LineEndingNormalizer

LineEndingNormalizer (LEN) is a Windows command-line tool that changes text line endings to CRLF, LF, or CR without changing the file's encoding.

It preserves the original byte-order mark (BOM), file attributes, and timestamps. Files that cannot be handled safely are left unchanged.

To change character encodings, use EncodingChecker.

Safe workflow

1. Preview the operation with -WhatIf.
2. Review files that would change or be refused.
3. Run the same command without -WhatIf.
4. Add -Backup when you want an independently hash-verified .bak copy.

LineEndingNormalizer.exe -BasePath C:\Source -Include "*.cs,*.txt" -Target LF -WhatIf
LineEndingNormalizer.exe -BasePath C:\Source -Include "*.cs,*.txt" -Target LF -Backup

What LEN does

• Detects ASCII, UTF-8, UTF-16, UTF-32, and a conservative set of legacy encodings.
• Normalizes CRLF, LF, and CR. Unicode NEL, LS, and PS are also normalized in Unicode files.
• Streams large files instead of loading them fully into memory.
• Strictly decodes and re-encodes Unicode; invalid input is refused.
• Copies legacy text byte-for-byte except for CR and LF bytes, and only when that operation has been verified safe for the detected encoding.
• Writes to a temporary file, verifies the result, and installs it only after all checks pass.
• Skips linked files and linked directories in every mode.

Important safety limit

Without a byte-order mark, UTF-16 bytes usually read as valid text in both byte orders, so the file itself cannot say which one is right. LEN refuses to normalize BOM-less UTF-16 unless the bytes prove the byte order -- which for ordinary Latin text they do not, because byte-swapped Latin characters land in a valid CJK range. Expect most BOM-less UTF-16 files to be refused.

Conversion proceeds only when the opposite byte order is structurally impossible. Adding a correct BOM, or converting the encoding with an explicitly chosen source codec, resolves the ambiguity.

See Safety and recovery for the complete safety model and known limits.

Common commands

Preview conversion:

LineEndingNormalizer.exe -BasePath . -Include "*" -Target CRLF -WhatIf

Validate in CI without writing:

LineEndingNormalizer.exe -BasePath . -Include "*.cs" -Target CRLF -ValidateOnly -FailOnChanges -Quiet

Detect encoding and line endings only:

LineEndingNormalizer.exe -BasePath . -Include "*" -DetectOnly > detection.csv

Write a detailed CSV report:

LineEndingNormalizer.exe -BasePath . -Include "*" -Target LF -WhatIf -Report report.csv

Run LineEndingNormalizer.exe --help for the complete switch reference and examples, or read Command-line reference.

Documentation

• Command-line reference
• Safety and recovery
• Encoding detection and supported text
• Independent audit
• Release checklist

Reports

-Report writes UTF-8 CSV with these fields:

File,Encoding,BOM,LineEnding,Target,Result,ReasonCode,Diagnostic

ReasonCode is stable for scripts. Diagnostic explains the particular failure for people. The report always describes the original file's detected state, even when normalization succeeds.

Requirements and downloads

LEN targets Windows and .NET 10. Download the framework-dependent or self-contained build from the latest release.

To build locally:

dotnet build LineEndingNormalizer.slnx --configuration Release
dotnet test LineEndingNormalizer.slnx --configuration Release

Related projects

• EncodingChecker converts text encodings through a strict, verified pipeline.
• CorpusTesters audits the shared detector and conversion behavior against public corpora.

License

MPL 2.0. Third-party notices are in THIRD-PARTY-NOTICES.md.