Skip to content

Latest commit

 

History

History
1046 lines (814 loc) · 40.7 KB

FAQ.md

File metadata and controls

1046 lines (814 loc) · 40.7 KB

FAQ

Does ripgrep support configuration files?

Yes. See the guide's section on configuration files.

What's changed in ripgrep recently?

Please consult ripgrep's CHANGELOG.

When is the next release?

ripgrep is a project whose contributors are volunteers. A release schedule adds undue stress to said volunteers. Therefore, releases are made on a best effort basis and no dates will ever be given.

An exception to this can be high impact bugs. If a ripgrep release contains a significant regression, then there will generally be a strong push to get a patch release out with a fix. However, no promises are made.

Does ripgrep have a man page?

Yes. If you installed ripgrep through a package manager on a Unix system, then it would have ideally been installed for you in the proper location. In which case, man rg should just work.

Otherwise, you can ask ripgrep to generate the man page:

$ mkdir -p man/man1
$ rg --generate man > man/man1/rg.1
$ MANPATH="$PWD/man" man rg

Or, if your version of man supports the -l/--local-file flag, then this will suffice:

$ rg --generate man | man -l -

Note that the man page's documentation for options is equivalent to the output shown in rg --help. To see more condensed documentation (one line per flag), run rg -h.

The man page is also included in all ripgrep binary releases.

Does ripgrep have support for shell auto-completion?

Yes! If you installed ripgrep through a package manager on a Unix system, then the shell completion files included in the release archive should have been installed for you automatically. If not, you can generate completes using ripgrep's command line interface.

For bash:

$ dir="$XDG_CONFIG_HOME/bash_completion"
$ mkdir -p "$dir"
$ rg --generate complete-bash > "$dir/rg.bash"

For fish:

$ dir="$XDG_CONFIG_HOME/fish/completions"
$ mkdir -p "$dir"
$ rg --generate complete-fish > "$dir/rg.fish"

For zsh:

$ dir="$HOME/.zsh-complete"
$ mkdir -p "$dir"
$ rg --generate complete-zsh > "$dir/_rg"

For PowerShell, create the completions:

$ rg --generate complete-powershell > _rg.ps1

And then add . _rg.ps1 to your PowerShell profile (note the leading period). If the _rg.ps1 file is not on your PATH, do . /path/to/_rg.ps1 instead.

How can I get results in a consistent order?

By default, ripgrep uses parallelism to execute its search because this makes the search much faster on most modern systems. This in turn means that ripgrep has a non-deterministic aspect to it, since the interleaving of threads during the execution of the program is itself non-deterministic. This has the effect of printing results in a somewhat arbitrary order, and this order can change from run to run of ripgrep.

The only way to make the order of results consistent is to ask ripgrep to sort the output. Currently, this will disable all parallelism. (On smaller repositories, you might not notice much of a performance difference!) You can achieve this with the --sort path flag.

There is more discussion on this topic here: BurntSushi#152

How do I search files that aren't UTF-8?

See the guide's section on file encoding.

How do I search compressed files?

ripgrep's -z/--search-zip flag will cause it to search compressed files automatically. Currently, this supports gzip, bzip2, xz, lzma, lz4, Brotli and Zstd. Each of these requires the corresponding gzip, bzip2, xz, lz4, brotli and zstd binaries to be installed on your system. (That is, ripgrep does decompression by shelling out to another process.)

ripgrep currently does not search archive formats, so *.tar.gz files, for example, are skipped.

How do I search over multiple lines?

The -U/--multiline flag enables ripgrep to report results that span over multiple lines.

How do I use lookaround and/or backreferences?

ripgrep's default regex engine does not support lookaround or backreferences. This is primarily because the default regex engine is implemented using finite state machines in order to guarantee a linear worst case time complexity on all inputs. Backreferences are not possible to implement in this paradigm, and lookaround appears difficult to do efficiently.

However, ripgrep optionally supports using PCRE2 as the regex engine instead of the default one based on finite state machines. You can enable PCRE2 with the -P/--pcre2 flag. For example, in the root of the ripgrep repo, you can easily find all palindromes:

$ rg -P '(\w{10})\1'
tests/misc.rs
483:    cmd.arg("--max-filesize").arg("44444444444444444444");
globset/src/glob.rs
1206:    matches!(match7, "a*a*a*a*a*a*a*a*a", "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa");

If your version of ripgrep doesn't support PCRE2, then you'll get an error message when you try to use the -P/--pcre2 flag:

$ rg -P '(\w{10})\1'
PCRE2 is not available in this build of ripgrep

Most of the releases distributed by the ripgrep project here on GitHub will come bundled with PCRE2 enabled. If you installed ripgrep through a different means (like your system's package manager), then please reach out to the maintainer of that package to see whether it's possible to enable the PCRE2 feature.

How do I configure ripgrep's colors?

ripgrep has two flags related to colors:

  • --color controls when to use colors.
  • --colors controls which colors to use.

The --color flag accepts one of the following possible values: never, auto, always or ansi. The auto value is the default and will cause ripgrep to only enable colors when it is printing to a terminal. But if you pipe ripgrep to a file or some other process, then it will suppress colors.

The --colors flag is a bit more complicated. The general format is:

--colors '{type}:{attribute}:{value}'
  • {type} should be one of path, line, column or match. Each of these correspond to the four different types of things that ripgrep will add color to in its output. Select the type whose color you want to change.
  • {attribute} should be one of fg, bg or style, corresponding to foreground color, background color, or miscellaneous styling (such as whether to bold the output or not).
  • {value} is determined by the value of {attribute}. If {attribute} is style, then {value} should be one of nobold, bold, nointense, intense, nounderline or underline. If {attribute} is fg or bg, then {value} should be a color.

A color is specified by either one of eight of English names, a single 256-bit number or an RGB triple (with over 16 million possible values, or "true color").

The color names are red, blue, green, cyan, magenta, yellow, white or black.

A single 256-bit number is a value in the range 0-255 (inclusive). It can either be in decimal format (e.g., 62) or hexadecimal format (e.g., 0x3E).

An RGB triple corresponds to three numbers (decimal or hexadecimal) separated by commas.

As a special case, --colors '{type}:none' will clear all colors and styles associated with {type}, which lets you start with a clean slate (instead of building on top of ripgrep's default color settings).

Here's an example that makes highlights the matches with a nice blue background with bolded white text:

$ rg somepattern \
    --colors 'match:none' \
    --colors 'match:bg:0x33,0x66,0xFF' \
    --colors 'match:fg:white' \
    --colors 'match:style:bold'

Colors are an ideal candidate to set in your configuration file. See the question on emulating The Silver Searcher's output style for an example specific to colors.

How do I enable true colors on Windows?

First, see the previous question's answer on configuring colors.

Secondly, coloring on Windows is a bit complicated. If you're using a terminal like Cygwin, then it's likely true color support already works out of the box. However, if you are using a normal Windows console (cmd or PowerShell) and a version of Windows prior to 10, then there is no known way to get true color support. If you are on Windows 10 and using a Windows console, then true colors should work out of the box with one caveat: you might need to clear ripgrep's default color settings first. That is, instead of this:

$ rg somepattern --colors 'match:fg:0x33,0x66,0xFF'

you should do this

$ rg somepattern --colors 'match:none' --colors 'match:fg:0x33,0x66,0xFF'

This is because ripgrep might set the default style for match to bold, and it seems like Windows 10's VT100 support doesn't permit bold and true color ANSI escapes to be used simultaneously. The work-around above will clear ripgrep's default styling, allowing you to craft it exactly as desired.

How do I stop ripgrep from messing up colors when I kill it?

Type in color in cmd.exe (Command Prompt) and echo -ne "\033[0m" on Unix-like systems to restore your original foreground color.

In PowerShell, you can add the following code to your profile which will restore the original foreground color when Reset-ForegroundColor is called. Including the Set-Alias line will allow you to call it with simply color.

$OrigFgColor = $Host.UI.RawUI.ForegroundColor
function Reset-ForegroundColor {
	$Host.UI.RawUI.ForegroundColor = $OrigFgColor
}
Set-Alias -Name color -Value Reset-ForegroundColor

PR #187 fixed this, and it was later deprecated in #281. A full explanation is available here.

Why does using a leading `/` on Windows fail?

If you're using cygwin on Windows and try to search for a pattern beginning with a /, then it's possible that cygwin is mangling that pattern without your knowledge. For example, if you tried running rg /foo in a cygwin shell on Windows, then cygwin might mistakenly perform path translation on /foo, which would result in rg C:/msys64/foo being searched instead.

You can fix this in one of three ways:

  1. Stop using cygwin.
  2. Escape the leading slash with an additional slash. e.g., rg //foo.
  3. Temporarily disable path translation by setting MSYS_NO_PATHCONV=1. e.g., MSYS_NO_PATHCONV=1 rg /foo.

For more details, see BurntSushi#1277

How do I get around the regex size limit?

If you've given ripgrep a particularly large pattern (or a large number of smaller patterns), then it is possible that it will fail to compile because it hit a pre-set limit. For example:

$ rg '\pL{1000}'
Compiled regex exceeds size limit of 10485760 bytes.

(Note: \pL{1000} may look small, but \pL is the character class containing all Unicode letters, which is quite large. And it's repeated 1000 times.)

In this case, you can work around by simply increasing the limit:

$ rg '\pL{1000}' --regex-size-limit 1G

Increasing the limit to 1GB does not necessarily mean that ripgrep will use that much memory. The limit just says that it's allowed to (approximately) use that much memory for constructing the regular expression.

How do I make the -f/--file flag faster?

The -f/--file permits one to give a file to ripgrep which contains a pattern on each line. ripgrep will then report any line that matches any of the patterns.

If this pattern file gets too big, then it is possible ripgrep will slow down dramatically. Typically this is because an internal cache is too small, and will cause ripgrep to spill over to a slower but more robust regular expression engine. If this is indeed the problem, then it is possible to increase this cache and regain speed. The cache can be controlled via the --dfa-size-limit flag. For example, using --dfa-size-limit 1G will set the cache size to 1GB. (Note that this doesn't mean ripgrep will use 1GB of memory automatically, but it will allow the regex engine to if it needs to.)

How do I make the output look like The Silver Searcher's output?

Use the --colors flag, like so:

rg --colors line:fg:yellow      \
   --colors line:style:bold     \
   --colors path:fg:green       \
   --colors path:style:bold     \
   --colors match:fg:black      \
   --colors match:bg:yellow     \
   --colors match:style:nobold  \
   foo

Alternatively, add your color configuration to your ripgrep config file (which is activated by setting the RIPGREP_CONFIG_PATH environment variable to point to your config file). For example:

$ cat $HOME/.config/ripgrep/rc
--colors=line:fg:yellow
--colors=line:style:bold
--colors=path:fg:green
--colors=path:style:bold
--colors=match:fg:black
--colors=match:bg:yellow
--colors=match:style:nobold
$ RIPGREP_CONFIG_PATH=$HOME/.config/ripgrep/rc rg foo

Why does ripgrep get slower when I enable PCRE2 regexes?

When you use the --pcre2 (-P for short) flag, ripgrep will use the PCRE2 regex engine instead of the default. Both regex engines are quite fast, but PCRE2 provides a number of additional features such as look-around and backreferences that many enjoy using. This is largely because PCRE2 uses a backtracking implementation where as the default regex engine uses a finite automaton based implementation. The former provides the ability to add lots of bells and whistles over the latter, but the latter executes with worst case linear time complexity.

With that out of the way, if you've used -P with ripgrep, you may have noticed that it can be slower. The reasons for why this is are quite complex, and they are complex because the optimizations that ripgrep uses to implement fast search are complex.

The task ripgrep has before it is somewhat simple; all it needs to do is search a file for occurrences of some pattern and then print the lines containing those occurrences. The problem lies in what is considered a valid match and how exactly we read the bytes from a file.

In terms of what is considered a valid match, remember that ripgrep will only report matches spanning a single line by default. The problem here is that some patterns can match across multiple lines, and ripgrep needs to prevent that from happening. For example, foo\sbar will match foo\nbar. The most obvious way to achieve this is to read the data from a file, and then apply the pattern search to that data for each line. The problem with this approach is that it can be quite slow; it would be much faster to let the pattern search across as much data as possible. It's faster because it gets rid of the overhead of finding the boundaries of every line, and also because it gets rid of the overhead of starting and stopping the pattern search for every single line. (This is operating under the general assumption that matching lines are much rarer than non-matching lines.)

It turns out that we can use the faster approach by applying a very simple restriction to the pattern: statically prevent the pattern from matching through a \n character. Namely, when given a pattern like foo\sbar, ripgrep will remove \n from the \s character class automatically. In some cases, a simple removal is not so easy. For example, ripgrep will return an error when your pattern includes a \n literal:

$ rg '\n'
the literal '"\n"' is not allowed in a regex

So what does this have to do with PCRE2? Well, ripgrep's default regex engine exposes APIs for doing syntactic analysis on the pattern in a way that makes it quite easy to strip \n from the pattern (or otherwise detect it and report an error if stripping isn't possible). PCRE2 seemingly does not provide a similar API, so ripgrep does not do any stripping when PCRE2 is enabled. This forces ripgrep to use the "slow" search strategy of searching each line individually.

OK, so if enabling PCRE2 slows down the default method of searching because it forces matches to be limited to a single line, then why is PCRE2 also sometimes slower when performing multiline searches? Well, that's because there are multiple reasons why using PCRE2 in ripgrep can be slower than the default regex engine. This time, blame PCRE2's Unicode support, which ripgrep enables by default. In particular, PCRE2 cannot simultaneously enable Unicode support and search arbitrary data. That is, when PCRE2's Unicode support is enabled, the data must be valid UTF-8 (to do otherwise is to invoke undefined behavior). This is in contrast to ripgrep's default regex engine, which can enable Unicode support and still search arbitrary data. ripgrep's default regex engine simply won't match invalid UTF-8 for a pattern that can otherwise only match valid UTF-8. Why doesn't PCRE2 do the same? This author isn't familiar with its internals, so we can't comment on it here.

The bottom line here is that we can't enable PCRE2's Unicode support without simultaneously incurring a performance penalty for ensuring that we are searching valid UTF-8. In particular, ripgrep will transcode the contents of each file to UTF-8 while replacing invalid UTF-8 data with the Unicode replacement codepoint. ripgrep then disables PCRE2's own internal UTF-8 checking, since we've guaranteed the data we hand it will be valid UTF-8. The reason why ripgrep takes this approach is because if we do hand PCRE2 invalid UTF-8, then it will report a match error if it comes across an invalid UTF-8 sequence. This is not good news for ripgrep, since it will stop it from searching the rest of the file, and will also print potentially undesirable error messages to users.

All right, the above is a lot of information to swallow if you aren't already familiar with ripgrep internals. Let's make this concrete with some examples. First, let's get some data big enough to magnify the performance differences:

$ curl -O 'https://burntsushi.net/stuff/subtitles2016-sample.gz'
$ gzip -d subtitles2016-sample
$ md5sum subtitles2016-sample
e3cb796a20bbc602fbfd6bb43bda45f5   subtitles2016-sample

To search this data, we will use the pattern ^\w{42}$, which contains exactly one hit in the file and has no literals. Having no literals is important, because it ensures that the regex engine won't use literal optimizations to speed up the search. In other words, it lets us reason coherently about the actual task that the regex engine is performing.

Let's now walk through a few examples in light of the information above. First, let's consider the default search using ripgrep's default regex engine and then the same search with PCRE2:

$ time rg '^\w{42}$' subtitles2016-sample
21225780:EverymajordevelopmentinthehistoryofAmerica

real    0m1.783s
user    0m1.731s
sys     0m0.051s

$ time rg -P '^\w{42}$' subtitles2016-sample
21225780:EverymajordevelopmentinthehistoryofAmerica

real    0m2.458s
user    0m2.419s
sys     0m0.038s

In this particular example, both pattern searches are using a Unicode aware \w character class and both are counting lines in order to report line numbers. The key difference here is that the first search will not search line by line, but the second one will. We can observe which strategy ripgrep uses by passing the --trace flag:

$ rg '^\w{42}$' subtitles2016-sample --trace
[... snip ...]
TRACE|grep_searcher::searcher|grep-searcher/src/searcher/mod.rs:622: Some("subtitles2016-sample"): searching via memory map
TRACE|grep_searcher::searcher|grep-searcher/src/searcher/mod.rs:712: slice reader: searching via slice-by-line strategy
TRACE|grep_searcher::searcher::core|grep-searcher/src/searcher/core.rs:61: searcher core: will use fast line searcher
[... snip ...]

$ rg -P '^\w{42}$' subtitles2016-sample --trace
[... snip ...]
TRACE|grep_searcher::searcher|grep-searcher/src/searcher/mod.rs:622: Some("subtitles2016-sample"): searching via memory map
TRACE|grep_searcher::searcher|grep-searcher/src/searcher/mod.rs:705: slice reader: needs transcoding, using generic reader
TRACE|grep_searcher::searcher|grep-searcher/src/searcher/mod.rs:685: generic reader: searching via roll buffer strategy
TRACE|grep_searcher::searcher::core|grep-searcher/src/searcher/core.rs:63: searcher core: will use slow line searcher
[... snip ...]

The first says it is using the "fast line searcher" where as the latter says it is using the "slow line searcher." The latter also shows that we are decoding the contents of the file, which also impacts performance.

Interestingly, in this case, the pattern does not match a \n and the file we're searching is valid UTF-8, so neither the slow line-by-line search strategy nor the decoding are necessary. We could fix the former issue with better PCRE2 introspection APIs. We can actually fix the latter issue with ripgrep's --no-encoding flag, which prevents the automatic UTF-8 decoding, but will enable PCRE2's own UTF-8 validity checking. Unfortunately, it's slower in my build of ripgrep:

$ time rg -P '^\w{42}$' subtitles2016-sample --no-encoding
21225780:EverymajordevelopmentinthehistoryofAmerica

real    0m3.074s
user    0m3.021s
sys     0m0.051s

(Tip: use the --trace flag to verify that no decoding in ripgrep is happening.)

A possible reason why PCRE2's UTF-8 checking is slower is because it might not be better than the highly optimized UTF-8 checking routines found in the encoding_rs library, which is what ripgrep uses for UTF-8 decoding. Moreover, my build of ripgrep enables encoding_rs's SIMD optimizations, which may be in play here.

Also, note that using the --no-encoding flag can cause PCRE2 to report invalid UTF-8 errors, which causes ripgrep to stop searching the file:

$ cat invalid-utf8
foobar

$ xxd invalid-utf8
00000000: 666f 6fff 6261 720a                      foo.bar.

$ rg foo invalid-utf8
1:foobar

$ rg -P foo invalid-utf8
1:foo�bar

$ rg -P foo invalid-utf8 --no-encoding
invalid-utf8: PCRE2: error matching: UTF-8 error: illegal byte (0xfe or 0xff)

All right, so at this point, you might think that we could remove the penalty for line-by-line searching by enabling multiline search. After all, our particular pattern can't match across multiple lines anyway, so we'll still get the results we want. Let's try it:

$ time rg -U '^\w{42}$' subtitles2016-sample
21225780:EverymajordevelopmentinthehistoryofAmerica

real    0m1.803s
user    0m1.748s
sys     0m0.054s

$ time rg -P -U '^\w{42}$' subtitles2016-sample
21225780:EverymajordevelopmentinthehistoryofAmerica

real    0m2.962s
user    0m2.246s
sys     0m0.713s

Search times remain the same with the default regex engine, but the PCRE2 search gets slower. What happened? The secrets can be revealed with the --trace flag once again. In the former case, ripgrep actually detects that the pattern can't match across multiple lines, and so will fall back to the "fast line search" strategy as with our search without -U.

However, for PCRE2, things are much worse. Namely, since Unicode mode is still enabled, ripgrep is still going to decode UTF-8 to ensure that it hands only valid UTF-8 to PCRE2. Unfortunately, one key downside of multiline search is that ripgrep cannot do it incrementally. Since matches can be arbitrarily long, ripgrep actually needs the entire file in memory at once. Normally, we can use a memory map for this, but because we need to UTF-8 decode the file before searching it, ripgrep winds up reading the entire contents of the file on to the heap before executing a search. Owch.

OK, so Unicode is killing us here. The file we're searching is mostly ASCII, so maybe we're OK with missing some data. (Try rg '[\w--\p{ascii}]' to see non-ASCII word characters that an ASCII-only \w character class would miss.) We can disable Unicode in both searches, but this is done differently depending on the regex engine we use:

$ time rg '(?-u)^\w{42}$' subtitles2016-sample
21225780:EverymajordevelopmentinthehistoryofAmerica

real    0m1.714s
user    0m1.669s
sys     0m0.044s

$ time rg -P '^\w{42}$' subtitles2016-sample --no-pcre2-unicode
21225780:EverymajordevelopmentinthehistoryofAmerica

real    0m1.997s
user    0m1.958s
sys     0m0.037s

For the most part, ripgrep's default regex engine performs about the same. PCRE2 does improve a little bit, and is now almost as fast as the default regex engine. If you look at the output of --trace, you'll see that ripgrep will no longer perform UTF-8 decoding, but it does still use the slow line-by-line searcher.

At this point, we can combine all of our insights above: let's try to get off of the slow line-by-line searcher by enabling multiline mode, and let's stop UTF-8 decoding by disabling Unicode support:

$ time rg -U '(?-u)^\w{42}$' subtitles2016-sample
21225780:EverymajordevelopmentinthehistoryofAmerica

real    0m1.714s
user    0m1.655s
sys     0m0.058s

$ time rg -P -U '^\w{42}$' subtitles2016-sample --no-pcre2-unicode
21225780:EverymajordevelopmentinthehistoryofAmerica

real    0m1.121s
user    0m1.071s
sys     0m0.048s

Ah, there's PCRE2's JIT shining! ripgrep's default regex engine once again remains about the same, but PCRE2 no longer needs to search line-by-line and it no longer needs to do any kind of UTF-8 checks. This allows the file to get memory mapped and passed right through PCRE2's JIT at impressive speeds. (As a brief and interesting historical note, the configuration of "memory map + multiline + no-Unicode" is exactly the configuration used by The Silver Searcher. This analysis perhaps sheds some reasoning as to why that configuration is useful!)

In summary, if you want PCRE2 to go as fast as possible and you don't care about Unicode and you don't care about matches possibly spanning across multiple lines, then enable multiline mode with -U and disable PCRE2's Unicode support with the --no-pcre2-unicode flag.

Caveat emptor: This author is not a PCRE2 expert, so there may be APIs that can improve performance that the author missed. Similarly, there may be alternative designs for a searching tool that are more amenable to how PCRE2 works.

When I run rg, why does it execute some other command?

It's likely that you have a shell alias or even another tool called rg which is interfering with ripgrep. Run which rg to see what it is.

(Notably, the Rails plug-in for Oh My Zsh sets up an rg alias for rails generate.)

Problems like this can be resolved in one of several ways:

  • If you're using the OMZ Rails plug-in, disable it by editing the plugins array in your zsh configuration.
  • Temporarily bypass an existing rg alias by calling ripgrep as command rg, \rg, or 'rg'.
  • Temporarily bypass an existing alias or another tool named rg by calling ripgrep by its full path (e.g., /usr/bin/rg or /usr/local/bin/rg).
  • Permanently disable an existing rg alias by adding unalias rg to the bottom of your shell configuration file (e.g., .bash_profile or .zshrc).
  • Give ripgrep its own alias that doesn't conflict with other tools/aliases by adding a line like the following to the bottom of your shell configuration file: alias ripgrep='command rg'.

How do I create an alias for ripgrep on Windows?

Often you can find a need to make alias for commands you use a lot that set certain flags. But PowerShell function aliases do not behave like your typical linux shell alias. You always need to propagate arguments and stdin input. But it cannot be done simply as function grep() { $input | rg.exe --hidden $args }

Use below example as reference to how setup alias in PowerShell.

function grep {
    $count = @($input).Count
    $input.Reset()

    if ($count) {
        $input | rg.exe --hidden $args
    }
    else {
        rg.exe --hidden $args
    }
}

PowerShell special variables:

  • input - is powershell stdin object that allows you to access its content.
  • args - is array of arguments passed to this function.

This alias checks whether there is stdin input and propagates only if there is some lines. Otherwise empty $input will make powershell to trigger rg to search empty stdin.

How do I create a PowerShell profile?

To customize powershell on start-up, there is a special PowerShell script that has to be created. In order to find its location, type $profile. See Microsoft's documentation for more details.

Any PowerShell code in this file gets evaluated at the start of console. This way you can have own aliases to be created at start.

How do I pipe non-ASCII content to ripgrep on Windows?

When piping input into native executables in PowerShell, the encoding of the input is controlled by the $OutputEncoding variable. By default, this is set to US-ASCII, and any characters in the pipeline that don't have encodings in US-ASCII are converted to ? (question mark) characters.

To change this setting, set $OutputEncoding to a different encoding, as represented by a .NET encoding object. Some common examples are below. The value of this variable is reset when PowerShell restarts, so to make this change take effect every time PowerShell is started add a line setting the variable into your PowerShell profile.

Example $OutputEncoding settings:

  • UTF-8 without BOM: $OutputEncoding = [System.Text.UTF8Encoding]::new()
  • The console's output encoding: $OutputEncoding = [System.Console]::OutputEncoding

If you continue to have encoding problems, you can also force the encoding that the console will use for printing to UTF-8 with [System.Console]::OutputEncoding = [System.Text.Encoding]::UTF8. This will also reset when PowerShell is restarted, so you can add that line to your profile as well if you want to make the setting permanent.

How can I search and replace with ripgrep?

Using ripgrep alone, you can't. ripgrep is a search tool that will never touch your files. However, the output of ripgrep can be piped to other tools that do modify files on disk. See this issue for more information.

sed is one such tool that can modify files on disk. sed can take a filename and a substitution command to search and replace in the specified file. Files containing matching patterns can be provided to sed using

rg foo --files-with-matches

The output of this command is a list of filenames that contain a match for the foo pattern.

This list can be piped into xargs, which will split the filenames from standard input into arguments for the command following xargs. You can use this combination to pipe a list of filenames into sed for replacement. For example:

rg foo --files-with-matches | xargs sed -i 's/foo/bar/g'

will replace all instances of 'foo' with 'bar' in the files in which ripgrep finds the foo pattern. The -i flag to sed indicates that you are editing files in place, and s/foo/bar/g says that you are performing a substitution of the pattern foo for bar, and that you are doing this substitution globally (all occurrences of the pattern in each file).

Note: the above command assumes that you are using GNU sed. If you are using BSD sed (the default on macOS and FreeBSD) then you must modify the above command to be the following:

rg foo --files-with-matches | xargs sed -i '' 's/foo/bar/g'

The -i flag in BSD sed requires a file extension to be given to make backups for all modified files. Specifying the empty string prevents file backups from being made.

Finally, if any of your file paths contain whitespace in them, then you might need to delimit your file paths with a NUL terminator. This requires telling ripgrep to output NUL bytes between each path, and telling xargs to read paths delimited by NUL bytes:

rg foo --files-with-matches -0 | xargs -0 sed -i 's/foo/bar/g'

To learn more about sed, see the sed manual here.

Additionally, Facebook has a tool called fastmod that uses some of the same libraries as ripgrep and might provide a more ergonomic search-and-replace experience.

How is ripgrep licensed?

ripgrep is dual licensed under the Unlicense and MIT licenses. Specifically, you may use ripgrep under the terms of either license.

The reason why ripgrep is dual licensed this way is two-fold:

  1. I, as ripgrep's author, would like to participate in a small bit of ideological activism by promoting the Unlicense's goal: to disclaim copyright monopoly interest.
  2. I, as ripgrep's author, would like as many people to use ripgrep as possible. Since the Unlicense is not a proven or well known license, ripgrep is also offered under the MIT license, which is ubiquitous and accepted by almost everyone.

More specifically, ripgrep and all its dependencies are compatible with this licensing choice. In particular, ripgrep's dependencies (direct and transitive) will always be limited to permissive licenses. That is, ripgrep will never depend on code that is not permissively licensed. This means rejecting any dependency that uses a copyleft license such as the GPL, LGPL, MPL or any of the Creative Commons ShareAlike licenses. Whether the license is "weak" copyleft or not does not matter; ripgrep will not depend on it.

Can ripgrep replace grep?

Yes and no.

If, upon hearing that "ripgrep can replace grep," you actually hear, "ripgrep can be used in every instance grep can be used, in exactly the same way, for the same use cases, with exactly the same bug-for-bug behavior," then no, ripgrep trivially cannot replace grep. Moreover, ripgrep will never replace grep.

If, upon hearing that "ripgrep can replace grep," you actually hear, "ripgrep can replace grep in some cases and not in other use cases," then yes, that is indeed true!

Let's go over some of those use cases in favor of ripgrep. Some of these may not apply to you. That's OK. There may be other use cases not listed here that do apply to you. That's OK too.

(For all claims related to performance in the following words, see my blog post introducing ripgrep.)

  • Are you frequently searching a repository of code? If so, ripgrep might be a good choice since there's likely a good chunk of your repository that you don't want to search. grep, can, of course, be made to filter files using recursive search, and if you don't mind writing out the requisite --exclude rules or writing wrapper scripts, then grep might be sufficient. (I'm not kidding, I myself did this with grep for almost a decade before writing ripgrep.) But if you instead enjoy having a search tool respect your .gitignore, then ripgrep might be perfect for you!
  • Are you frequently searching non-ASCII text that is UTF-8 encoded? One of ripgrep's key features is that it can handle Unicode features in your patterns in a way that tends to be faster than GNU grep. Unicode features in ripgrep are enabled by default; there is no need to configure your locale settings to use ripgrep properly because ripgrep doesn't respect your locale settings.
  • Do you need to search UTF-16 files and you don't want to bother explicitly transcoding them? Great. ripgrep does this for you automatically. No need to enable it.
  • Do you need to search a large directory of large files? ripgrep uses parallelism by default, which tends to make it faster than a standard grep -r search. However, if you're OK writing the occasional find ./ -print0 | xargs -P8 -0 grep command, then maybe grep is good enough.

Here are some cases where you might not want to use ripgrep. The same caveats for the previous section apply.

  • Are you writing portable shell scripts intended to work in a variety of environments? Great, probably not a good idea to use ripgrep! ripgrep has nowhere near the ubiquity of grep, so if you do use ripgrep, you might need to futz with the installation process more than you would with grep.
  • Do you care about POSIX compatibility? If so, then you can't use ripgrep because it never was, isn't and never will be POSIX compatible.
  • Do you hate tools that try to do something smart? If so, ripgrep is all about being smart, so you might prefer to just stick with grep.
  • Is there a particular feature of grep you rely on that ripgrep either doesn't have or never will have? If the former, file a bug report, maybe ripgrep can do it! If the latter, well, then, just use grep.

What does the "rip" in ripgrep mean?

When I first started writing ripgrep, I called it rep, intending it to be a shorter variant of grep. Soon after, I renamed it to xrep since rep wasn't obvious enough of a name for my taste. And also because adding x to anything always makes it better, right?

Before ripgrep's first public release, I decided that I didn't like xrep. I thought it was slightly awkward to type, and despite my previous praise of the letter x, I kind of thought it was pretty lame. Being someone who really likes Rust, I wanted to call it "rustgrep" or maybe "rgrep" for short. But I thought that was just as lame, and maybe a little too in-your-face. But I wanted to continue using r so I could at least pretend Rust had something to do with it.

I spent a couple of days trying to think of very short words that began with the letter r that were even somewhat related to the task of searching. I don't remember how it popped into my head, but "rip" came up as something that meant "fast," as in, "to rip through your text." The fact that RIP is also an initialism for "Rest in Peace" (as in, "ripgrep kills grep") never really dawned on me. Perhaps the coincidence is too striking to believe that, but I didn't realize it until someone explicitly pointed it out to me after the initial public release. I admit that I found it mildly amusing, but if I had realized it myself before the public release, I probably would have pressed on and chose a different name. Alas, renaming things after a release is hard, so I decided to mush on.

Given the fact that ripgrep never was, is or will be a 100% drop-in replacement for grep, ripgrep is neither actually a "grep killer" nor was it ever intended to be. It certainly does eat into some of its use cases, but that's nothing that other tools like ack or The Silver Searcher weren't already doing.

How can I donate to ripgrep or its maintainers?

I welcome sponsorship.

Or if you'd prefer, donating to a charitably organization that you like would also be most welcome. My favorites are: