endings only, then use Download CSV to keep the table values for review.
Choose the right file analysis tool
Use File Management Analyze Tools for local file or folder inspection. Choose File Type Detector for a file's format, signature, MIME type, or extension mismatch. Choose Duplicate File Finder for byte-identical files grouped by content hash. Choose Folder Size Analyzer for storage totals, item counts, largest items, or a treemap.
Frequently asked questions
How do I analyze file encoding and line endings?
Select files with browse files or drop them into the drop area, then read the Encoding, BOM, Line endings, count, flag, and Preview columns in the table.
Can this detect UTF-16 without a BOM?
Yes. A null-byte pattern can produce a UTF-16 LE or UTF-16 BE result with Medium (heuristic) confidence. Other uncertain byte sequences remain ambiguous.
How do I find mixed line endings in files?
Check Mixed line endings only. The table keeps files whose decoded text contains more than one of LF, CRLF, and CR.
Why is a file marked likely binary?
A file is marked likely binary when its sampled control-byte ratio exceeds 2%, with tab, LF, and CR excluded. Its text counts and Preview are unavailable.
Can I export the full file analysis?
Yes. Download CSV and Download JSON export the complete batch, including files hidden by Search files or filters.