twin/ripgrep
1
0
Fork 0
mirror of https://github.com/BurntSushi/ripgrep.git synced 2025-08-01 20:52:03 -07:00
Code Issues Packages Projects Releases Wiki Activity

Compare commits

base: twin:0.1.17
Branches Tags
twin:master twin:ag/rollup twin:ag/fix-ci twin:ag/index twin:ag/fix-multiline-replace-duplicate twin:ag/fix-cve-2021-3013 twin:ag/moarci twin:ag/ghactions twin:ag/snafu twin:ag/disable-bom-sniff twin:ag/partial-bstr-migration twin:ag/bstr-migration twin:ag/prepare-0.10.0 twin:ag/libripgrep-freeze-2 twin:ag/libripgrep-freeze-1 twin:ag/changelog
twin:globset-0.4.16 twin:14.1.1 twin:14.1.0 twin:ignore-0.4.22 twin:14.0.3 twin:14.0.2 twin:grep-0.3.1 twin:grep-printer-0.2.1 twin:grep-searcher-0.1.13 twin:14.0.1 twin:grep-0.3.0 twin:grep-printer-0.2.0 twin:14.0.0 twin:grep-0.2.13 twin:grep-searcher-0.1.12 twin:grep-pcre2-0.1.7 twin:grep-regex-0.1.12 twin:grep-matcher-0.1.7 twin:grep-cli-0.1.10 twin:ignore-0.4.21 twin:globset-0.4.14 twin:globset-0.4.13 twin:globset-0.4.12 twin:grep-cli-0.1.9 twin:globset-0.4.11 twin:grep-cli-0.1.8 twin:grep-0.2.12 twin:ignore-0.4.20 twin:grep-0.2.11 twin:grep-printer-0.1.7 twin:grep-searcher-0.1.11 twin:grep-pcre2-0.1.6 twin:grep-regex-0.1.11 twin:grep-matcher-0.1.6 twin:grep-cli-0.1.7 twin:ignore-0.4.19 twin:globset-0.4.10 twin:grep-0.2.10 twin:grep-searcher-0.1.10 twin:grep-0.2.9 twin:grep-searcher-0.1.9 twin:grep-regex-0.1.10 twin:globset-0.4.9 twin:globset-0.4.8 twin:13.0.0 twin:grep-0.2.8 twin:grep-printer-0.1.6 twin:grep-searcher-0.1.8 twin:grep-pcre2-0.1.5 twin:grep-regex-0.1.9 twin:grep-matcher-0.1.5 twin:grep-cli-0.1.6 twin:ignore-0.4.18 twin:globset-0.4.7 twin:ignore-0.4.17 twin:globset-0.4.6 twin:12.1.1 twin:grep-0.2.7 twin:grep-cli-0.1.5 twin:ignore-0.4.16 twin:12.1.0 twin:grep-0.2.6 twin:grep-printer-0.1.5 twin:grep-regex-0.1.8 twin:ignore-0.4.15 twin:12.0.1 twin:ignore-0.4.14 twin:ignore-0.4.13 twin:grep-regex-0.1.7 twin:12.0.0 twin:grep-searcher-0.1.7 twin:grep-regex-0.1.6 twin:grep-printer-0.1.4 twin:grep-pcre2-0.1.4 twin:grep-matcher-0.1.4 twin:ignore-0.4.12 twin:grep-0.2.5 twin:globset-0.4.5 twin:grep-cli-0.1.4 twin:ignore-0.4.11 twin:grep-regex-0.1.5 twin:grep-searcher-0.1.6 twin:ignore-0.4.10 twin:11.0.2 twin:grep-matcher-0.1.3 twin:grep-regex-0.1.4 twin:ignore-0.4.9 twin:ignore-0.4.8 twin:grep-searcher-0.1.5 twin:grep-printer-0.1.3 twin:grep-cli-0.1.3 twin:globset-0.4.4 twin:11.0.1 twin:11.0.0 twin:grep-0.2.4 twin:grep-cli-0.1.2 twin:grep-printer-0.1.2 twin:grep-searcher-0.1.4 twin:grep-regex-0.1.3 twin:grep-pcre2-0.1.3 twin:grep-matcher-0.1.2 twin:ignore-0.4.7 twin:globset-0.4.3 twin:grep-regex-0.1.2 twin:grep-searcher-0.1.3 twin:grep-searcher-0.1.2 twin:ignore-0.4.6 twin:ignore-0.4.5 twin:grep-0.2.3 twin:grep-pcre2-0.1.2 twin:0.10.0 twin:grep-searcher-0.1.1 twin:globset-0.4.2 twin:grep-pcre2-0.1.1 twin:grep-0.2.2 twin:grep-printer-0.1.1 twin:ignore-0.4.4 twin:grep-regex-0.1.1 twin:grep-matcher-0.1.1 twin:grep-cli-0.1.1 twin:grep-0.2.1 twin:grep-cli-0.1.0 twin:grep-0.2.0 twin:grep-search-0.1.0 twin:grep-matcher-0.1.0 twin:grep-printer-0.1.0 twin:grep-pcre2-0.1.0 twin:grep-regex-0.1.0 twin:0.9.0 twin:grep-0.1.9 twin:ignore-0.4.3 twin:globset-0.4.1 twin:ignore-0.4.2 twin:globset-0.4.0 twin:termcolor-0.3.6 twin:0.8.1 twin:termcolor-0.3.5 twin:ignore-0.4.1 twin:0.8.0 twin:ignore-0.4.0 twin:globset-0.3.0 twin:termcolor-0.3.4 twin:wincolor-0.1.6 twin:grep-0.1.8 twin:wincolor-0.1.5 twin:0.7.1 twin:ignore-0.3.1 twin:ignore-0.3.0 twin:grep-0.1.7 twin:globset-0.2.1 twin:0.7.0 twin:termcolor-0.3.3 twin:0.6.0 twin:ignore-0.2.2 twin:ignore-0.2.1 twin:wincolor-0.1.4 twin:0.5.2 twin:ignore-0.1.9 twin:0.5.1 twin:termcolor-0.3.2 twin:0.5.0 twin:grep-0.1.6 twin:ignore-0.1.8 twin:globset-0.1.4 twin:termcolor-0.3.1 twin:wincolor-0.1.3 twin:termcolor-0.3.0 twin:wincolor-0.1.2 twin:0.4.0 twin:0.3.2 twin:ignore-0.1.6 twin:0.3.1 twin:termcolor-0.1.1 twin:0.3.0 twin:termcolor-0.1.0 twin:wincolor-0.1.0 twin:0.2.9 twin:ignore-0.1.5 twin:0.2.8 twin:0.2.7 twin:ignore-0.1.4 twin:grep-0.1.4 twin:globset-0.1.2 twin:0.2.6 twin:ignore-0.1.3 twin:ignore-0.1.2 twin:0.2.5 twin:0.2.4 twin:ignore-0.1.1 twin:ignore-0.1.0 twin:globset-0.1.1 twin:0.2.3 twin:0.2.2 twin:globset-0.1.0 twin:0.2.1 twin:0.2.0 twin:0.1.18 twin:0.1.17 twin:0.1.16 twin:0.1.15 twin:0.1.14 twin:0.1.13 twin:0.1.12 twin:0.1.11 twin:0.1.10 twin:0.1.9 twin:0.1.8 twin:0.1.7 twin:0.1.6 twin:0.1.5 twin:0.1.4 twin:0.1.3 twin:0.1.2 twin:0.1.1 twin:0.1.0 twin:0.0.19 twin:0.0.18 twin:0.0.17 twin:0.0.16 twin:0.0.15 twin:0.0.14 twin:0.0.13 twin:0.0.12 twin:0.0.11 twin:0.0.10 twin:0.0.9 twin:0.0.8 twin:0.0.7 twin:0.0.6 twin:0.0.5 twin:0.0.4 twin:0.0.3 twin:0.0.2 twin:0.0.1
...
compare: twin:0.2.1
Branches Tags
twin:ag/rollup twin:master twin:ag/fix-ci twin:ag/index twin:ag/fix-multiline-replace-duplicate twin:ag/fix-cve-2021-3013 twin:ag/moarci twin:ag/ghactions twin:ag/snafu twin:ag/disable-bom-sniff twin:ag/partial-bstr-migration twin:ag/bstr-migration twin:ag/prepare-0.10.0 twin:ag/libripgrep-freeze-2 twin:ag/libripgrep-freeze-1 twin:ag/changelog
twin:globset-0.4.16 twin:14.1.1 twin:14.1.0 twin:ignore-0.4.22 twin:14.0.3 twin:14.0.2 twin:grep-0.3.1 twin:grep-printer-0.2.1 twin:grep-searcher-0.1.13 twin:14.0.1 twin:grep-0.3.0 twin:grep-printer-0.2.0 twin:14.0.0 twin:grep-0.2.13 twin:grep-searcher-0.1.12 twin:grep-pcre2-0.1.7 twin:grep-regex-0.1.12 twin:grep-matcher-0.1.7 twin:grep-cli-0.1.10 twin:ignore-0.4.21 twin:globset-0.4.14 twin:globset-0.4.13 twin:globset-0.4.12 twin:grep-cli-0.1.9 twin:globset-0.4.11 twin:grep-cli-0.1.8 twin:grep-0.2.12 twin:ignore-0.4.20 twin:grep-0.2.11 twin:grep-printer-0.1.7 twin:grep-searcher-0.1.11 twin:grep-pcre2-0.1.6 twin:grep-regex-0.1.11 twin:grep-matcher-0.1.6 twin:grep-cli-0.1.7 twin:ignore-0.4.19 twin:globset-0.4.10 twin:grep-0.2.10 twin:grep-searcher-0.1.10 twin:grep-0.2.9 twin:grep-searcher-0.1.9 twin:grep-regex-0.1.10 twin:globset-0.4.9 twin:globset-0.4.8 twin:13.0.0 twin:grep-0.2.8 twin:grep-printer-0.1.6 twin:grep-searcher-0.1.8 twin:grep-pcre2-0.1.5 twin:grep-regex-0.1.9 twin:grep-matcher-0.1.5 twin:grep-cli-0.1.6 twin:ignore-0.4.18 twin:globset-0.4.7 twin:ignore-0.4.17 twin:globset-0.4.6 twin:12.1.1 twin:grep-0.2.7 twin:grep-cli-0.1.5 twin:ignore-0.4.16 twin:12.1.0 twin:grep-0.2.6 twin:grep-printer-0.1.5 twin:grep-regex-0.1.8 twin:ignore-0.4.15 twin:12.0.1 twin:ignore-0.4.14 twin:ignore-0.4.13 twin:grep-regex-0.1.7 twin:12.0.0 twin:grep-searcher-0.1.7 twin:grep-regex-0.1.6 twin:grep-printer-0.1.4 twin:grep-pcre2-0.1.4 twin:grep-matcher-0.1.4 twin:ignore-0.4.12 twin:grep-0.2.5 twin:globset-0.4.5 twin:grep-cli-0.1.4 twin:ignore-0.4.11 twin:grep-regex-0.1.5 twin:grep-searcher-0.1.6 twin:ignore-0.4.10 twin:11.0.2 twin:grep-matcher-0.1.3 twin:grep-regex-0.1.4 twin:ignore-0.4.9 twin:ignore-0.4.8 twin:grep-searcher-0.1.5 twin:grep-printer-0.1.3 twin:grep-cli-0.1.3 twin:globset-0.4.4 twin:11.0.1 twin:11.0.0 twin:grep-0.2.4 twin:grep-cli-0.1.2 twin:grep-printer-0.1.2 twin:grep-searcher-0.1.4 twin:grep-regex-0.1.3 twin:grep-pcre2-0.1.3 twin:grep-matcher-0.1.2 twin:ignore-0.4.7 twin:globset-0.4.3 twin:grep-regex-0.1.2 twin:grep-searcher-0.1.3 twin:grep-searcher-0.1.2 twin:ignore-0.4.6 twin:ignore-0.4.5 twin:grep-0.2.3 twin:grep-pcre2-0.1.2 twin:0.10.0 twin:grep-searcher-0.1.1 twin:globset-0.4.2 twin:grep-pcre2-0.1.1 twin:grep-0.2.2 twin:grep-printer-0.1.1 twin:ignore-0.4.4 twin:grep-regex-0.1.1 twin:grep-matcher-0.1.1 twin:grep-cli-0.1.1 twin:grep-0.2.1 twin:grep-cli-0.1.0 twin:grep-0.2.0 twin:grep-search-0.1.0 twin:grep-matcher-0.1.0 twin:grep-printer-0.1.0 twin:grep-pcre2-0.1.0 twin:grep-regex-0.1.0 twin:0.9.0 twin:grep-0.1.9 twin:ignore-0.4.3 twin:globset-0.4.1 twin:ignore-0.4.2 twin:globset-0.4.0 twin:termcolor-0.3.6 twin:0.8.1 twin:termcolor-0.3.5 twin:ignore-0.4.1 twin:0.8.0 twin:ignore-0.4.0 twin:globset-0.3.0 twin:termcolor-0.3.4 twin:wincolor-0.1.6 twin:grep-0.1.8 twin:wincolor-0.1.5 twin:0.7.1 twin:ignore-0.3.1 twin:ignore-0.3.0 twin:grep-0.1.7 twin:globset-0.2.1 twin:0.7.0 twin:termcolor-0.3.3 twin:0.6.0 twin:ignore-0.2.2 twin:ignore-0.2.1 twin:wincolor-0.1.4 twin:0.5.2 twin:ignore-0.1.9 twin:0.5.1 twin:termcolor-0.3.2 twin:0.5.0 twin:grep-0.1.6 twin:ignore-0.1.8 twin:globset-0.1.4 twin:termcolor-0.3.1 twin:wincolor-0.1.3 twin:termcolor-0.3.0 twin:wincolor-0.1.2 twin:0.4.0 twin:0.3.2 twin:ignore-0.1.6 twin:0.3.1 twin:termcolor-0.1.1 twin:0.3.0 twin:termcolor-0.1.0 twin:wincolor-0.1.0 twin:0.2.9 twin:ignore-0.1.5 twin:0.2.8 twin:0.2.7 twin:ignore-0.1.4 twin:grep-0.1.4 twin:globset-0.1.2 twin:0.2.6 twin:ignore-0.1.3 twin:ignore-0.1.2 twin:0.2.5 twin:0.2.4 twin:ignore-0.1.1 twin:ignore-0.1.0 twin:globset-0.1.1 twin:0.2.3 twin:0.2.2 twin:globset-0.1.0 twin:0.2.1 twin:0.2.0 twin:0.1.18 twin:0.1.17 twin:0.1.16 twin:0.1.15 twin:0.1.14 twin:0.1.13 twin:0.1.12 twin:0.1.11 twin:0.1.10 twin:0.1.9 twin:0.1.8 twin:0.1.7 twin:0.1.6 twin:0.1.5 twin:0.1.4 twin:0.1.3 twin:0.1.2 twin:0.1.1 twin:0.1.0 twin:0.0.19 twin:0.0.18 twin:0.0.17 twin:0.0.16 twin:0.0.15 twin:0.0.14 twin:0.0.13 twin:0.0.12 twin:0.0.11 twin:0.0.10 twin:0.0.9 twin:0.0.8 twin:0.0.7 twin:0.0.6 twin:0.0.5 twin:0.0.4 twin:0.0.3 twin:0.0.2 twin:0.0.1

82 Commits
0.1.17 ... 0.2.1

Author SHA1 Message Date
Andrew Gallant
de79be2db2 0.2.1 2016-09-26 20:02:58 -04:00
Andrew Gallant
416b69bae5 changelog 0.2.1 2016-09-26 20:02:47 -04:00
Andrew Gallant
3e78fce3a3 Don't print empty lines in single threaded mode. 
Fixes #99.
 2016-09-26 19:57:23 -04:00
Andrew Gallant
7a3fd1f23f Add a --null flag. 
This flag causes a NUL byte to follow any file path in ripgrep's output.

Closes #89.
 2016-09-26 19:21:17 -04:00
Andrew Gallant
d306403440 Fix an off-by-one error with --column. 
Fixes #105.
 2016-09-26 19:09:59 -04:00
Andrew Gallant
ebabe1df6a Merge branch 'gitignore_blank_lines' 2016-09-26 18:55:43 -04:00
Andrew Gallant
f27aa3ff6f Add regression test. 
Fixes #106.
 2016-09-26 18:55:26 -04:00
Tom Jackson
20ccd441f2 Allow (and ignore) whitespace-only lines in .gitignore files 
Git considers these to be blank lines.
 2016-09-26 18:52:57 -04:00
Andrew Gallant
104d740f76 Don't quit if opening a file fails. 
This was already working correctly in multithreaded mode, but in single
threaded mode, a file failing to open caused search to stop. That's bad.

Fixes #98.
 2016-09-26 18:44:19 -04:00
Andrew Gallant
2da0eab2b8 Don't initialize ignores for file arguments. 
We'll never use them, so it's wasted effort.
 2016-09-26 18:44:19 -04:00
Andrew Gallant
b8c7864a02 Merge pull request #107 from kaushalmodi/add-systemverilog-type 
Add SystemVerilog (SV) type
 2016-09-26 16:54:49 -04:00
Kaushal Modi
ec26995655 Add SystemVerilog (SV) type 2016-09-26 15:24:35 -04:00
Andrew Gallant
a41235a3b5 Merge pull request #100 from emlyn/patch-1 
Recognise cljc and cljx extensions as Clojure(script)
 2016-09-26 08:59:59 -04:00
Emlyn Corrin
1a91b900e7 Clojure files can also end in cljc or cljx 
(see https://github.com/clojure/clojurescript/wiki/Using-cljc)
 2016-09-26 09:53:46 +01:00
Andrew Gallant
2b15832655 update brew formula to 0.2.0 2016-09-25 22:50:50 -04:00
Andrew Gallant
b1c52b52d6 0.2.0 2016-09-25 22:32:14 -04:00
Andrew Gallant
109bc3f78e bump grep to 0.1.3 2016-09-25 22:30:17 -04:00
Andrew Gallant
b62195b33f grep 0.1.3 2016-09-25 22:29:35 -04:00
Andrew Gallant
baebfd7add changelog 0.2.0 2016-09-25 22:27:58 -04:00
Andrew Gallant
19e405e5c5 fix windows 2016-09-25 21:48:01 -04:00
Andrew Gallant
f85822266f Don't use an intermediate buffer when --threads=1. 
Fixes #8
 2016-09-25 21:27:17 -04:00
Andrew Gallant
b034b77798 Don't replace NUL bytes when searching binary files as text. 
This was a result of misinterpreting a feature in grep where NUL bytes
are replaced with \n. The primary reason for doing this is to avoid
excessive memory usage on truly binary data. However, grep only does this
when searching binary files as if they were binary, and which only reports
whether the file matched or not. When grep is told to search binary data
as text (the -a/--text flag), then it doesn't do any replacement so we
shouldn't either.

In general, this makes sense, because the user is essentially asserting
that a particular file that looks like binary is actually text. In that
case, we shouldn't try to replace any NUL bytes.

ripgrep doesn't actually support searching binary data for whether it
matches or not, so we don't actually need the replace_buf function.
However, it does seem like a potentially useful feature.
 2016-09-25 21:26:49 -04:00
Andrew Gallant
278e1168bf Make printing paths a bit faster. 
It seems silly, but on *nix, we can just dump the bytes of the path
straight to the terminal. There's no need to do a UTF-8 check, which
can be costly when printing lots of matches.
 2016-09-25 21:23:26 -04:00
Andrew Gallant
6a8051b258 Don't union inner literals of repetitions. 
If we do, this results in extracting `foofoofoo` from `(\wfoo){3}`,
which is wrong. This does prevent us from extracting `foofoofoo` from
`foo{3}`, which is unfortunate, but we miss plenty of other stuff too.
Literal extracting needs a good rethink (all the way down into the regex
engine).

Fixes #93
 2016-09-25 20:10:28 -04:00
Andrew Gallant
a13ac3e3d4 On Windows, always consider stdin to be a tty. 
This means that `rg pat < file` won't do the expected thing and search
`fil`. Instead, it will recursively search the current directory for `pat`.
This isn't ideal, but is better than the previous behavior, which was to
wait for stdin when running `rg pat`, given the appearance of hanging
forever. The former is an important use case, but the latter is the
*central* use case of ripgrep, so we should make that work.

`rg` can still be used to search stdin on Windows, it just needs to be
done explicitly. e.g., `rg pat - < file` will search for `pat` in `file`.

Fixes #19
 2016-09-25 20:00:29 -04:00
Andrew Gallant
a72467996b Fix Windows compilation error. 2016-09-25 20:00:29 -04:00
Andrew Gallant
9395076468 Merge pull request #92 from svenstaro/patch-1 
ripgrep is now in [community]
 2016-09-25 18:50:03 -04:00
Sven-Hendrik Haase
a12c63957b ripgrep is now in [community] 
The README should reflect that.
 2016-09-26 00:48:41 +02:00
Andrew Gallant
982265af70 Move --files-with-matches to less common options. 2016-09-25 18:32:41 -04:00
Andrew Gallant
ed94aedf27 Permit whitelisting hidden files in ignores. 
Fixes #90
 2016-09-25 18:31:41 -04:00
Andrew Gallant
fd5ae2f795 Add curly brace alternates to glob format. 
Closes #80.
 2016-09-25 17:28:23 -04:00
Andrew Gallant
3d6a39be06 Fix tests on Windows. 
Mostly this is just using \\ instead of / in paths reported by the OS.
 2016-09-25 15:45:51 -04:00
Andrew Gallant
e7839f2200 Merge pull request #71 from catchmrbharath/issue46 
[Fixes #46] Use 1 less worker thread than number of threads
 2016-09-25 15:02:38 -04:00
Andrew Gallant
9dc5464c84 Stop after first match is found with --quiet. 
Fixes #77.
 2016-09-25 15:01:29 -04:00
Andrew Gallant
95edcd4d3a Merge pull request #42 from andschwa/files-with-matches 
Files with matches
 2016-09-25 14:53:31 -04:00
Andrew Gallant
d97f404970 Stupid docopt. 
It thinks `--type-clear is` is a flag spec.
 2016-09-25 14:47:35 -04:00
Andrew Gallant
b2bbd46178 Clarify documentation of --type-add. 
This explains it a bit more based on end user feedback. We also fix
the example, which was wrong.

Fixes #82.
 2016-09-25 14:37:01 -04:00
Andrew Gallant
82542df5cb Merge pull request #84 from martinlindhe/ts 
Add ts type for typescript
 2016-09-25 11:33:11 -04:00
Martin Lindhe
e4329037aa Add ts type for typescript 2016-09-25 17:16:15 +02:00
Andrew Gallant
ab0d1c1c79 Be more conservative with stdin. 
If no paths are given to ripgrep, only read from stdin if it's a file or
a FIFO. In particular, if something like `rg foo < /dev/null` is used,
then don't try to read from stdin.

Fixes #35, #81
 2016-09-25 11:14:54 -04:00
Andrew Gallant
2015c56e8d Merge pull request #62 from martinlindhe/js-wc 
--type js: include more extensions
 2016-09-25 11:10:47 -04:00
Martin Lindhe
23ad8b989d --type js: include more extensions 2016-09-25 17:06:13 +02:00
Andrew Schwartzmeyer
a8f3d9e87e Add --files-with-matches flag. 
Closes #26.

Acts like --count but emits only the paths of files with matches,
suitable for piping to xargs. Both mmap and no-mmap searches terminate
after the first match is found. Documentation updated and tests added.
 2016-09-24 21:40:17 -07:00
Bharath M R
9f1aae64f8 [Fixes #46] Use 1 less worker thread than number of threads 
The main thread does directory traversal. Hence
number of threads = main Thread + number of worker threads.
We should have atleast one worker thread.
 2016-09-24 19:48:26 -07:00
Andrew Gallant
1595f0faf5 Add --smart-case. 
It does what it says on the tin.

Closes #70.
 2016-09-24 21:51:04 -04:00
Andrew Gallant
8eeb0c0b60 Add --no-ignore-vcs flag. 
This flag will respect .ignore but not .gitignore.

Closes #68.
 2016-09-24 21:31:24 -04:00
Andrew Gallant
423f2a1927 Permit options with --help/--version. 
Fixes #47.
 2016-09-24 21:13:24 -04:00
Andrew Gallant
4b5e789a2a Strip trailing whitespace in gitignore patterns. 
Fixes #38.
 2016-09-24 20:56:24 -04:00
Andrew Gallant
37b731a048 Update brew package to 0.1.17. 
Closes #58, Fixes #13
 2016-09-24 20:51:07 -04:00
Andrew Gallant
a44735aa87 Tweak memory maps on darwin. 
Namely, don't automatically pick memory maps on darwin, ever. They
appear slower than standard read calls.

Closes #36.
 2016-09-24 20:48:05 -04:00
Andrew Gallant
6b2efd4d88 If a file is empty, still try to search it. 
Files like /proc/cpuinfo will advertise themselves as a normal file with
size 0. Normally, this isn't a problem, but if ripgrep decides to use a
memory map, it skipped searching if the file was empty since it's an error
to memory map an empty file. Instead of returning 0, we should just fall
back to standard read calls.

Fixes #55.
 2016-09-24 20:45:06 -04:00
Andrew Gallant
c8227e0cf3 Don't ignore first path when using --files. 
This is a docopt oddity, but probably not a bug. If --files is given,
then just interpret the pattern (if not empty) as the first file path.

Fixes #64.
 2016-09-24 20:22:02 -04:00
Andrew Gallant
b941c10b90 Fix directory whitelisting. 
There was a bug in the translation from a gitignore pattern to a standard
glob where `!/dir` wasn't being interpreted as an absolute path.

Fixes #67.
 2016-09-24 20:10:30 -04:00
Andrew Gallant
872a107658 Fix whitelisting precedence. 
Once a file is known to be whitelisted, we shouldn't check any ancestor
gitignores.
 2016-09-24 20:09:29 -04:00
Andrew Gallant
71ad9bf393 Fix trailing recursive globs in gitignore. 
A standard glob of `foo/**` will match `foo`, but gitignore semantics
specify that `foo/**` should only match the contents of `foo` and not
`foo` itself. We capture those semantics by translating `foo/**` to
`foo/**/*`.

Fixes #30.
 2016-09-24 19:44:06 -04:00
Andrew Gallant
f733e9ebe4 Fix typo. 
Thanks @dmit!
 2016-09-24 19:43:01 -04:00
Andrew Gallant
ce85df1d2e Clarify what rg does in --help. 
Fixes #24.
 2016-09-24 19:26:28 -04:00
Andrew Gallant
a6e3cab65a Add --no-filename flag. 
When this flag is set, a filename is never shown for a match.

Closes #20
 2016-09-24 19:24:24 -04:00
Andrew Gallant
7b860affbe Change the default output of --files to elide './'. 
This is kind of a ticky-tack change. I do think ./ as a prefix is
reasonable default, *but* we strip ./ when showing search results, so it
does make sense to be consistent.

Fixes #21.
 2016-09-24 19:18:48 -04:00
Andrew Gallant
af4dc78537 Update to docopt 0.6.85. 
The new version won't panic if printing to stdout fails.

Fixes #22.
 2016-09-24 19:14:19 -04:00
Andrew Gallant
9ce0484670 Clarify the documentation of the --type-* flags. 
Fixes #15
 2016-09-24 18:55:48 -04:00
Andrew Gallant
346bad7dfc Fix handling of absolute patterns in parent gitignore files. 
If a gitignore file in a *parent* directory is used, then it must be
matched relative to the directory it's in. ripgrep wasn't actually
adhering to this rule. Consider an example:

  .gitignore
  src
    llvm
      foo

Where `.gitignore` contains `/llvm/` and `foo` contains `test`. When
running `rg test` at the top-level directory, `foo` is correctly searched.
If you `cd` into `src` and re-run the same search, `foo` is ignored because
the `/llvm/` pattern is interpreted with respect to the current working
directory, which is wrong. The problem is that the path of `llvm` is
`./llvm`, which makes it look like it should match.

We fix this by rebuilding the directory path of each file when traversing
gitignores in parent directories. This does come with a small performance
hit.

Fixes #25.
 2016-09-24 18:40:50 -04:00
Andrew Gallant
56fe93d343 Fix an absolute path name bug. 
Namely, if a .gitignore inside a sub-directory has an absolute pattern,
e.g., `/foo/`, then we should match it relative to the directory containing
the .gitignore.
 2016-09-24 17:31:24 -04:00
Andrew Gallant
155676b474 Fixes #43. 2016-09-24 16:34:34 -04:00
Andrew Gallant
a3fc4cdded Fix a bug in the translation from a gitignore pattern to a glob. 
We were erroneously neglecting to prefix a pattern like `foo/`
with `**/` (to make `**/foo/`) because it had a slash in it. In fact, the
only reason to neglect a **/ prefix is if the pattern already starts
with **/, or if the pattern is absolute.

Fixes #16, #49, #50, #65
 2016-09-24 16:29:25 -04:00
Andrew Gallant
3bec8f3f0a Impl Debug for IgnoreDir. 2016-09-24 16:29:25 -04:00
Andrew Gallant
3b37f12ec0 Merge pull request #69 from dloss/nim-filetype 
Add support for the Nim programming language file type
 2016-09-24 16:04:58 -04:00
Dirk Loss
a2ed677e03 Add support for the Nim programming language file type 2016-09-24 21:48:33 +02:00
Andrew Gallant
2fb9c3c42c Merge pull request #56 from chrisdoc/feature/swift-file-type 
Add support for the Swift programming language file type
 2016-09-24 14:34:06 -04:00
Andrew Gallant
447e1ba0e2 Merge pull request #66 from kontomondo/master 
FSharp language file type
 2016-09-24 14:33:01 -04:00
Konto Mondo
3b45059212 FSharp language file type 2016-09-24 10:30:30 -04:00
chrisdoc
f74078af5b Add support for the Swift programming language file type 2016-09-24 08:42:44 +02:00
Andrew Gallant
5ff9b2f2a2 Merge pull request #41 from BurntSushi/generic-ignore 
Switch from .rgignore to .ignore.
 2016-09-23 23:14:38 -04:00
Andrew Gallant
cc90511ab2 Switch from .rgignore to .ignore. 
But don't actually remove support for .rgignore until the next semver
bump.

Note that this puts us in line with the silver searcher:
https://github.com/ggreer/the_silver_searcher/pull/974

Fixes #40
 2016-09-23 22:44:33 -04:00
Andrew Gallant
f5d60a80a8 Merge pull request #28 from ledge23/patch-1 
Add VB files to default type list
 2016-09-23 22:33:12 -04:00
Andrew Gallant
6fa158f6d3 Merge pull request #29 from jimhester/r_extensions 
Add a few more R relevant extensions
 2016-09-23 22:32:10 -04:00
Andrew Gallant
ef6dea40ff Merge pull request #39 from JohnVillalovos/master 
Prefer https:// over git://
 2016-09-23 22:28:20 -04:00
John L. Villalovos
9035c6b7b3 Prefer https:// over git:// 
1) git is not a secure protocol and vulnerable to man-in-the-middle
   attacks.
2) git:// is a pain for users behind proxy servers :(

Change-Id: I1901bebbaf8f64b23b070dee8732a6fb13cbdfdd
 2016-09-23 16:34:24 -07:00
Andrew Gallant
f5eb36baac Fixing VC++ wording and link. 
Kudos to @retep998
 2016-09-23 18:39:07 -04:00
Andrew Gallant
6367dd61ba Column numbers should start at 1. 
ripgrep was documented to do 1-based indexing, so this is a bug and not
a breaking change.

Fixes #18
 2016-09-23 17:11:09 -04:00
Jim Hester
98892de1c1 Add a few more R relevant extensions 2016-09-23 14:48:15 -04:00
Zack Schuster
273c14a45a Add VB files to default type list 
Use-case: While not a vogue technology, VB is still a common file type taught in many university settings and used in many commercial settings. Working with VB files out-of-the-box would provide a lot of value to `ripgrep` users.

Example: I'm working on converting a legacy app to a modern infrastructure. The legacy app mixes CS and VB files liberally, so I always need to check both. For portability, it would be nice to just be able to ask for `-tcs -tvb` without registering with `--type-add` first.

Tests: I didn't notice any coverage aimed at this part of the code, but if I'm mistaken I'll amend the PR.
 2016-09-23 11:44:53 -07:00
24 changed files with 1251 additions and 248 deletions
Expand all files Collapse all files

102
CHANGELOG.md Normal file
View File

@@ -0,0 +1,102 @@
0.2.1
=====
Feature enhancements:
* Added or improved file type filtering for Clojure and SystemVerilog.
* [FEATURE #89](https://github.com/BurntSushi/ripgrep/issues/89):
Add a --null flag that outputs a NUL byte after every file path.
Bug fixes:
* [BUG #98](https://github.com/BurntSushi/ripgrep/issues/98):
Fix a bug in single threaded mode when if opening a file failed, ripgrep
quit instead of continuing the search.
* [BUG #99](https://github.com/BurntSushi/ripgrep/issues/99):
Fix another bug in single threaded mode where empty lines were being printed
by mistake.
* [BUG #105](https://github.com/BurntSushi/ripgrep/issues/105):
Fix an off-by-one error with --column.
* [BUG #106](https://github.com/BurntSushi/ripgrep/issues/106):
Fix a bug where a whitespace only line in a gitignore file caused ripgrep
to panic (i.e., crash).
0.2.0
=====
Feature enhancements:
* Added or improved file type filtering for VB, R, F#, Swift, Nim, Javascript,
TypeScript
* [FEATURE #20](https://github.com/BurntSushi/ripgrep/issues/20):
Adds a --no-filename flag.
* [FEATURE #26](https://github.com/BurntSushi/ripgrep/issues/26):
Adds --files-with-matches flag. Like --count, but only prints file paths
and doesn't need to count every match.
* [FEATURE #40](https://github.com/BurntSushi/ripgrep/issues/40):
Switch from using `.rgignore` to `.ignore`. Note that `.rgignore` is
still supported, but deprecated.
* [FEATURE #68](https://github.com/BurntSushi/ripgrep/issues/68):
Add --no-ignore-vcs flag that ignores .gitignore but not .ignore.
* [FEATURE #70](https://github.com/BurntSushi/ripgrep/issues/70):
Add -S/--smart-case flag (but is disabled by default).
* [FEATURE #80](https://github.com/BurntSushi/ripgrep/issues/80):
Add support for `{foo,bar}` globs.
Many many bug fixes. Thanks every for reporting these and helping make
`ripgrep` better! (Note that I haven't captured every tracking issue here,
some were closed as duplicates.)
* [BUG #8](https://github.com/BurntSushi/ripgrep/issues/8):
Don't use an intermediate buffer when --threads=1. (Permits constant memory
usage.)
* [BUG #15](https://github.com/BurntSushi/ripgrep/issues/15):
Improves the documentation for --type-add.
* [BUG #16](https://github.com/BurntSushi/ripgrep/issues/16),
[BUG #49](https://github.com/BurntSushi/ripgrep/issues/49),
[BUG #50](https://github.com/BurntSushi/ripgrep/issues/50),
[BUG #65](https://github.com/BurntSushi/ripgrep/issues/65):
Some gitignore globs were being treated as anchored when they weren't.
* [BUG #18](https://github.com/BurntSushi/ripgrep/issues/18):
--vimgrep reported incorrect column number.
* [BUG #19](https://github.com/BurntSushi/ripgrep/issues/19):
ripgrep was hanging waiting on stdin in some Windows terminals. Note that
this introduced a new bug:
[#94](https://github.com/BurntSushi/ripgrep/issues/94).
* [BUG #21](https://github.com/BurntSushi/ripgrep/issues/21):
Removes leading `./` when printing file paths.
* [BUG #22](https://github.com/BurntSushi/ripgrep/issues/22):
Running `rg --help | echo` caused `rg` to panic.
* [BUG #24](https://github.com/BurntSushi/ripgrep/issues/22):
Clarify the central purpose of rg in its usage message.
* [BUG #25](https://github.com/BurntSushi/ripgrep/issues/25):
Anchored gitignore globs weren't applied in subdirectories correctly.
* [BUG #30](https://github.com/BurntSushi/ripgrep/issues/30):
Globs like `foo/**` should match contents of `foo`, but not `foo` itself.
* [BUG #35](https://github.com/BurntSushi/ripgrep/issues/35),
[BUG #81](https://github.com/BurntSushi/ripgrep/issues/81):
When automatically detecting stdin, only read if it's a file or a fifo.
i.e., ignore stdin in `rg foo < /dev/null`.
* [BUG #36](https://github.com/BurntSushi/ripgrep/issues/36):
Don't automatically pick memory maps on MacOS. Ever.
* [BUG #38](https://github.com/BurntSushi/ripgrep/issues/38):
Trailing whitespace in gitignore wasn't being ignored.
* [BUG #43](https://github.com/BurntSushi/ripgrep/issues/43):
--glob didn't work with directories.
* [BUG #46](https://github.com/BurntSushi/ripgrep/issues/46):
Use one fewer worker thread than what is provided on CLI.
* [BUG #47](https://github.com/BurntSushi/ripgrep/issues/47):
--help/--version now work even if other options are set.
* [BUG #55](https://github.com/BurntSushi/ripgrep/issues/55):
ripgrep was refusing to search /proc/cpuinfo. Fixed by disabling memory
maps for files with zero size.
* [BUG #64](https://github.com/BurntSushi/ripgrep/issues/64):
The first path given with --files set was ignored.
* [BUG #67](https://github.com/BurntSushi/ripgrep/issues/67):
Sometimes whitelist globs like `!/dir` weren't interpreted as anchored.
* [BUG #77](https://github.com/BurntSushi/ripgrep/issues/77):
When -q/--quiet flag was passed, ripgrep kept searching even after a match
was found.
* [BUG #90](https://github.com/BurntSushi/ripgrep/issues/90):
Permit whitelisting hidden files.
* [BUG #93](https://github.com/BurntSushi/ripgrep/issues/93):
ripgrep was extracting an erroneous inner literal from a repeated pattern.

12
Cargo.lock generated
View File

@@ -1,13 +1,13 @@
[root]
name = "ripgrep"
version = "0.1.17"
version = "0.2.1"
dependencies = [
"deque 0.3.1 (registry+https://github.com/rust-lang/crates.io-index)",
"docopt 0.6.84 (registry+https://github.com/rust-lang/crates.io-index)",
"docopt 0.6.85 (registry+https://github.com/rust-lang/crates.io-index)",
"env_logger 0.3.5 (registry+https://github.com/rust-lang/crates.io-index)",
"fnv 1.0.5 (registry+https://github.com/rust-lang/crates.io-index)",
"glob 0.2.11 (registry+https://github.com/rust-lang/crates.io-index)",
"grep 0.1.2",
"grep 0.1.3",
"kernel32-sys 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)",
"lazy_static 0.2.1 (registry+https://github.com/rust-lang/crates.io-index)",
"libc 0.2.16 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -40,7 +40,7 @@ dependencies = [
[[package]]
name = "docopt"
version = "0.6.84"
version = "0.6.85"
source = "registry+https://github.com/rust-lang/crates.io-index"
dependencies = [
"lazy_static 0.2.1 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -80,7 +80,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
[[package]]
name = "grep"
version = "0.1.2"
version = "0.1.3"
dependencies = [
"log 0.3.6 (registry+https://github.com/rust-lang/crates.io-index)",
"memchr 0.1.11 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -234,7 +234,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
[metadata]
"checksum aho-corasick 0.5.3 (registry+https://github.com/rust-lang/crates.io-index)" = "ca972c2ea5f742bfce5687b9aef75506a764f61d37f8f649047846a9686ddb66"
"checksum deque 0.3.1 (registry+https://github.com/rust-lang/crates.io-index)" = "1614659040e711785ed8ea24219140654da1729f3ec8a47a9719d041112fe7bf"
"checksum docopt 0.6.84 (registry+https://github.com/rust-lang/crates.io-index)" = "5cbd4e858d9719f44d8c22567afd9eb4a3d12a23f5be67bd3ddc98ffdcf31e5c"
"checksum docopt 0.6.85 (registry+https://github.com/rust-lang/crates.io-index)" = "1b88d783674021c5570e7238e17985b9b8c7141d90f33de49031b8d56e7f0bf9"
"checksum env_logger 0.3.5 (registry+https://github.com/rust-lang/crates.io-index)" = "15abd780e45b3ea4f76b4e9a26ff4843258dd8a3eed2775a0e7368c2e7936c2f"
"checksum fnv 1.0.5 (registry+https://github.com/rust-lang/crates.io-index)" = "6cc484842f1e2884faf56f529f960cc12ad8c71ce96cc7abba0a067c98fee344"
"checksum fs2 0.2.5 (registry+https://github.com/rust-lang/crates.io-index)" = "bcd414e5a1a979b931bb92f41b7a54106d3f6d2e6c253e9ce943b7cd468251ef"

4
Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "ripgrep"
version = "0.1.17" #:version
version = "0.2.1" #:version
authors = ["Andrew Gallant <jamslam@gmail.com>"]
description = """
Line oriented search tool using Rust's regex library. Combines the raw
@@ -27,7 +27,7 @@ deque = "0.3"
docopt = "0.6"
env_logger = "0.3"
fnv = "1.0"
grep = { version = "0.1.2", path = "grep" }
grep = { version = "0.1.3", path = "grep" }
lazy_static = "0.2"
libc = "0.2"
log = "0.3"

16
README.md
View File

@@ -94,8 +94,7 @@ Linux.](https://github.com/BurntSushi/ripgrep/releases) Linux binaries are
static executables. Windows binaries are available either as built with MinGW
(GNU) or with Microsoft Visual C++ (MSVC). When possible, prefer MSVC over GNU,
but you'll need to have the
[Microsoft Visual C++ Build
Tools](http://landinghub.visualstudio.com/visual-cpp-build-tools)
[Microsoft VC++ 2015 redistributable](https://www.microsoft.com/en-us/download/details.aspx?id=48145)
installed.
If you're a **Homebrew** user, then you can install it with a custom formula
@@ -106,11 +105,10 @@ directly):
$ brew install https://raw.githubusercontent.com/BurntSushi/ripgrep/master/pkg/brew/ripgrep.rb
```
If you're an **Archlinux** user, then you can install `ripgrep` from the
[`ripgrep` AUR package](https://aur.archlinux.org/packages/ripgrep/), e.g.,
If you're an **Arch Linux** user, then you can install `ripgrep` from the official repos:
```
$ yaourt -S ripgrep
$ pacman -Syu ripgrep
```
If you're a **Rust programmer**, `ripgrep` can be installed with `cargo`:
@@ -146,9 +144,9 @@ files, ignore hidden files and directories and skip binary files:
$ rg foobar
```
The above command also respects all `.rgignore` files, including in parent
directories. `.rgignore` files can be used when `.gitignore` files are
insufficient. In all cases, `.rgignore` patterns take precedence over
The above command also respects all `.ignore` files, including in parent
directories. `.ignore` files can be used when `.gitignore` files are
insufficient. In all cases, `.ignore` patterns take precedence over
`.gitignore`.
To ignore all ignore files, use `-u`. To additionally search hidden files
@@ -237,7 +235,7 @@ The syntax supported is
`ripgrep` compiles with Rust 1.9 (stable) or newer. Building is easy:
```
$ git clone git://github.com/BurntSushi/ripgrep
$ git clone https://github.com/BurntSushi/ripgrep
$ cd ripgrep
$ cargo build --release
$ ./target/release/rg --version

1
ci/script.sh
View File

@@ -16,6 +16,7 @@ disable_cross_doctests() {
}
run_test_suite() {
cargo clean --target $TARGET --verbose
cargo build --target $TARGET --verbose
cargo test --target $TARGET --verbose

53
doc/rg.1
View File

@@ -16,9 +16,9 @@ rg [\f[I]options\f[]] \-\-files [\f[I]<\f[]path\f[I]> ...\f[]]
.PP
rg [\f[I]options\f[]] \-\-type\-list
.PP
rg \-\-help
rg [\f[I]options\f[]] \-\-help
.PP
rg \-\-version
rg [\f[I]options\f[]] \-\-version
.SH DESCRIPTION
.PP
rg (ripgrep) combines the usability of The Silver Searcher (an ack
@@ -86,6 +86,7 @@ Suppress line numbers.
.TP
.B \-q, \-\-quiet
Do not print anything to stdout.
If a match is found in a file, stop searching that file.
.RS
.RE
.TP
@@ -169,12 +170,23 @@ Print each file that would be searched (but don\[aq]t search).
.RS
.RE
.TP
.B \-l, \-\-files\-with\-matches
Only show path of each file with matches.
.RS
.RE
.TP
.B \-H, \-\-with\-filename
Prefix each match with the file name that contains it.
This is the default when more than one file is searched.
.RS
.RE
.TP
.B \-\-no\-filename
Never show the filename for a match.
This is the default when one file is searched.
.RS
.RE
.TP
.B \-\-heading
Show the file name above clusters of matches from each file.
This is the default mode at a tty.
@@ -211,8 +223,8 @@ Never use memory maps, even when they might be faster.
.RE
.TP
.B \-\-no\-ignore
Don\[aq]t respect ignore files (.gitignore, .rgignore, etc.) This
implies \-\-no\-ignore\-parent.
Don\[aq]t respect ignore files (.gitignore, .ignore, etc.) This implies
\-\-no\-ignore\-parent.
.RS
.RE
.TP
@@ -221,11 +233,31 @@ Don\[aq]t respect ignore files in parent directories.
.RS
.RE
.TP
.B \-\-no\-ignore\-vcs
Don\[aq]t respect version control ignore files (e.g., .gitignore).
Note that .ignore files will continue to be respected.
.RS
.RE
.TP
.B \-\-null
Whenever a file name is printed, follow it with a NUL byte.
This includes printing filenames before matches, and when printing a
list of matching files such as with \-\-count, \-\-files\-with\-matches
and \-\-files.
.RS
.RE
.TP
.B \-p, \-\-pretty
Alias for \-\-color=always \-\-heading \-n.
.RS
.RE
.TP
.B \-S, \-\-smart\-case
Search case insensitively if the pattern is all lowercase.
Search case sensitively otherwise.
.RS
.RE
.TP
.B \-j, \-\-threads \f[I]ARG\f[]
The number of threads to use.
Defaults to the number of logical CPUs (capped at 6).
@@ -254,11 +286,20 @@ Show all supported file types and their associated globs.
.TP
.B \-\-type\-add \f[I]ARG\f[] ...
Add a new glob for a particular file type.
Example: \-\-type\-add html:\f[I]\&.html,\f[].htm
Only one glob can be added at a time.
Multiple \-\-type\-add flags can be provided.
Unless \-\-type\-clear is used, globs are added to any existing globs
inside of ripgrep.
Note that this must be passed to every invocation of rg.
.RS
.RE
.PP
Example: \f[C]\-\-type\-add\ html:*.html\f[]
.TP
.B \-\-type\-clear \f[I]TYPE\f[] ...
Clear the file type globs for TYPE.
Clear the file type globs previously defined for TYPE.
This only clears the default type definitions that are found inside of
ripgrep.
Note that this must be passed to every invocation of rg.
.RS
.RE

42
doc/rg.1.md
View File

@@ -12,9 +12,9 @@ rg [*options*] --files [*<*path*> ...*]
rg [*options*] --type-list
rg --help
rg [*options*] --help
rg --version
rg [*options*] --version
# DESCRIPTION
@@ -58,7 +58,8 @@ the raw speed of grep.
: Suppress line numbers.
-q, --quiet
: Do not print anything to stdout.
: Do not print anything to stdout. If a match is found in a file, stop
searching that file.
-r, --replace *ARG*
: Replace every match with the string given. Capture group indices (e.g., $5)
@@ -110,10 +111,17 @@ the raw speed of grep.
--files
: Print each file that would be searched (but don't search).
-l, --files-with-matches
: Only show path of each file with matches.
-H, --with-filename
: Prefix each match with the file name that contains it. This is the
default when more than one file is searched.
--no-filename
: Never show the filename for a match. This is the default when
one file is searched.
--heading
: Show the file name above clusters of matches from each file.
This is the default mode at a tty.
@@ -137,15 +145,29 @@ the raw speed of grep.
: Never use memory maps, even when they might be faster.
--no-ignore
: Don't respect ignore files (.gitignore, .rgignore, etc.)
: Don't respect ignore files (.gitignore, .ignore, etc.)
This implies --no-ignore-parent.
--no-ignore-parent
: Don't respect ignore files in parent directories.
--no-ignore-vcs
: Don't respect version control ignore files (e.g., .gitignore).
Note that .ignore files will continue to be respected.
--null
: Whenever a file name is printed, follow it with a NUL byte.
This includes printing filenames before matches, and when printing
a list of matching files such as with --count, --files-with-matches
and --files.
-p, --pretty
: Alias for --color=always --heading -n.
-S, --smart-case
: Search case insensitively if the pattern is all lowercase.
Search case sensitively otherwise.
-j, --threads *ARG*
: The number of threads to use. Defaults to the number of logical CPUs
(capped at 6). [default: 0]
@@ -164,8 +186,14 @@ the raw speed of grep.
: Show all supported file types and their associated globs.
--type-add *ARG* ...
: Add a new glob for a particular file type.
Example: --type-add html:*.html,*.htm
: Add a new glob for a particular file type. Only one glob can be added
at a time. Multiple --type-add flags can be provided. Unless --type-clear
is used, globs are added to any existing globs inside of ripgrep. Note that
this must be passed to every invocation of rg.
Example: `--type-add html:*.html`
--type-clear *TYPE* ...
: Clear the file type globs for TYPE.
: Clear the file type globs previously defined for TYPE. This only clears
the default type definitions that are found inside of ripgrep. Note
that this must be passed to every invocation of rg.

2
grep/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "grep"
version = "0.1.2" #:version
version = "0.1.3" #:version
authors = ["Andrew Gallant <jamslam@gmail.com>"]
description = """
Fast line oriented regex searching as a library.

11
grep/src/literals.rs
View File

@@ -8,7 +8,6 @@ Note that this implementation is incredibly suspicious. We need something more
principled.
*/
use std::cmp;
use std::iter;
use regex::bytes::Regex;
use syntax::{
@@ -181,8 +180,6 @@ fn repeat_range_literals<F: FnMut(&Expr, &mut Literals)>(
lits: &mut Literals,
mut f: F,
) {
use syntax::Expr::*;
if min == 0 {
// This is a bit conservative. If `max` is set, then we could
// treat this as a finite set of alternations. For now, we
@@ -190,8 +187,12 @@ fn repeat_range_literals<F: FnMut(&Expr, &mut Literals)>(
lits.cut();
} else {
let n = cmp::min(lits.limit_size(), min as usize);
let es = iter::repeat(e.clone()).take(n).collect();
f(&Concat(es), lits);
// We only extract literals from a single repetition, even though
// we could do more. e.g., `a{3}` will have `a` extracted instead of
// `aaa`. The reason is that inner literal extraction can't be unioned
// across repetitions. e.g., extracting `foofoofoo` from `(\w+foo){3}`
// is wrong.
f(e, lits);
if n < min as usize {
lits.cut();
}

36
grep/src/search.rs
View File

@@ -52,6 +52,7 @@ pub struct GrepBuilder {
#[derive(Clone, Debug)]
struct Options {
case_insensitive: bool,
case_smart: bool,
line_terminator: u8,
size_limit: usize,
dfa_size_limit: usize,
@@ -61,6 +62,7 @@ impl Default for Options {
fn default() -> Options {
Options {
case_insensitive: false,
case_smart: false,
line_terminator: b'\n',
size_limit: 10 * (1 << 20),
dfa_size_limit: 10 * (1 << 20),
@@ -98,6 +100,18 @@ impl GrepBuilder {
self
}
/// Whether to enable smart case search or not (disabled by default).
///
/// Smart case uses case insensitive search if the regex is contains all
/// lowercase literal characters. Otherwise, a case sensitive search is
/// used instead.
///
/// Enabling the case_insensitive flag overrides this.
pub fn case_smart(mut self, yes: bool) -> GrepBuilder {
self.opts.case_smart = yes;
self
}
/// Set the approximate size limit of the compiled regular expression.
///
/// This roughly corresponds to the number of bytes occupied by a
@@ -148,8 +162,11 @@ impl GrepBuilder {
/// Creates a new regex from the given expression with the current
/// configuration.
fn regex(&self, expr: &Expr) -> Result<Regex> {
let casei =
self.opts.case_insensitive
|| (self.opts.case_smart && !has_uppercase_literal(expr));
RegexBuilder::new(&expr.to_string())
.case_insensitive(self.opts.case_insensitive)
.case_insensitive(casei)
.multi_line(true)
.unicode(true)
.size_limit(self.opts.size_limit)
@@ -274,6 +291,23 @@ impl<'b, 's> Iterator for Iter<'b, 's> {
}
}
fn has_uppercase_literal(expr: &Expr) -> bool {
use syntax::Expr::*;
match *expr {
Literal { ref chars, casei } => {
casei || chars.iter().any(|c| c.is_uppercase())
}
LiteralBytes { ref bytes, casei } => {
casei || bytes.iter().any(|&b| b'A' <= b && b <= b'Z')
}
Group { ref e, .. } => has_uppercase_literal(e),
Repeat { ref e, .. } => has_uppercase_literal(e),
Concat(ref es) => es.iter().any(has_uppercase_literal),
Alternate(ref es) => es.iter().any(has_uppercase_literal),
_ => false,
}
}
#[cfg(test)]
mod tests {
#![allow(unused_imports)]

6
pkg/brew/ripgrep.rb
View File

@@ -1,15 +1,15 @@
require 'formula'
class Ripgrep < Formula
version '0.1.15'
version '0.2.0'
desc "Search tool like grep and The Silver Searcher."
homepage "https://github.com/BurntSushi/ripgrep"
if Hardware::CPU.is_64_bit?
url "https://github.com/BurntSushi/ripgrep/releases/download/#{version}/ripgrep-#{version}-x86_64-apple-darwin.tar.gz"
sha256 "fc138cd57b533bd65739f3f695322e483fe648736358d853ddb9bcd26d84fdc5"
sha256 "f55ef5dac04178bcae0d6c5ba2d09690d326e8c7c3f28e561025b04e1ab81d80"
else
url "https://github.com/BurntSushi/ripgrep/releases/download/#{version}/ripgrep-#{version}-i686-apple-darwin.tar.gz"
sha256 "3ce1f12e49a463bc9dd4cfe2537aa9989a0dc81f7aa6f959ee0d0d82b5f768cb"
sha256 "d901d55ccb48c19067f563d42652dfd8642bf50d28a40c0e2a4d3e866857a93b"
end
def install

118
src/args.rs
View File

@@ -39,10 +39,10 @@ Usage: rg [options] -e PATTERN ... [<path> ...]
rg [options] <pattern> [<path> ...]
rg [options] --files [<path> ...]
rg [options] --type-list
rg --help
rg --version
rg [options] --help
rg [options] --version
rg combines the usability of The Silver Searcher with the raw speed of grep.
rg recursively searches your current directory for a regex pattern.
Common options:
-a, --text Search binary files as if they were text.
@@ -65,7 +65,8 @@ Common options:
-n, --line-number Show line numbers (1-based). This is enabled
by default at a tty.
-N, --no-line-number Suppress line numbers.
-q, --quiet Do not print anything to stdout.
-q, --quiet Do not print anything to stdout. If a match is
found in a file, stop searching that file.
-r, --replace ARG Replace every match with the string given.
Capture group indices (e.g., $5) and names
(e.g., $foo) are supported.
@@ -110,10 +111,17 @@ Less common options:
--files
Print each file that would be searched (but don't search).
-l, --files-with-matches
Only show path of each file with matches.
-H, --with-filename
Prefix each match with the file name that contains it. This is the
default when more than one file is searched.
--no-filename
Never show the filename for a match. This is the default when
one file is searched.
--heading
Show the file name above clusters of matches from each file.
This is the default mode at a tty.
@@ -137,15 +145,29 @@ Less common options:
Never use memory maps, even when they might be faster.
--no-ignore
Don't respect ignore files (.gitignore, .rgignore, etc.)
Don't respect ignore files (.gitignore, .ignore, etc.)
This implies --no-ignore-parent.
--no-ignore-parent
Don't respect ignore files in parent directories.
--no-ignore-vcs
Don't respect version control ignore files (e.g., .gitignore).
Note that .ignore files will continue to be respected.
--null
Whenever a file name is printed, follow it with a NUL byte.
This includes printing filenames before matches, and when printing
a list of matching files such as with --count, --files-with-matches
and --files.
-p, --pretty
Alias for --color=always --heading -n.
-S, --smart-case
Search case insensitively if the pattern is all lowercase.
Search case sensitively otherwise.
-j, --threads ARG
The number of threads to use. Defaults to the number of logical CPUs
(capped at 6). [default: 0]
@@ -163,11 +185,17 @@ File type management options:
Show all supported file types and their associated globs.
--type-add ARG ...
Add a new glob for a particular file type.
Example: --type-add html:*.html,*.htm
Add a new glob for a particular file type. Only one glob can be added
at a time. Multiple type-add flags can be provided. Unless type-clear
is used, globs are added to any existing globs inside of ripgrep. Note
that this must be passed to every invocation of rg.
Example: `--type-add html:*.html`
--type-clear TYPE ...
Clear the file type globs for TYPE.
Clear the file type globs previously defined for TYPE. This only clears
the default type definitions that are found inside of ripgrep. Note
that this must be passed to every invocation of rg.
";
/// RawArgs are the args as they are parsed from Docopt. They aren't used
@@ -183,6 +211,7 @@ pub struct RawArgs {
flag_context: usize,
flag_context_separator: String,
flag_count: bool,
flag_files_with_matches: bool,
flag_debug: bool,
flag_files: bool,
flag_follow: bool,
@@ -197,12 +226,16 @@ pub struct RawArgs {
flag_no_heading: bool,
flag_no_ignore: bool,
flag_no_ignore_parent: bool,
flag_no_ignore_vcs: bool,
flag_no_line_number: bool,
flag_no_mmap: bool,
flag_no_filename: bool,
flag_null: bool,
flag_pretty: bool,
flag_quiet: bool,
flag_regexp: Vec<String>,
flag_replace: Option<String>,
flag_smart_case: bool,
flag_text: bool,
flag_threads: usize,
flag_type: Vec<String>,
@@ -227,6 +260,7 @@ pub struct Args {
column: bool,
context_separator: Vec<u8>,
count: bool,
files_with_matches: bool,
eol: u8,
files: bool,
follow: bool,
@@ -241,6 +275,8 @@ pub struct Args {
mmap: bool,
no_ignore: bool,
no_ignore_parent: bool,
no_ignore_vcs: bool,
null: bool,
quiet: bool,
replace: Option<Vec<u8>>,
text: bool,
@@ -259,7 +295,8 @@ impl RawArgs {
if self.arg_path.is_empty() {
if atty::on_stdin()
|| self.flag_files
|| self.flag_type_list {
|| self.flag_type_list
|| !atty::stdin_is_readable() {
vec![Path::new("./").to_path_buf()]
} else {
vec![Path::new("-").to_path_buf()]
@@ -283,6 +320,9 @@ impl RawArgs {
} else if cfg!(windows) {
// On Windows, memory maps appear faster than read calls. Neat.
true
} else if cfg!(darwin) {
// On Mac, memory maps appear to suck. Neat.
false
} else {
// If we're only searching a few paths and all of them are
// files, then memory maps are probably faster.
@@ -317,16 +357,20 @@ impl RawArgs {
self.flag_color == "always"
};
let eol = b'\n';
let mut with_filename = self.flag_with_filename;
if !with_filename {
with_filename = paths.len() > 1 || paths[0].is_dir();
}
with_filename = with_filename && !self.flag_no_filename;
let mut btypes = TypesBuilder::new();
btypes.add_defaults();
try!(self.add_types(&mut btypes));
let types = try!(btypes.build());
let grep = try!(
GrepBuilder::new(&pattern)
.case_smart(self.flag_smart_case)
.case_insensitive(self.flag_ignore_case)
.line_terminator(eol)
.build()
@@ -343,6 +387,7 @@ impl RawArgs {
column: self.flag_column,
context_separator: unescape(&self.flag_context_separator),
count: self.flag_count,
files_with_matches: self.flag_files_with_matches,
eol: eol,
files: self.flag_files,
follow: self.flag_follow,
@@ -359,6 +404,10 @@ impl RawArgs {
no_ignore_parent:
// --no-ignore implies --no-ignore-parent
self.flag_no_ignore_parent || no_ignore,
no_ignore_vcs:
// --no-ignore implies --no-ignore-vcs
self.flag_no_ignore_vcs || no_ignore,
null: self.flag_null,
quiet: self.flag_quiet,
replace: self.flag_replace.clone().map(|s| s.into_bytes()),
text: text,
@@ -451,7 +500,7 @@ impl Args {
}
}
}
let raw: RawArgs =
let mut raw: RawArgs =
Docopt::new(USAGE)
.and_then(|d| d.argv(argv).version(Some(version())).decode())
.unwrap_or_else(|e| e.exit());
@@ -466,6 +515,13 @@ impl Args {
errored!("failed to initialize logger: {}", err);
}
// *sigh*... If --files is given, then the first path ends up in
// pattern.
if raw.flag_files {
if !raw.arg_pattern.is_empty() {
raw.arg_path.insert(0, raw.arg_pattern.clone());
}
}
raw.to_args().map_err(From::from)
}
@@ -506,6 +562,7 @@ impl Args {
.heading(self.heading)
.line_per_match(self.line_per_match)
.quiet(self.quiet)
.null(self.null)
.with_filename(self.with_filename);
if let Some(ref rep) = self.replace {
p = p.replace(rep.clone());
@@ -517,14 +574,23 @@ impl Args {
/// to the writer given.
pub fn out(&self) -> Out {
let mut out = Out::new(self.color);
if self.heading && !self.count {
out = out.file_separator(b"".to_vec());
} else if self.before_context > 0 || self.after_context > 0 {
out = out.file_separator(self.context_separator.clone());
if let Some(filesep) = self.file_separator() {
out = out.file_separator(filesep);
}
out
}
/// Retrieve the configured file separator.
pub fn file_separator(&self) -> Option<Vec<u8>> {
if self.heading && !self.count && !self.files_with_matches {
Some(b"".to_vec())
} else if self.before_context > 0 || self.after_context > 0 {
Some(self.context_separator.clone())
} else {
None
}
}
/// Create a new buffer for use with searching.
#[cfg(not(windows))]
pub fn outbuf(&self) -> ColoredTerminal<term::TerminfoTerminal<Vec<u8>>> {
@@ -571,6 +637,7 @@ impl Args {
.after_context(self.after_context)
.before_context(self.before_context)
.count(self.count)
.files_with_matches(self.files_with_matches)
.eol(self.eol)
.line_number(self.line_number)
.invert_match(self.invert_match)
@@ -589,6 +656,7 @@ impl Args {
) -> BufferSearcher<'a, W> {
BufferSearcher::new(printer, grep, path, buf)
.count(self.count)
.files_with_matches(self.files_with_matches)
.eol(self.eol)
.line_number(self.line_number)
.invert_match(self.invert_match)
@@ -615,14 +683,20 @@ impl Args {
pub fn walker(&self, path: &Path) -> Result<walk::Iter> {
let wd = WalkDir::new(path).follow_links(self.follow);
let mut ig = Ignore::new();
ig.ignore_hidden(!self.hidden);
ig.no_ignore(self.no_ignore);
ig.add_types(self.types.clone());
if !self.no_ignore_parent {
try!(ig.push_parents(path));
}
if let Some(ref overrides) = self.glob_overrides {
ig.add_override(overrides.clone());
// Only register ignore rules if this is a directory. If it's a file,
// then it was explicitly given by the end user, so we always search
// it.
if path.is_dir() {
ig.ignore_hidden(!self.hidden);
ig.no_ignore(self.no_ignore);
ig.no_ignore_vcs(self.no_ignore_vcs);
ig.add_types(self.types.clone());
if !self.no_ignore_parent {
try!(ig.push_parents(path));
}
if let Some(ref overrides) = self.glob_overrides {
ig.add_override(overrides.clone());
}
}
Ok(walk::Iter::new(ig, wd))
}

44
src/atty.rs
View File

@@ -4,30 +4,58 @@ from (or to) a terminal. Windows and Unix do this differently, so implement
both here.
*/
#[cfg(unix)]
pub fn stdin_is_readable() -> bool {
use std::fs::File;
use std::os::unix::fs::FileTypeExt;
use std::os::unix::io::{FromRawFd, IntoRawFd};
use libc;
let file = unsafe { File::from_raw_fd(libc::STDIN_FILENO) };
let md = file.metadata();
let _ = file.into_raw_fd();
let ft = match md {
Err(_) => return false,
Ok(md) => md.file_type(),
};
ft.is_file() || ft.is_fifo()
}
#[cfg(windows)]
pub fn stdin_is_readable() -> bool {
// ???
true
}
/// Returns true if there is a tty on stdin.
#[cfg(unix)]
pub fn on_stdin() -> bool {
use libc;
0 < unsafe { libc::isatty(libc::STDIN_FILENO) }
}
/// Returns true if there is a tty on stdout.
#[cfg(unix)]
pub fn on_stdout() -> bool {
use libc;
0 < unsafe { libc::isatty(libc::STDOUT_FILENO) }
}
/// Returns true if there is a tty on stdin.
#[cfg(windows)]
pub fn on_stdin() -> bool {
use kernel32;
use winapi;
unsafe {
let fd = winapi::winbase::STD_INPUT_HANDLE;
let mut out = 0;
kernel32::GetConsoleMode(kernel32::GetStdHandle(fd), &mut out) != 0
}
// BUG: https://github.com/BurntSushi/ripgrep/issues/19
// It's not clear to me how to determine whether there is a tty on stdin.
// Checking GetConsoleMode(GetStdHandle(stdin)) != 0 appears to report
// that stdin is a pipe, even if it's not in a cygwin terminal, for
// example.
//
// To fix this, we just assume there is always a tty on stdin. If Windows
// users need to search stdin, they'll have to pass -. Ug.
true
}
/// Returns true if there is a tty on stdout.
#[cfg(windows)]
pub fn on_stdout() -> bool {
use kernel32;

81
src/gitignore.rs
View File

@@ -9,7 +9,7 @@ The motivation for this submodule is performance and portability:
2. We could shell out to a `git` sub-command like ls-files or status, but it
seems better to not rely on the existence of external programs for a search
tool. Besides, we need to implement this logic anyway to support things like
an .rgignore file.
an .ignore file.
The key implementation detail here is that a single gitignore file is compiled
into a single RegexSet, which can be used to report which globs match a
@@ -31,7 +31,7 @@ use std::path::{Path, PathBuf};
use regex;
use glob;
use pathutil::strip_prefix;
use pathutil::{is_file_name, strip_prefix};
/// Represents an error that can occur when parsing a gitignore file.
#[derive(Debug)]
@@ -115,7 +115,17 @@ impl Gitignore {
if let Some(p) = strip_prefix("./", path) {
path = p;
}
if let Some(p) = strip_prefix(&self.root, path) {
// Strip any common prefix between the candidate path and the root
// of the gitignore, to make sure we get relative matching right.
// BUT, a file name might not have any directory components to it,
// in which case, we don't want to accidentally strip any part of the
// file name.
if !is_file_name(path) {
if let Some(p) = strip_prefix(&self.root, path) {
path = p;
}
}
if let Some(p) = strip_prefix("/", path) {
path = p;
}
self.matched_stripped(path, is_dir)
@@ -225,9 +235,10 @@ impl GitignoreBuilder {
/// The path given should be the path at which the globs for this gitignore
/// file should be matched.
pub fn new<P: AsRef<Path>>(root: P) -> GitignoreBuilder {
let root = strip_prefix("./", root.as_ref()).unwrap_or(root.as_ref());
GitignoreBuilder {
builder: glob::SetBuilder::new(),
root: root.as_ref().to_path_buf(),
root: root.to_path_buf(),
patterns: vec![],
}
}
@@ -272,6 +283,12 @@ impl GitignoreBuilder {
from: P,
mut line: &str,
) -> Result<(), Error> {
if line.starts_with("#") {
return Ok(());
}
if !line.ends_with("\\ ") {
line = line.trim_right();
}
if line.is_empty() {
return Ok(());
}
@@ -284,24 +301,15 @@ impl GitignoreBuilder {
};
let mut opts = glob::MatchOptions::default();
let has_slash = line.chars().any(|c| c == '/');
// If the line starts with an escaped '!', then remove the escape.
// Otherwise, if it starts with an unescaped '!', then this is a
// whitelist pattern.
match line.chars().nth(0) {
Some('#') => return Ok(()),
Some('\\') => {
match line.chars().nth(1) {
Some('!') | Some('#') => {
line = &line[1..];
}
_ => {}
}
}
Some('!') => {
let is_absolute = line.chars().nth(0).unwrap() == '/';
if line.starts_with("\\!") || line.starts_with("\\#") {
line = &line[1..];
} else {
if line.starts_with("!") {
pat.whitelist = true;
line = &line[1..];
}
Some('/') => {
if line.starts_with("/") {
// `man gitignore` says that if a glob starts with a slash,
// then the glob can only match the beginning of a path
// (relative to the location of gitignore). We achieve this by
@@ -309,7 +317,6 @@ impl GitignoreBuilder {
opts.require_literal_separator = true;
line = &line[1..];
}
_ => {}
}
// If it ends with a slash, then this should only match directories,
// but the slash should otherwise not be used while globbing.
@@ -320,14 +327,25 @@ impl GitignoreBuilder {
}
}
// If there is a literal slash, then we note that so that globbing
// doesn't let wildcards match slashes. Otherwise, we need to let
// the pattern match anywhere, so we add a `**/` prefix to achieve
// that behavior.
// doesn't let wildcards match slashes.
pat.pat = line.to_string();
if has_slash {
opts.require_literal_separator = true;
} else {
pat.pat = format!("**/{}", pat.pat);
}
// If there was a leading slash, then this is a pattern that must
// match the entire path name. Otherwise, we should let it match
// anywhere, so use a **/ prefix.
if !is_absolute {
// ... but only if we don't already have a **/ prefix.
if !pat.pat.starts_with("**/") {
pat.pat = format!("**/{}", pat.pat);
}
}
// If the pattern ends with `/**`, then we should only match everything
// inside a directory, but not the directory itself. Standard globs
// will match the directory. So we add `/*` to force the issue.
if pat.pat.ends_with("/**") {
pat.pat = format!("{}/*", pat.pat);
}
try!(self.builder.add_with(&pat.pat, &opts));
self.patterns.push(pat);
@@ -393,10 +411,14 @@ mod tests {
ignored!(ig24, ROOT, "target", "grep/target");
ignored!(ig25, ROOT, "Cargo.lock", "./tabwriter-bin/Cargo.lock");
ignored!(ig26, ROOT, "/foo/bar/baz", "./foo/bar/baz");
ignored!(ig27, ROOT, "foo/", "xyz/foo", true);
ignored!(ig28, ROOT, "src/*.rs", "src/grep/src/main.rs");
ignored!(ig29, "./src", "/llvm/", "./src/llvm", true);
ignored!(ig30, ROOT, "node_modules/ ", "node_modules", true);
not_ignored!(ignot1, ROOT, "amonths", "months");
not_ignored!(ignot2, ROOT, "monthsa", "months");
not_ignored!(ignot3, ROOT, "src/*.rs", "src/grep/src/main.rs");
not_ignored!(ignot3, ROOT, "/src/*.rs", "src/grep/src/main.rs");
not_ignored!(ignot4, ROOT, "/*.c", "mozilla-sha1/sha1.c");
not_ignored!(ignot5, ROOT, "/src/*.rs", "src/grep/src/main.rs");
not_ignored!(ignot6, ROOT, "*.rs\n!src/main.rs", "src/main.rs");
@@ -406,4 +428,11 @@ mod tests {
not_ignored!(ignot10, ROOT, "**/foo/bar", "foo/src/bar");
not_ignored!(ignot11, ROOT, "#foo", "#foo");
not_ignored!(ignot12, ROOT, "\n\n\n", "foo");
not_ignored!(ignot13, ROOT, "foo/**", "foo", true);
// See: https://github.com/BurntSushi/ripgrep/issues/106
#[test]
fn regression_106() {
Gitignore::from_str("/", " ").unwrap();
}
}

185
src/glob.rs
View File

@@ -43,12 +43,19 @@ use regex::bytes::RegexSet;
use pathutil::file_name;
lazy_static! {
static ref FILE_SEPARATORS: String = regex::quote(r"/\");
}
/// Represents an error that can occur when parsing a glob pattern.
#[derive(Clone, Debug, Eq, PartialEq)]
pub enum Error {
InvalidRecursive,
UnclosedClass,
InvalidRange(char, char),
UnopenedAlternates,
UnclosedAlternates,
NestedAlternates,
}
impl StdError for Error {
@@ -63,6 +70,17 @@ impl StdError for Error {
Error::InvalidRange(_, _) => {
"invalid character range"
}
Error::UnopenedAlternates => {
"unopened alternate group; missing '{' \
(maybe escape '}' with '[}]'?)"
}
Error::UnclosedAlternates => {
"unclosed alternate group; missing '}' \
(maybe escape '{' with '[{]'?)"
}
Error::NestedAlternates => {
"nested alternate groups are not allowed"
}
}
}
}
@@ -70,7 +88,11 @@ impl StdError for Error {
impl fmt::Display for Error {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
match *self {
Error::InvalidRecursive | Error::UnclosedClass => {
Error::InvalidRecursive
| Error::UnclosedClass
| Error::UnopenedAlternates
| Error::UnclosedAlternates
| Error::NestedAlternates => {
write!(f, "{}", self.description())
}
Error::InvalidRange(s, e) => {
@@ -322,7 +344,7 @@ impl SetBuilder {
///
/// It cannot be used directly to match file paths, but it can be converted
/// to a regular expression string.
#[derive(Clone, Debug, Default)]
#[derive(Clone, Debug, Default, Eq, PartialEq)]
pub struct Pattern {
tokens: Vec<Token>,
}
@@ -350,6 +372,7 @@ enum Token {
negated: bool,
ranges: Vec<(char, char)>,
},
Alternates(Vec<Pattern>),
}
impl Pattern {
@@ -358,13 +381,19 @@ impl Pattern {
/// If the pattern is not a valid glob, then an error is returned.
pub fn new(pat: &str) -> Result<Pattern, Error> {
let mut p = Parser {
p: Pattern::default(),
stack: vec![Pattern::default()],
chars: pat.chars().peekable(),
prev: None,
cur: None,
};
try!(p.parse());
Ok(p.p)
if p.stack.is_empty() {
Err(Error::UnopenedAlternates)
} else if p.stack.len() > 1 {
Err(Error::UnclosedAlternates)
} else {
Ok(p.stack.pop().unwrap())
}
}
/// Returns an extension if this pattern exclusively matches it.
@@ -506,7 +535,6 @@ impl Pattern {
/// regular expression and will represent the matching semantics of this
/// glob pattern and the options given.
pub fn to_regex_with(&self, options: &MatchOptions) -> String {
let seps = regex::quote(r"/\");
let mut re = String::new();
re.push_str("(?-u)");
if options.case_insensitive {
@@ -520,7 +548,20 @@ impl Pattern {
re.push('$');
return re;
}
for tok in &self.tokens {
self.tokens_to_regex(options, &self.tokens, &mut re);
re.push('$');
re
}
fn tokens_to_regex(
&self,
options: &MatchOptions,
tokens: &[Token],
re: &mut String,
) {
let seps = &*FILE_SEPARATORS;
for tok in tokens {
match *tok {
Token::Literal(c) => {
re.push_str(&regex::quote(&c.to_string()));
@@ -566,15 +607,22 @@ impl Pattern {
}
re.push(']');
}
Token::Alternates(ref patterns) => {
let mut parts = vec![];
for pat in patterns {
let mut altre = String::new();
self.tokens_to_regex(options, &pat.tokens, &mut altre);
parts.push(altre);
}
re.push_str(&parts.join("|"));
}
}
}
re.push('$');
re
}
}
struct Parser<'a> {
p: Pattern,
stack: Vec<Pattern>,
chars: iter::Peekable<str::Chars<'a>>,
prev: Option<char>,
cur: Option<char>,
@@ -584,44 +632,101 @@ impl<'a> Parser<'a> {
fn parse(&mut self) -> Result<(), Error> {
while let Some(c) = self.bump() {
match c {
'?' => self.p.tokens.push(Token::Any),
'?' => try!(self.push_token(Token::Any)),
'*' => try!(self.parse_star()),
'[' => try!(self.parse_class()),
c => self.p.tokens.push(Token::Literal(c)),
'{' => try!(self.push_alternate()),
'}' => try!(self.pop_alternate()),
',' => try!(self.parse_comma()),
c => try!(self.push_token(Token::Literal(c))),
}
}
Ok(())
}
fn push_alternate(&mut self) -> Result<(), Error> {
if self.stack.len() > 1 {
return Err(Error::NestedAlternates);
}
Ok(self.stack.push(Pattern::default()))
}
fn pop_alternate(&mut self) -> Result<(), Error> {
let mut alts = vec![];
while self.stack.len() >= 2 {
alts.push(self.stack.pop().unwrap());
}
self.push_token(Token::Alternates(alts))
}
fn push_token(&mut self, tok: Token) -> Result<(), Error> {
match self.stack.last_mut() {
None => Err(Error::UnopenedAlternates),
Some(ref mut pat) => Ok(pat.tokens.push(tok)),
}
}
fn pop_token(&mut self) -> Result<Token, Error> {
match self.stack.last_mut() {
None => Err(Error::UnopenedAlternates),
Some(ref mut pat) => Ok(pat.tokens.pop().unwrap()),
}
}
fn have_tokens(&self) -> Result<bool, Error> {
match self.stack.last() {
None => Err(Error::UnopenedAlternates),
Some(ref pat) => Ok(!pat.tokens.is_empty()),
}
}
fn parse_comma(&mut self) -> Result<(), Error> {
// If we aren't inside a group alternation, then don't
// treat commas specially. Otherwise, we need to start
// a new alternate.
if self.stack.len() <= 1 {
self.push_token(Token::Literal(','))
} else {
Ok(self.stack.push(Pattern::default()))
}
}
fn parse_star(&mut self) -> Result<(), Error> {
let prev = self.prev;
if self.chars.peek() != Some(&'*') {
self.p.tokens.push(Token::ZeroOrMore);
try!(self.push_token(Token::ZeroOrMore));
return Ok(());
}
assert!(self.bump() == Some('*'));
if self.p.tokens.is_empty() {
self.p.tokens.push(Token::RecursivePrefix);
if !try!(self.have_tokens()) {
try!(self.push_token(Token::RecursivePrefix));
let next = self.bump();
if !next.is_none() && next != Some('/') {
return Err(Error::InvalidRecursive);
}
return Ok(());
}
self.p.tokens.pop().unwrap();
try!(self.pop_token());
if prev != Some('/') {
return Err(Error::InvalidRecursive);
if self.stack.len() <= 1
|| (prev != Some(',') && prev != Some('{')) {
return Err(Error::InvalidRecursive);
}
}
let next = self.bump();
if next.is_none() {
self.p.tokens.push(Token::RecursiveSuffix);
return Ok(());
match self.chars.peek() {
None => {
assert!(self.bump().is_none());
self.push_token(Token::RecursiveSuffix)
}
Some(&',') | Some(&'}') if self.stack.len() >= 2 => {
self.push_token(Token::RecursiveSuffix)
}
Some(&'/') => {
assert!(self.bump() == Some('/'));
self.push_token(Token::RecursiveZeroOrMore)
}
_ => Err(Error::InvalidRecursive),
}
if next != Some('/') {
return Err(Error::InvalidRecursive);
}
self.p.tokens.push(Token::RecursiveZeroOrMore);
Ok(())
}
fn parse_class(&mut self) -> Result<(), Error> {
@@ -691,11 +796,10 @@ impl<'a> Parser<'a> {
// it as a literal.
ranges.push(('-', '-'));
}
self.p.tokens.push(Token::Class {
self.push_token(Token::Class {
negated: negated,
ranges: ranges,
});
Ok(())
})
}
fn bump(&mut self) -> Option<char> {
@@ -934,16 +1038,6 @@ mod tests {
baseliteral!(lit6, "[ab]", false);
baseliteral!(lit7, "?", false);
/*
issuffix!(suf1, "", false);
issuffix!(suf2, "a", true);
issuffix!(suf3, "ab", true);
issuffix!(suf4, "*ab", true);
issuffix!(suf5, "*.ab", true);
issuffix!(suf6, "?.ab", true);
issuffix!(suf7, "ab*", false);
*/
matches!(match1, "a", "a");
matches!(match2, "a*b", "a_b");
matches!(match3, "a*b*c", "abc");
@@ -975,6 +1069,7 @@ mod tests {
matches!(matchrec20, "**/.*", "abc/.abc");
matches!(matchrec21, ".*/**", ".abc");
matches!(matchrec22, ".*/**", ".abc/abc");
matches!(matchnot23, "foo/**", "foo");
matches!(matchrange1, "a[0-9]b", "a0b");
matches!(matchrange2, "a[0-9]b", "a9b");
@@ -1004,6 +1099,20 @@ mod tests {
matches!(matchcasei3, "aBcDeFg", "ABCDEFG", CASEI);
matches!(matchcasei4, "aBcDeFg", "AbCdEfG", CASEI);
matches!(matchalt1, "a,b", "a,b");
matches!(matchalt2, ",", ",");
matches!(matchalt3, "{a,b}", "a");
matches!(matchalt4, "{a,b}", "b");
matches!(matchalt5, "{**/src/**,foo}", "abc/src/bar");
matches!(matchalt6, "{**/src/**,foo}", "foo");
matches!(matchalt7, "{[}],foo}", "}");
matches!(matchalt8, "{foo}", "foo");
matches!(matchalt9, "{}", "");
matches!(matchalt10, "{,}", "");
matches!(matchalt11, "{*.foo,*.bar,*.wat}", "test.foo");
matches!(matchalt12, "{*.foo,*.bar,*.wat}", "test.bar");
matches!(matchalt13, "{*.foo,*.bar,*.wat}", "test.wat");
matches!(matchslash1, "abc/def", "abc/def", SLASHLIT);
nmatches!(matchslash2, "abc?def", "abc/def", SLASHLIT);
nmatches!(matchslash2_win, "abc?def", "abc\\def", SLASHLIT);

130
src/ignore.rs
View File

@@ -5,7 +5,7 @@ whether a *single* file path should be searched or not.
In general, there are two ways to ignore a particular file:
1. Specify an ignore rule in some "global" configuration, such as a
$HOME/.rgignore or on the command line.
$HOME/.ignore or on the command line.
2. A specific ignore file (like .gitignore) found during directory traversal.
The `IgnoreDir` type handles ignore patterns for any one particular directory
@@ -14,16 +14,18 @@ of `IgnoreDir`s for use during directory traversal.
*/
use std::error::Error as StdError;
use std::ffi::OsString;
use std::fmt;
use std::io;
use std::path::{Path, PathBuf};
use gitignore::{self, Gitignore, GitignoreBuilder, Match, Pattern};
use pathutil::is_hidden;
use pathutil::{file_name, is_hidden};
use types::Types;
const IGNORE_NAMES: &'static [&'static str] = &[
".gitignore",
".ignore",
".rgignore",
];
@@ -77,7 +79,9 @@ impl From<gitignore::Error> for Error {
pub struct Ignore {
/// A stack of ignore patterns at each directory level of traversal.
/// A directory that contributes no ignore patterns is `None`.
stack: Vec<Option<IgnoreDir>>,
stack: Vec<IgnoreDir>,
/// A stack of parent directories above the root of the current search.
parent_stack: Vec<IgnoreDir>,
/// A set of override globs that are always checked first. A match (whether
/// it's whitelist or blacklist) trumps anything in stack.
overrides: Overrides,
@@ -85,9 +89,11 @@ pub struct Ignore {
types: Types,
/// Whether to ignore hidden files or not.
ignore_hidden: bool,
/// When true, don't look at .gitignore or .agignore files for ignore
/// When true, don't look at .gitignore or .ignore files for ignore
/// rules.
no_ignore: bool,
/// When true, don't look at .gitignore files for ignore rules.
no_ignore_vcs: bool,
}
impl Ignore {
@@ -95,10 +101,12 @@ impl Ignore {
pub fn new() -> Ignore {
Ignore {
stack: vec![],
parent_stack: vec![],
overrides: Overrides::new(None),
types: Types::empty(),
ignore_hidden: true,
no_ignore: false,
no_ignore_vcs: true,
}
}
@@ -114,6 +122,12 @@ impl Ignore {
self
}
/// When set, VCS ignore files are ignored.
pub fn no_ignore_vcs(&mut self, yes: bool) -> &mut Ignore {
self.no_ignore_vcs = yes;
self
}
/// Add a set of globs that overrides all other match logic.
pub fn add_override(&mut self, gi: Gitignore) -> &mut Ignore {
self.overrides = Overrides::new(Some(gi));
@@ -138,10 +152,13 @@ impl Ignore {
let mut path = &*path;
let mut saw_git = path.join(".git").is_dir();
let mut ignore_names = IGNORE_NAMES.to_vec();
if self.no_ignore_vcs {
ignore_names.retain(|&name| name != ".gitignore");
}
let mut ignore_dir_results = vec![];
while let Some(parent) = path.parent() {
if self.no_ignore {
ignore_dir_results.push(Ok(None));
ignore_dir_results.push(Ok(IgnoreDir::empty(parent)));
} else {
if saw_git {
ignore_names.retain(|&name| name != ".gitignore");
@@ -156,7 +173,7 @@ impl Ignore {
}
for ignore_dir_result in ignore_dir_results.into_iter().rev() {
try!(self.push_ignore_dir(ignore_dir_result));
self.parent_stack.push(try!(ignore_dir_result));
}
Ok(())
}
@@ -167,10 +184,13 @@ impl Ignore {
/// stack (and therefore should be popped).
pub fn push<P: AsRef<Path>>(&mut self, path: P) -> Result<(), Error> {
if self.no_ignore {
self.stack.push(None);
return Ok(());
self.stack.push(IgnoreDir::empty(path));
Ok(())
} else if self.no_ignore_vcs {
self.push_ignore_dir(IgnoreDir::without_vcs(path))
} else {
self.push_ignore_dir(IgnoreDir::new(path))
}
self.push_ignore_dir(IgnoreDir::new(path))
}
/// Pushes the result of building a directory matcher on to the stack.
@@ -178,7 +198,7 @@ impl Ignore {
/// If the result given contains an error, then it is returned.
pub fn push_ignore_dir(
&mut self,
result: Result<Option<IgnoreDir>, Error>,
result: Result<IgnoreDir, Error>,
) -> Result<(), Error> {
match result {
Ok(id) => {
@@ -187,7 +207,7 @@ impl Ignore {
}
Err(err) => {
// Don't leave the stack in an inconsistent state.
self.stack.push(None);
self.stack.push(IgnoreDir::empty("error"));
Err(err)
}
}
@@ -207,12 +227,9 @@ impl Ignore {
if let Some(is_ignored) = self.ignore_match(path, mat) {
return is_ignored;
}
if self.ignore_hidden && is_hidden(&path) {
debug!("{} ignored because it is hidden", path.display());
return true;
}
let mut whitelisted = false;
if !self.no_ignore {
for id in self.stack.iter().rev().filter_map(|id| id.as_ref()) {
for id in self.stack.iter().rev() {
let mat = id.matched(path, is_dir);
if let Some(is_ignored) = self.ignore_match(path, mat) {
if is_ignored {
@@ -220,13 +237,43 @@ impl Ignore {
}
// If this path is whitelisted by an ignore, then
// fallthrough and let the file type matcher have a say.
whitelisted = true;
break;
}
}
// If the file has been whitelisted, then we have to stop checking
// parent directories. The only thing that can override a whitelist
// at this point is a type filter.
if !whitelisted {
let mut path = path.to_path_buf();
for id in self.parent_stack.iter().rev() {
if let Some(ref dirname) = id.name {
path = Path::new(dirname).join(path);
}
let mat = id.matched(&*path, is_dir);
if let Some(is_ignored) = self.ignore_match(&*path, mat) {
if is_ignored {
return true;
}
// If this path is whitelisted by an ignore, then
// fallthrough and let the file type matcher have a
// say.
whitelisted = true;
break;
}
}
}
}
let mat = self.types.matched(path, is_dir);
if let Some(is_ignored) = self.ignore_match(path, mat) {
return is_ignored;
if is_ignored {
return true;
}
whitelisted = true;
}
if !whitelisted && self.ignore_hidden && is_hidden(&path) {
debug!("{} ignored because it is hidden", path.display());
return true;
}
false
}
@@ -256,9 +303,12 @@ impl Ignore {
/// IgnoreDir represents a set of ignore patterns retrieved from a single
/// directory.
#[derive(Debug)]
pub struct IgnoreDir {
/// The path to this directory as given.
path: PathBuf,
/// The directory name, if one exists.
name: Option<OsString>,
/// A single accumulation of glob patterns for this directory, matched
/// using gitignore semantics.
///
@@ -272,13 +322,27 @@ pub struct IgnoreDir {
impl IgnoreDir {
/// Create a new matcher for the given directory.
///
/// If no ignore glob patterns could be found in the directory then `None`
/// is returned.
pub fn new<P: AsRef<Path>>(path: P) -> Result<Option<IgnoreDir>, Error> {
pub fn new<P: AsRef<Path>>(path: P) -> Result<IgnoreDir, Error> {
IgnoreDir::with_ignore_names(path, IGNORE_NAMES.iter())
}
/// Create a new matcher for the given directory.
///
/// Don't respect VCS ignore files.
pub fn without_vcs<P: AsRef<Path>>(path: P) -> Result<IgnoreDir, Error> {
let names = IGNORE_NAMES.iter().filter(|name| **name != ".gitignore");
IgnoreDir::with_ignore_names(path, names)
}
/// Create a new IgnoreDir that never matches anything with the given path.
pub fn empty<P: AsRef<Path>>(path: P) -> IgnoreDir {
IgnoreDir {
path: path.as_ref().to_path_buf(),
name: file_name(path.as_ref()).map(|s| s.to_os_string()),
gi: None,
}
}
/// Create a new matcher for the given directory using only the ignore
/// patterns found in the file names given.
///
@@ -291,12 +355,9 @@ impl IgnoreDir {
pub fn with_ignore_names<P: AsRef<Path>, S, I>(
path: P,
names: I,
) -> Result<Option<IgnoreDir>, Error>
) -> Result<IgnoreDir, Error>
where P: AsRef<Path>, S: AsRef<str>, I: Iterator<Item=S> {
let mut id = IgnoreDir {
path: path.as_ref().to_path_buf(),
gi: None,
};
let mut id = IgnoreDir::empty(path);
let mut ok = false;
let mut builder = GitignoreBuilder::new(&id.path);
// The ordering here is important. Later globs have higher precedence.
@@ -304,11 +365,10 @@ impl IgnoreDir {
ok = builder.add_path(id.path.join(name.as_ref())).is_ok() || ok;
}
if !ok {
Ok(None)
} else {
id.gi = Some(try!(builder.build()));
Ok(Some(id))
return Ok(id);
}
id.gi = Some(try!(builder.build()));
Ok(id)
}
/// Returns true if and only if the given file path should be ignored
@@ -359,10 +419,6 @@ impl Overrides {
/// Match::None (and interpreting non-matches as ignored) unless is_dir
/// is true.
pub fn matched<P: AsRef<Path>>(&self, path: P, is_dir: bool) -> Match {
// File types don't apply to directories.
if is_dir {
return Match::None;
}
let path = path.as_ref();
self.gi.as_ref()
.map(|gi| {
@@ -394,6 +450,9 @@ mod tests {
let gi = builder.build().unwrap();
let id = IgnoreDir {
path: Path::new($root).to_path_buf(),
name: Path::new($root).file_name().map(|s| {
s.to_os_string()
}),
gi: Some(gi),
};
assert!(id.matched($path, false).is_ignored());
@@ -411,6 +470,9 @@ mod tests {
let gi = builder.build().unwrap();
let id = IgnoreDir {
path: Path::new($root).to_path_buf(),
name: Path::new($root).file_name().map(|s| {
s.to_os_string()
}),
gi: Some(gi),
};
assert!(!id.matched($path, false).is_ignored());

78
src/main.rs
View File

@@ -23,11 +23,12 @@ extern crate winapi;
use std::error::Error;
use std::fs::File;
use std::io;
use std::path::{Path, PathBuf};
use std::path::Path;
use std::process;
use std::result;
use std::sync::{Arc, Mutex};
use std::thread;
use std::cmp;
use deque::{Stealer, Stolen};
use grep::Grep;
@@ -87,14 +88,17 @@ fn main() {
fn run(args: Args) -> Result<u64> {
let args = Arc::new(args);
let paths = args.paths();
let threads = cmp::max(1, args.threads() - 1);
let isone =
paths.len() == 1 && (paths[0] == Path::new("-") || paths[0].is_file());
if args.files() {
return run_files(args.clone());
}
if args.type_list() {
return run_types(args.clone());
}
if paths.len() == 1 && (paths[0] == Path::new("-") || paths[0].is_file()) {
return run_one(args.clone(), &paths[0]);
if threads == 1 || isone {
return run_one_thread(args.clone());
}
let out = Arc::new(Mutex::new(args.out()));
@@ -102,7 +106,7 @@ fn run(args: Args) -> Result<u64> {
let workq = {
let (workq, stealer) = deque::new();
for _ in 0..args.threads() {
for _ in 0..threads {
let worker = MultiWorker {
chan_work: stealer.clone(),
out: out.clone(),
@@ -145,22 +149,52 @@ fn run(args: Args) -> Result<u64> {
Ok(match_count)
}
fn run_one(args: Arc<Args>, path: &Path) -> Result<u64> {
fn run_one_thread(args: Arc<Args>) -> Result<u64> {
let mut worker = Worker {
args: args.clone(),
inpbuf: args.input_buffer(),
grep: args.grep(),
match_count: 0,
};
let term = args.stdout();
let mut printer = args.printer(term);
let work =
if path == Path::new("-") {
WorkReady::Stdin
let paths = args.paths();
let mut term = args.stdout();
let mut paths_searched: u64 = 0;
for p in paths {
if p == Path::new("-") {
paths_searched += 1;
let mut printer = args.printer(&mut term);
if worker.match_count > 0 {
if let Some(sep) = args.file_separator() {
printer = printer.file_separator(sep);
}
}
worker.do_work(&mut printer, WorkReady::Stdin);
} else {
WorkReady::PathFile(path.to_path_buf(), try!(File::open(path)))
};
worker.do_work(&mut printer, work);
for ent in try!(args.walker(p)) {
paths_searched += 1;
let mut printer = args.printer(&mut term);
if worker.match_count > 0 {
if let Some(sep) = args.file_separator() {
printer = printer.file_separator(sep);
}
}
let file = match File::open(ent.path()) {
Ok(file) => file,
Err(err) => {
eprintln!("{}: {}", ent.path().display(), err);
continue;
}
};
worker.do_work(&mut printer, WorkReady::DirFile(ent, file));
}
}
}
if !paths.is_empty() && paths_searched == 0 {
eprintln!("No files were searched, which means ripgrep probably \
applied a filter you didn't expect. \
Try running again with --debug.");
}
Ok(worker.match_count)
}
@@ -202,7 +236,6 @@ enum Work {
enum WorkReady {
Stdin,
DirFile(DirEntry, File),
PathFile(PathBuf, File),
}
struct MultiWorker {
@@ -277,17 +310,6 @@ impl Worker {
self.search(printer, path, file)
}
}
WorkReady::PathFile(path, file) => {
let mut path = &*path;
if let Some(p) = strip_prefix("./", path) {
path = p;
}
if self.args.mmap() {
self.search_mmap(printer, path, &file)
} else {
self.search(printer, path, file)
}
}
};
match result {
Ok(count) => {
@@ -322,7 +344,11 @@ impl Worker {
) -> Result<u64> {
if try!(file.metadata()).len() == 0 {
// Opening a memory map with an empty file results in an error.
return Ok(0);
// However, this may not actually be an empty file! For example,
// /proc/cpuinfo reports itself as an empty file, but it can
// produce data when it's read from. Therefore, we fall back to
// regular read calls.
return self.search(printer, path, file);
}
let mmap = try!(Mmap::open(file, Protection::Read));
Ok(self.args.searcher_buffer(

59
src/out.rs
View File

@@ -48,8 +48,6 @@ impl Out {
/// If set, the separator is printed between matches from different files.
/// By default, no separator is printed.
///
/// If sep is empty, then no file separator is printed.
pub fn file_separator(mut self, sep: Vec<u8>) -> Out {
self.file_separator = Some(sep);
self
@@ -317,3 +315,60 @@ impl<T: Terminal + Send> term::Terminal for ColoredTerminal<T> {
}
}
}
impl<'a, T: Terminal + Send> term::Terminal for &'a mut ColoredTerminal<T> {
type Output = T::Output;
fn fg(&mut self, fg: term::color::Color) -> term::Result<()> {
(**self).fg(fg)
}
fn bg(&mut self, bg: term::color::Color) -> term::Result<()> {
(**self).bg(bg)
}
fn attr(&mut self, attr: term::Attr) -> term::Result<()> {
(**self).attr(attr)
}
fn supports_attr(&self, attr: term::Attr) -> bool {
(**self).supports_attr(attr)
}
fn reset(&mut self) -> term::Result<()> {
(**self).reset()
}
fn supports_reset(&self) -> bool {
(**self).supports_reset()
}
fn supports_color(&self) -> bool {
(**self).supports_color()
}
fn cursor_up(&mut self) -> term::Result<()> {
(**self).cursor_up()
}
fn delete_line(&mut self) -> term::Result<()> {
(**self).delete_line()
}
fn carriage_return(&mut self) -> term::Result<()> {
(**self).carriage_return()
}
fn get_ref(&self) -> &Self::Output {
(**self).get_ref()
}
fn get_mut(&mut self) -> &mut Self::Output {
(**self).get_mut()
}
fn into_inner(self) -> Self::Output {
// Good golly miss molly...
unimplemented!()
}
}

18
src/pathutil.rs
View File

@@ -98,3 +98,21 @@ pub fn is_hidden<P: AsRef<Path>>(path: P) -> bool {
false
}
}
/// Returns true if this file path is just a file name. i.e., Its parent is
/// the empty string.
#[cfg(unix)]
pub fn is_file_name<P: AsRef<Path>>(path: P) -> bool {
use std::os::unix::ffi::OsStrExt;
use memchr::memchr;
let path = path.as_ref().as_os_str().as_bytes();
memchr(b'/', path).is_none()
}
/// Returns true if this file path is just a file name. i.e., Its parent is
/// the empty string.
#[cfg(not(unix))]
pub fn is_file_name<P: AsRef<Path>>(path: P) -> bool {
path.as_ref().parent().map(|p| p.as_os_str().is_empty()).unwrap_or(false)
}

98
src/printer.rs
View File

@@ -4,6 +4,7 @@ use regex::bytes::Regex;
use term::{Attr, Terminal};
use term::color;
use pathutil::strip_prefix;
use types::FileTypeDef;
/// Printer encapsulates all output logic for searching.
@@ -24,6 +25,8 @@ pub struct Printer<W> {
/// printed via the match directly, but occasionally we need to insert them
/// ourselves (for example, to print a context separator).
eol: u8,
/// A file separator to show before any matches are printed.
file_separator: Option<Vec<u8>>,
/// Whether to show file name as a heading or not.
///
/// N.B. If with_filename is false, then this setting has no effect.
@@ -32,6 +35,9 @@ pub struct Printer<W> {
line_per_match: bool,
/// Whether to suppress all output.
quiet: bool,
/// Whether to print NUL bytes after a file path instead of new lines
/// or `:`.
null: bool,
/// A string to use as a replacement of each match in a matching line.
replace: Option<Vec<u8>>,
/// Whether to prefix each match with the corresponding file name.
@@ -47,9 +53,11 @@ impl<W: Terminal + Send> Printer<W> {
column: false,
context_separator: "--".to_string().into_bytes(),
eol: b'\n',
file_separator: None,
heading: false,
line_per_match: false,
quiet: false,
null: false,
replace: None,
with_filename: false,
}
@@ -74,6 +82,13 @@ impl<W: Terminal + Send> Printer<W> {
self
}
/// If set, the separator is printed before any matches. By default, no
/// separator is printed.
pub fn file_separator(mut self, sep: Vec<u8>) -> Printer<W> {
self.file_separator = Some(sep);
self
}
/// Whether to show file name as a heading or not.
///
/// N.B. If with_filename is false, then this setting has no effect.
@@ -88,6 +103,13 @@ impl<W: Terminal + Send> Printer<W> {
self
}
/// Whether to cause NUL bytes to follow file paths instead of other
/// visual separators (like `:`, `-` and `\n`).
pub fn null(mut self, yes: bool) -> Printer<W> {
self.null = yes;
self
}
/// When set, all output is suppressed.
pub fn quiet(mut self, yes: bool) -> Printer<W> {
self.quiet = yes;
@@ -115,6 +137,11 @@ impl<W: Terminal + Send> Printer<W> {
self.has_printed
}
/// Returns true if the printer has been configured to be quiet.
pub fn is_quiet(&self) -> bool {
self.quiet
}
/// Flushes the underlying writer and returns it.
pub fn into_inner(mut self) -> W {
let _ = self.wtr.flush();
@@ -138,15 +165,24 @@ impl<W: Terminal + Send> Printer<W> {
/// Prints the given path.
pub fn path<P: AsRef<Path>>(&mut self, path: P) {
self.write(path.as_ref().to_string_lossy().as_bytes());
self.write_eol();
let path = strip_prefix("./", path.as_ref()).unwrap_or(path.as_ref());
self.write_path(path);
if self.null {
self.write(b"\x00");
} else {
self.write_eol();
}
}
/// Prints the given path and a count of the number of matches found.
pub fn path_count<P: AsRef<Path>>(&mut self, path: P, count: u64) {
if self.with_filename {
self.write(path.as_ref().to_string_lossy().as_bytes());
self.write(b":");
self.write_path(path);
if self.null {
self.write(b"\x00");
} else {
self.write(b":");
}
}
self.write(count.to_string().as_bytes());
self.write_eol();
@@ -179,7 +215,7 @@ impl<W: Terminal + Send> Printer<W> {
let column =
if self.column {
Some(re.find(&buf[start..end])
.map(|(s, _)| s + 1).unwrap_or(0) as u64)
.map(|(s, _)| s).unwrap_or(0) as u64)
} else {
None
};
@@ -204,16 +240,21 @@ impl<W: Terminal + Send> Printer<W> {
column: Option<u64>,
) {
if self.heading && self.with_filename && !self.has_printed {
self.write_file_sep();
self.write_heading(path.as_ref());
} else if !self.heading && self.with_filename {
self.write(path.as_ref().to_string_lossy().as_bytes());
self.write(b":");
self.write_path(path.as_ref());
if self.null {
self.write(b"\x00");
} else {
self.write(b":");
}
}
if let Some(line_number) = line_number {
self.line_number(line_number, b':');
}
if let Some(c) = column {
self.write(c.to_string().as_bytes());
self.write((c + 1).to_string().as_bytes());
self.write(b":");
}
if self.replace.is_some() {
@@ -254,10 +295,15 @@ impl<W: Terminal + Send> Printer<W> {
line_number: Option<u64>,
) {
if self.heading && self.with_filename && !self.has_printed {
self.write_file_sep();
self.write_heading(path.as_ref());
} else if !self.heading && self.with_filename {
self.write(path.as_ref().to_string_lossy().as_bytes());
self.write(b"-");
self.write_path(path.as_ref());
if self.null {
self.write(b"\x00");
} else {
self.write(b"-");
}
}
if let Some(line_number) = line_number {
self.line_number(line_number, b'-');
@@ -273,8 +319,12 @@ impl<W: Terminal + Send> Printer<W> {
let _ = self.wtr.fg(color::BRIGHT_GREEN);
let _ = self.wtr.attr(Attr::Bold);
}
self.write(path.as_ref().to_string_lossy().as_bytes());
self.write_eol();
self.write_path(path.as_ref());
if self.null {
self.write(b"\x00");
} else {
self.write_eol();
}
if self.wtr.supports_color() {
let _ = self.wtr.reset();
}
@@ -292,6 +342,19 @@ impl<W: Terminal + Send> Printer<W> {
self.write(&[sep]);
}
#[cfg(unix)]
fn write_path<P: AsRef<Path>>(&mut self, path: P) {
use std::os::unix::ffi::OsStrExt;
let path = path.as_ref().as_os_str().as_bytes();
self.write(path);
}
#[cfg(not(unix))]
fn write_path<P: AsRef<Path>>(&mut self, path: P) {
self.write(path.as_ref().to_string_lossy().as_bytes());
}
fn write(&mut self, buf: &[u8]) {
if self.quiet {
return;
@@ -304,4 +367,15 @@ impl<W: Terminal + Send> Printer<W> {
let eol = self.eol;
self.write(&[eol]);
}
fn write_file_sep(&mut self) {
if self.quiet {
return;
}
if let Some(ref sep) = self.file_separator {
self.has_printed = true;
let _ = self.wtr.write_all(sep);
let _ = self.wtr.write_all(b"\n");
}
}
}

24
src/search_buffer.rs
View File

@@ -53,6 +53,14 @@ impl<'a, W: Send + Terminal> BufferSearcher<'a, W> {
self
}
/// If enabled, searching will print the path instead of each match.
///
/// Disabled by default.
pub fn files_with_matches(mut self, yes: bool) -> Self {
self.opts.files_with_matches = yes;
self
}
/// Set the end-of-line byte used by this searcher.
pub fn eol(mut self, eol: u8) -> Self {
self.opts.eol = eol;
@@ -96,6 +104,9 @@ impl<'a, W: Send + Terminal> BufferSearcher<'a, W> {
self.print_match(m.start(), m.end());
}
last_end = m.end();
if self.printer.is_quiet() || self.opts.files_with_matches {
break;
}
}
if self.opts.invert_match {
let upto = self.buf.len();
@@ -104,13 +115,16 @@ impl<'a, W: Send + Terminal> BufferSearcher<'a, W> {
if self.opts.count && self.match_count > 0 {
self.printer.path_count(self.path, self.match_count);
}
if self.opts.files_with_matches && self.match_count > 0 {
self.printer.path(self.path);
}
self.match_count
}
#[inline(always)]
pub fn print_match(&mut self, start: usize, end: usize) {
self.match_count += 1;
if self.opts.count {
if self.opts.skip_matches() {
return;
}
self.count_lines(start);
@@ -237,6 +251,14 @@ and exhibited clearly, with a label attached.\
assert_eq!(out, "/baz.rs:2\n");
}
#[test]
fn files_with_matches() {
let (count, out) = search(
"Sherlock", SHERLOCK, |s| s.files_with_matches(true));
assert_eq!(1, count);
assert_eq!(out, "/baz.rs\n");
}
#[test]
fn invert_match() {
let (count, out) = search(

58
src/search_stream.rs
View File

@@ -80,6 +80,7 @@ pub struct Options {
pub after_context: usize,
pub before_context: usize,
pub count: bool,
pub files_with_matches: bool,
pub eol: u8,
pub invert_match: bool,
pub line_number: bool,
@@ -92,12 +93,22 @@ impl Default for Options {
after_context: 0,
before_context: 0,
count: false,
files_with_matches: false,
eol: b'\n',
invert_match: false,
line_number: false,
text: false,
}
}
}
impl Options {
/// Both --count and --files-with-matches options imply that we should not
/// display matches at all.
pub fn skip_matches(&self) -> bool {
return self.count || self.files_with_matches;
}
}
impl<'a, R: io::Read, W: Terminal + Send> Searcher<'a, R, W> {
@@ -158,6 +169,14 @@ impl<'a, R: io::Read, W: Terminal + Send> Searcher<'a, R, W> {
self
}
/// If enabled, searching will print the path instead of each match.
///
/// Disabled by default.
pub fn files_with_matches(mut self, yes: bool) -> Self {
self.opts.files_with_matches = yes;
self
}
/// Set the end-of-line byte used by this searcher.
pub fn eol(mut self, eol: u8) -> Self {
self.opts.eol = eol;
@@ -193,7 +212,7 @@ impl<'a, R: io::Read, W: Terminal + Send> Searcher<'a, R, W> {
self.line_count = if self.opts.line_number { Some(0) } else { None };
self.last_match = Match::default();
self.after_context_remaining = 0;
loop {
while !self.terminate() {
let upto = self.inp.lastnl;
self.print_after_context(upto);
if !try!(self.fill()) {
@@ -202,7 +221,7 @@ impl<'a, R: io::Read, W: Terminal + Send> Searcher<'a, R, W> {
if !self.opts.text && self.inp.is_binary {
break;
}
while self.inp.pos < self.inp.lastnl {
while !self.terminate() && self.inp.pos < self.inp.lastnl {
let matched = self.grep.read_match(
&mut self.last_match,
&mut self.inp.buf[..self.inp.lastnl],
@@ -234,12 +253,22 @@ impl<'a, R: io::Read, W: Terminal + Send> Searcher<'a, R, W> {
}
}
}
if self.opts.count && self.match_count > 0 {
self.printer.path_count(self.path, self.match_count);
if self.match_count > 0 {
if self.opts.count {
self.printer.path_count(self.path, self.match_count);
} else if self.opts.files_with_matches {
self.printer.path(self.path);
}
}
Ok(self.match_count)
}
#[inline(always)]
fn terminate(&self) -> bool {
self.match_count > 0
&& (self.printer.is_quiet() || self.opts.files_with_matches)
}
#[inline(always)]
fn fill(&mut self) -> Result<bool, Error> {
let mut keep = self.inp.lastnl;
@@ -281,7 +310,7 @@ impl<'a, R: io::Read, W: Terminal + Send> Searcher<'a, R, W> {
#[inline(always)]
fn print_before_context(&mut self, upto: usize) {
if self.opts.count || self.opts.before_context == 0 {
if self.opts.skip_matches() || self.opts.before_context == 0 {
return;
}
let start = self.last_printed;
@@ -304,7 +333,7 @@ impl<'a, R: io::Read, W: Terminal + Send> Searcher<'a, R, W> {
#[inline(always)]
fn print_after_context(&mut self, upto: usize) {
if self.opts.count || self.after_context_remaining == 0 {
if self.opts.skip_matches() || self.after_context_remaining == 0 {
return;
}
let start = self.last_printed;
@@ -322,7 +351,7 @@ impl<'a, R: io::Read, W: Terminal + Send> Searcher<'a, R, W> {
#[inline(always)]
fn print_match(&mut self, start: usize, end: usize) {
self.match_count += 1;
if self.opts.count {
if self.opts.skip_matches() {
return;
}
self.print_separator(start);
@@ -503,10 +532,6 @@ impl InputBuffer {
if self.first && is_binary(&self.buf[self.end..self.end + n]) {
self.is_binary = true;
}
if self.is_binary {
replace_buf(
&mut self.buf[self.end..self.end + n], b'\x00', self.eol);
}
self.first = false;
// We assume that reading 0 bytes means we've hit EOF.
if n == 0 {
@@ -629,6 +654,7 @@ pub fn count_lines(buf: &[u8], eol: u8) -> u64 {
}
/// Replaces a with b in buf.
#[allow(dead_code)]
fn replace_buf(buf: &mut [u8], a: u8, b: u8) {
if a == b {
return;
@@ -970,7 +996,7 @@ fn main() {
let text = "Sherlock\n\x00Holmes\n";
let (count, out) = search("Sherlock|Holmes", text, |s| s.text(true));
assert_eq!(2, count);
assert_eq!(out, "/baz.rs:Sherlock\n/baz.rs:Holmes\n");
assert_eq!(out, "/baz.rs:Sherlock\n/baz.rs:\x00Holmes\n");
}
#[test]
@@ -992,6 +1018,14 @@ fn main() {
assert_eq!(out, "/baz.rs:2\n");
}
#[test]
fn files_with_matches() {
let (count, out) = search_smallcap(
"Sherlock", SHERLOCK, |s| s.files_with_matches(true));
assert_eq!(1, count);
assert_eq!(out, "/baz.rs\n");
}
#[test]
fn invert_match() {
let (count, out) = search_smallcap(

14
src/types.rs
View File

@@ -18,7 +18,7 @@ const TYPE_EXTENSIONS: &'static [(&'static str, &'static [&'static str])] = &[
("awk", &["*.awk"]),
("c", &["*.c", "*.h", "*.H"]),
("cbor", &["*.cbor"]),
("clojure", &["*.clj", "*.cljs"]),
("clojure", &["*.clj", "*.cljc", "*.cljs", "*.cljx"]),
("cmake", &["CMakeLists.txt"]),
("coffeescript", &["*.coffee"]),
("cpp", &[
@@ -36,12 +36,15 @@ const TYPE_EXTENSIONS: &'static [(&'static str, &'static [&'static str])] = &[
"*.f", "*.F", "*.f77", "*.F77", "*.pfo",
"*.f90", "*.F90", "*.f95", "*.F95",
]),
("fsharp", &["*.fs", "*.fsx", "*.fsi"]),
("go", &["*.go"]),
("groovy", &["*.groovy"]),
("haskell", &["*.hs", "*.lhs"]),
("html", &["*.htm", "*.html"]),
("java", &["*.java"]),
("js", &["*.js"]),
("js", &[
"*.js", "*.jsx", "*.vue",
]),
("json", &["*.json"]),
("jsonl", &["*.jsonl"]),
("lisp", &["*.el", "*.jl", "*.lisp", "*.lsp", "*.sc", "*.scm"]),
@@ -52,6 +55,7 @@ const TYPE_EXTENSIONS: &'static [(&'static str, &'static [&'static str])] = &[
("matlab", &["*.m"]),
("mk", &["mkfile"]),
("ml", &["*.ml"]),
("nim", &["*.nim"]),
("objc", &["*.h", "*.m"]),
("objcpp", &["*.h", "*.mm"]),
("ocaml", &["*.ml", "*.mli", "*.mll", "*.mly"]),
@@ -59,17 +63,21 @@ const TYPE_EXTENSIONS: &'static [(&'static str, &'static [&'static str])] = &[
("php", &["*.php", "*.php3", "*.php4", "*.php5", "*.phtml"]),
("py", &["*.py"]),
("readme", &["README*", "*README"]),
("rr", &["*.R"]),
("r", &["*.R", "*.r", "*.Rmd", "*.Rnw"]),
("rst", &["*.rst"]),
("ruby", &["*.rb"]),
("rust", &["*.rs"]),
("scala", &["*.scala"]),
("sh", &["*.bash", "*.csh", "*.ksh", "*.sh", "*.tcsh"]),
("sql", &["*.sql"]),
("sv", &["*.v", "*.vg", "*.sv", "*.svh", "*.h"]),
("swift", &["*.swift"]),
("tex", &["*.tex", "*.cls", "*.sty"]),
("ts", &["*.ts", "*.tsx"]),
("txt", &["*.txt"]),
("toml", &["*.toml", "Cargo.lock"]),
("vala", &["*.vala"]),
("vb", &["*.vb"]),
("vimscript", &["*.vim"]),
("xml", &["*.xml"]),
("yacc", &["*.y"]),

307
tests/tests.rs
View File

@@ -34,6 +34,33 @@ macro_rules! sherlock {
};
}
macro_rules! clean {
($name:ident, $query:expr, $path:expr, $fun:expr) => {
#[test]
fn $name() {
let wd = WorkDir::new(stringify!($name));
let mut cmd = wd.command();
cmd.arg($query).arg($path);
$fun(wd, cmd);
}
};
}
fn path(unix: &str) -> String {
if cfg!(windows) {
unix.replace("/", "\\")
} else {
unix.to_string()
}
}
fn sort_lines(lines: &str) -> String {
let mut lines: Vec<String> =
lines.trim().lines().map(|s| s.to_owned()).collect();
lines.sort();
format!("{}\n", lines.join("\n"))
}
sherlock!(single_file, |wd: WorkDir, mut cmd| {
let lines: String = wd.stdout(&mut cmd);
let expected = "\
@@ -118,7 +145,11 @@ be, to a very large extent, the result of luck. Sherlock Holmes
foo
Sherlock Holmes lives on Baker Street.
";
assert!(lines == expected1 || lines == expected2);
if lines != expected1 {
assert_eq!(lines, expected2);
} else {
assert_eq!(lines, expected1);
}
});
sherlock!(inverted, |wd: WorkDir, mut cmd: Command| {
@@ -280,6 +311,20 @@ sherlock!(glob_negate, "Sherlock", ".", |wd: WorkDir, mut cmd: Command| {
assert_eq!(lines, "file.py:Sherlock\n");
});
sherlock!(count, "Sherlock", ".", |wd: WorkDir, mut cmd: Command| {
cmd.arg("--count");
let lines: String = wd.stdout(&mut cmd);
let expected = "sherlock:2\n";
assert_eq!(lines, expected);
});
sherlock!(files_with_matches, "Sherlock", ".", |wd: WorkDir, mut cmd: Command| {
cmd.arg("--files-with-matches");
let lines: String = wd.stdout(&mut cmd);
let expected = "sherlock\n";
assert_eq!(lines, expected);
});
sherlock!(after_context, |wd: WorkDir, mut cmd: Command| {
cmd.arg("-A").arg("1");
let lines: String = wd.stdout(&mut cmd);
@@ -377,6 +422,11 @@ sherlock!(ignore_git, "Sherlock", ".", |wd: WorkDir, mut cmd: Command| {
wd.assert_err(&mut cmd);
});
sherlock!(ignore_generic, "Sherlock", ".", |wd: WorkDir, mut cmd: Command| {
wd.create(".ignore", "sherlock\n");
wd.assert_err(&mut cmd);
});
sherlock!(ignore_ripgrep, "Sherlock", ".", |wd: WorkDir, mut cmd: Command| {
wd.create(".rgignore", "sherlock\n");
wd.assert_err(&mut cmd);
@@ -510,19 +560,11 @@ sherlock!(symlink_follow, "Sherlock", ".", |wd: WorkDir, mut cmd: Command| {
cmd.current_dir(wd.path().join("foo/bar"));
let lines: String = wd.stdout(&mut cmd);
if cfg!(windows) {
let expected = "\
baz\\sherlock:For the Doctor Watsons of this world, as opposed to the Sherlock
baz\\sherlock:be, to a very large extent, the result of luck. Sherlock Holmes
";
assert_eq!(lines, expected);
} else {
let expected = "\
let expected = "\
baz/sherlock:For the Doctor Watsons of this world, as opposed to the Sherlock
baz/sherlock:be, to a very large extent, the result of luck. Sherlock Holmes
";
assert_eq!(lines, expected);
}
assert_eq!(lines, path(expected));
});
sherlock!(unrestricted1, "Sherlock", ".", |wd: WorkDir, mut cmd: Command| {
@@ -556,7 +598,7 @@ sherlock!(unrestricted3, "foo", ".", |wd: WorkDir, mut cmd: Command| {
cmd.arg("-uuu");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, "file:foo\nfile:foo\n");
assert_eq!(lines, "file:foo\x00bar\nfile:foo\x00baz\n");
});
// On Windows, this test uses memory maps, so the NUL bytes don't get replaced.
@@ -574,10 +616,217 @@ sherlock!(vimgrep, "Sherlock|Watson", ".", |wd: WorkDir, mut cmd: Command| {
let lines: String = wd.stdout(&mut cmd);
let expected = "\
sherlock:1:15:For the Doctor Watsons of this world, as opposed to the Sherlock
sherlock:1:56:For the Doctor Watsons of this world, as opposed to the Sherlock
sherlock:3:48:be, to a very large extent, the result of luck. Sherlock Holmes
sherlock:5:11:but Doctor Watson has to have it taken out for him and dusted,
sherlock:1:16:For the Doctor Watsons of this world, as opposed to the Sherlock
sherlock:1:57:For the Doctor Watsons of this world, as opposed to the Sherlock
sherlock:3:49:be, to a very large extent, the result of luck. Sherlock Holmes
sherlock:5:12:but Doctor Watson has to have it taken out for him and dusted,
";
assert_eq!(lines, expected);
});
// See: https://github.com/BurntSushi/ripgrep/issues/16
clean!(regression_16, "xyz", ".", |wd: WorkDir, mut cmd: Command| {
wd.create(".gitignore", "ghi/");
wd.create_dir("ghi");
wd.create_dir("def/ghi");
wd.create("ghi/toplevel.txt", "xyz");
wd.create("def/ghi/subdir.txt", "xyz");
wd.assert_err(&mut cmd);
});
// See: https://github.com/BurntSushi/ripgrep/issues/25
clean!(regression_25, "test", ".", |wd: WorkDir, mut cmd: Command| {
wd.create(".gitignore", "/llvm/");
wd.create_dir("src/llvm");
wd.create("src/llvm/foo", "test");
let lines: String = wd.stdout(&mut cmd);
let expected = path("src/llvm/foo:test\n");
assert_eq!(lines, expected);
cmd.current_dir(wd.path().join("src"));
let lines: String = wd.stdout(&mut cmd);
let expected = path("llvm/foo:test\n");
assert_eq!(lines, expected);
});
// See: https://github.com/BurntSushi/ripgrep/issues/30
clean!(regression_30, "test", ".", |wd: WorkDir, mut cmd: Command| {
if cfg!(windows) {
wd.create(".gitignore", "vendor/**\n!vendor\\manifest");
} else {
wd.create(".gitignore", "vendor/**\n!vendor/manifest");
}
wd.create_dir("vendor");
wd.create("vendor/manifest", "test");
cmd.arg("--debug");
let lines: String = wd.stdout(&mut cmd);
let expected = path("vendor/manifest:test\n");
assert_eq!(lines, expected);
});
// See: https://github.com/BurntSushi/ripgrep/issues/49
clean!(regression_49, "xyz", ".", |wd: WorkDir, mut cmd: Command| {
wd.create(".gitignore", "foo/bar");
wd.create_dir("test/foo/bar");
wd.create("test/foo/bar/baz", "test");
wd.assert_err(&mut cmd);
});
// See: https://github.com/BurntSushi/ripgrep/issues/50
clean!(regression_50, "xyz", ".", |wd: WorkDir, mut cmd: Command| {
wd.create(".gitignore", "XXX/YYY/");
wd.create_dir("abc/def/XXX/YYY");
wd.create_dir("ghi/XXX/YYY");
wd.create("abc/def/XXX/YYY/bar", "test");
wd.create("ghi/XXX/YYY/bar", "test");
wd.assert_err(&mut cmd);
});
// See: https://github.com/BurntSushi/ripgrep/issues/65
clean!(regression_65, "xyz", ".", |wd: WorkDir, mut cmd: Command| {
wd.create(".gitignore", "a/");
wd.create_dir("a");
wd.create("a/foo", "xyz");
wd.create("a/bar", "xyz");
wd.assert_err(&mut cmd);
});
// See: https://github.com/BurntSushi/ripgrep/issues/67
clean!(regression_67, "test", ".", |wd: WorkDir, mut cmd: Command| {
wd.create(".gitignore", "/*\n!/dir");
wd.create_dir("dir");
wd.create_dir("foo");
wd.create("foo/bar", "test");
wd.create("dir/bar", "test");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, path("dir/bar:test\n"));
});
// See: https://github.com/BurntSushi/ripgrep/issues/90
clean!(regression_90, "test", ".", |wd: WorkDir, mut cmd: Command| {
wd.create(".gitignore", "!.foo");
wd.create(".foo", "test");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, ".foo:test\n");
});
// See: https://github.com/BurntSushi/ripgrep/issues/93
clean!(regression_93, r"(\d{1,3}\.){3}\d{1,3}", ".",
|wd: WorkDir, mut cmd: Command| {
wd.create("foo", "192.168.1.1");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, "foo:192.168.1.1\n");
});
// See: https://github.com/BurntSushi/ripgrep/issues/99
clean!(regression_99, "test", ".",
|wd: WorkDir, mut cmd: Command| {
wd.create("foo1", "test");
wd.create("foo2", "zzz");
wd.create("bar", "test");
cmd.arg("-j1").arg("--heading");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(sort_lines(&lines), sort_lines("bar\ntest\n\nfoo1\ntest\n"));
});
// See: https://github.com/BurntSushi/ripgrep/issues/105
clean!(regression_105_part1, "test", ".", |wd: WorkDir, mut cmd: Command| {
wd.create("foo", "zztest");
cmd.arg("--vimgrep");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, "foo:1:3:zztest\n");
});
// See: https://github.com/BurntSushi/ripgrep/issues/105
clean!(regression_105_part2, "test", ".", |wd: WorkDir, mut cmd: Command| {
wd.create("foo", "zztest");
cmd.arg("--column");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, "foo:3:zztest\n");
});
// See: https://github.com/BurntSushi/ripgrep/issues/20
sherlock!(feature_20, "Sherlock", ".", |wd: WorkDir, mut cmd: Command| {
cmd.arg("--no-filename");
let lines: String = wd.stdout(&mut cmd);
let expected = "\
For the Doctor Watsons of this world, as opposed to the Sherlock
be, to a very large extent, the result of luck. Sherlock Holmes
";
assert_eq!(lines, expected);
});
// See: https://github.com/BurntSushi/ripgrep/issues/68
clean!(feature_68, "test", ".", |wd: WorkDir, mut cmd: Command| {
wd.create(".gitignore", "foo");
wd.create(".ignore", "bar");
wd.create("foo", "test");
wd.create("bar", "test");
cmd.arg("--no-ignore-vcs");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, "foo:test\n");
});
// See: https://github.com/BurntSushi/ripgrep/issues/70
sherlock!(feature_70, "sherlock", ".", |wd: WorkDir, mut cmd: Command| {
cmd.arg("--smart-case");
let lines: String = wd.stdout(&mut cmd);
let expected = "\
sherlock:For the Doctor Watsons of this world, as opposed to the Sherlock
sherlock:be, to a very large extent, the result of luck. Sherlock Holmes
";
assert_eq!(lines, expected);
});
// See: https://github.com/BurntSushi/ripgrep/issues/89
sherlock!(feature_89_files_with_matches, "Sherlock", ".",
|wd: WorkDir, mut cmd: Command| {
cmd.arg("--null").arg("--files-with-matches");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, "sherlock\x00");
});
// See: https://github.com/BurntSushi/ripgrep/issues/89
sherlock!(feature_89_count, "Sherlock", ".",
|wd: WorkDir, mut cmd: Command| {
cmd.arg("--null").arg("--count");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, "sherlock\x002\n");
});
// See: https://github.com/BurntSushi/ripgrep/issues/89
sherlock!(feature_89_files, "NADA", ".",
|wd: WorkDir, mut cmd: Command| {
cmd.arg("--null").arg("--files");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, "sherlock\x00");
});
// See: https://github.com/BurntSushi/ripgrep/issues/89
sherlock!(feature_89_match, "Sherlock", ".",
|wd: WorkDir, mut cmd: Command| {
cmd.arg("--null").arg("-C1");
let lines: String = wd.stdout(&mut cmd);
let expected = "\
sherlock\x00For the Doctor Watsons of this world, as opposed to the Sherlock
sherlock\x00Holmeses, success in the province of detective work must always
sherlock\x00be, to a very large extent, the result of luck. Sherlock Holmes
sherlock\x00can extract a clew from a wisp of straw or a flake of cigar ash;
";
assert_eq!(lines, expected);
});
@@ -619,7 +868,7 @@ fn binary_search_no_mmap() {
let mut cmd = wd.command();
cmd.arg("-a").arg("--no-mmap").arg("foo").arg("file");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, "foo\nfoo\n");
assert_eq!(lines, "foo\x00bar\nfoo\x00baz\n");
}
#[test]
@@ -632,13 +881,23 @@ fn files() {
let mut cmd = wd.command();
cmd.arg("--files");
let lines: String = wd.stdout(&mut cmd);
if cfg!(windows) {
assert!(lines == "./dir\\file\n./file\n"
|| lines == "./file\n./dir\\file\n");
} else {
assert!(lines == "./file\n./dir/file\n"
|| lines == "./dir/file\n./file\n");
}
assert!(lines == path("file\ndir/file\n")
|| lines == path("dir/file\nfile\n"));
}
// See: https://github.com/BurntSushi/ripgrep/issues/64
#[test]
fn regression_64() {
let wd = WorkDir::new("regression_64");
wd.create_dir("dir");
wd.create_dir("foo");
wd.create("dir/abc", "");
wd.create("foo/abc", "");
let mut cmd = wd.command();
cmd.arg("--files").arg("foo");
let lines: String = wd.stdout(&mut cmd);
assert_eq!(lines, path("foo/abc\n"));
}
#[test]