A simple and fast text search and replace tool for Windows.
- .NET 10 Runtime (download)
| Option | Description |
|---|---|
| Start directory | Directory to search in. |
| File mask | One or more masks separated by ;, , or |, e.g. *.cs;*.xaml. The last 10 masks are kept in the drop-down. |
| Text | Plain text, or a .NET regular expression if Regex is checked. The search works line by line, so a match cannot span several lines. |
| Replace with | Replacement text, see Replacing. |
| Search in | File content, File name or both. |
| Encoding | Encoding for files without byte order mark (BOM), see Encoding. |
| Case sensitive | Applies to plain text and regular expressions. |
| Include subdirectories | Searches the whole directory tree. |
All files matching the file mask are searched, including hidden, system and binary files. The .git directory is skipped. Files and directories that cannot be read, like locked files, files without access or broken ZIP archives, are skipped and counted in the status bar.
Click a file to expand its matches. The match is highlighted in its line. A line with several matches is listed once per match, each row highlights its own match. Long lines are cut around the match. The context menu opens a file in Explorer or a match in Notepad++.
A match in a file name is listed as the first row of the file's matches, marked with a blue Name badge instead of a line number. When searching in both content and file name, each file is listed once with all its matches.
You can search within ZIP files by including *.zip in your file mask. The file masks are applied to each file within the archive, allowing you to search large archives without prior extraction.
Nested archives are ignored. Matches inside archives cannot be replaced.
Files with a BOM are read with the encoding of the BOM (UTF-8, UTF-16 LE/BE, UTF-32 LE/BE).
Files without BOM are read with the encoding selected in Encoding: UTF-8 (default) or the ANSI code page of the system (e.g. Windows-1252). The encoding of such files cannot be detected reliably, so you have to choose.
The search tolerates a wrong encoding: ASCII text is still found, only non-ASCII characters may not match.
Binary files are searched like text files. Set Encoding to ANSI to find ASCII strings reliably, because every byte is one character then. Strings stored as UTF-16 (common in Windows and .NET binaries) are not found.
Replace replaces the selected matches in the current result with the text in Replace with. With Regex, the replacement can refer to capture groups, e.g. $1.
All matches are selected after a search. Use the check boxes to exclude single matches, or a whole file with the check box in its header.
Replacing never guesses, it refuses files instead of risking to corrupt them. A file is not modified and an error is shown if
- the file was modified after the search, so the found positions are no longer valid,
- the file is not valid in its encoding (e.g. an ANSI file searched as UTF-8),
- the replacement text contains characters that cannot be stored in the file's encoding (e.g.
→in Windows-1252).
Only the matched text is modified. All other bytes stay unchanged, including the BOM, the line endings (CR LF, LF or CR, even mixed) and a missing line break at the end of the file. This also allows replacing text in binary files, e.g. a version string, if Encoding is set to ANSI.
The file is written to a temporary file first, which replaces the original at the end. If anything fails, the original file is untouched. Large files are processed without loading them into memory.
After replacing, the result list is cleared, because the found positions and file names are outdated.
Matches in file names rename the file. The content is modified first and the file is renamed afterwards. A rename is refused if the new name is invalid or a file with that name already exists. Only files are renamed, not directories.
Run the tests with:
dotnet test
Issues and pull requests are welcome.