I recently installed silversearcher-ag with sudo and uninstalled it as it required sudo to run the command every single time. The command I used to install and uninstall this package respectively:

# installation
sudo apt install silversearcher-ag

# uninstallation
sudo apt remove silversearcher-ag
sudo apt autoremove

As mentioned by the comment, here's the output for type -a ag: ag is aliased to `apt-get update;apt-get dist-upgrade' After uninstallation, I found that if I do sudo ag, the command would correctly fail due to command not found; however, if I simply do a ag as user, I was still able to run the command ( although the command actually failed due to lack of permission, i.e., it has to be run with sudo every single time). I've looked at places like /usr/bin/, /usr/local/bin, /usr/sbin/, /usr/local/sbin, sbin/, bin/, the ag command didn't reside in any of them. whereis, locate returned no matching results. I've also tried sudo apt purge silversearcher-ag sudo apt purge silversearcher-ag* The first command actually removed some additional affliations, the second simply removed none. After these two commands the ag was still there. What do I need to do to completely remove this package? I am running MX Linux 23-ahs, based on Debian 12 bookworm.

There is an option --ignore which allows specifying files to ignore. At the moment I only managed to ignore multiple files by doing --ignore file1 --ignore file2....... Trying to use --ignore "*assets*|*scripts*" does nothing. So is there a catch I'm not aware of?

I don't want [ag](https://github.com/ggreer/the_silver_searcher) to search files larger than 2Mio. I can't find anything like a --file-size flag to the ag command, but is there any way to tell it that? I'm getting lots of

ERR: Skipping foo.bar: pcre_exec() can't handle files larger than 2147483647 bytes.

This will be easier to explain via example. Here are my input files:

file1:
x
x
a
b
c
x
x

file2:
x
x
c
b
a
x
x

file3:
x
x
x x a b c x x
x
x

file4:
x
x
x x c b a x x
x
x

file5:
x
x
a b
x
x

file6:
x
x
x x b c x x
x
x

file7:
x
x
x x b b x x
x
x

I want to search for files that have regexp a AND c. This will return files 1-4. I want the output to look as close as possible to ag -C --pager="less -R" regexp's output. Here's what that would look like (I am surrounding results with angle brackets to represent color highlighting):

file1:
2: x
3: 
4: b
5: 
6: x

file2:
2: x
3: 
4: b
5: 
6: x

file3:
2: x
3: x x  b  x x
4: x

file4:
2: x
3: x x  b  x x
4: x

Or maybe ag would print it like this:

file1:
2: x
3: 
4: b
--
4: b
5: 
6: x

I'm not sure, but this detail doesn't matter to me. Here's what does: 1. The highlighting 1. The relative path of the file above the results 1. The features less provides for navigation and search 1. It can find multiple regexps ANDed together 1. That -C option still exists on the command line The line numbers are a nice to have, but not necessary. --- I've tried many many things, and this is the closest I've gotten: ## Step 1: Precompile a file list of each individual regexp

for x in $array_of_regexp_file_names; 
  do r=${x/.txt/} ; # remove .txt from the end
  ag -il $r | sort > $x & ; # sort the list of FILES with this single regexp
done

This gives a list of files for each regexp, sorted by filename. ## Step 2: Use ag to search the intersection of 2+ file lists I will break down the following:

ag -C 1 --pager="less -R" "regexp1|regexp2" $(comm -12 regexp1.txt regexp2.txt)

### $(comm -12 regexp1.txt regexp2.txt) This command finds the intersection of two file lists. That is, the red in this picture: ### ag -C 1 --pager="less -R" "regexp1|regexp2" ... Here, I am giving ag a regexp that I *know* exists in every file in that intersection. That may seem redundant, but I'm doing it because want those words highlighted in the output. It makes my life 1000x easier. Here's the problem: that intersection is so many files, running the command gives me this output:

zsh: argument list too long: ag

Other than that, my workaround works. I've tested this by running a command like this:

ag -C 1 --pager="less -R" "regexp1|regexp2" $(comm -12 patter1.txt regexp2.txt | head -10)

The problem is the intersecting list is so long, it doesn't fit on the command line. If ag provided an option to pass a list of files to search, I could get past this, but it doesn't have that functionality. Regardless, I'm hoping I don't need to: I'm assuming there's a much easier solution to this problem, I just don't know what it is. --- Edit: To solidify some highlighting rules, here are some other examples: ### Example 1 regexs:

regex1 = a.
regex2 = .b

input

file8:
x
x
abc
x
x

Output:

2: x
3: c
4: x

### Example 2 regexs:

regex1 = foo
regex2 = oba

input

file9:
x
x
foobar
x
x

Output:

2: x
3: bar
4: x

I picked these outputs because that's what grep and ag already do, but I'm pretty ambivalent about both of these scenarios, so if these examples are challenging to implement, I don't mind if highlighting works differently in these edge cases; in general, *my* regexps won't overlap.

I'm using ag v 2.2. How do I configure ag so that it will search for files beginning with a "." in addition to all other files? I'm noticing I have a file like so $ cat client/.env.production REACT_APP_PROXY=https://map.chicommons.coop but when I search with "ag" using a term in that file, the term doesn't come up ... $ ag 'map\.chicom' . web/directory/settings.py 28:ALLOWED_HOSTS = ['127.0.0.1', 'localhost', 'dev.chicommons.coop', 'map.chicommons.coop'] client/config.js 3: ? "map.chicommons.com"

I define ag as bash function, to make usage relatively transparent

ag() { [ $# -gt 0 ] && vim -c silent\ SyntasticToggleMode -c copen -q 
Vim: Warning: Output is not to a terminal

Only way out is to kill stuck vim process from another shell. What's going wrong here and how can this bash function be upgraded (or find some alternative) to make it TAB-completion compatible ? EDIT ag doesn't ship any upstream completion

$ [git:master] complete -p ag
-bash: complete: ag: no completion specification

I'm using Silver Searcher, ag, version 2.1.0 on Ubuntu. I'm trying to search for the string aStar in Haskell files in the current directory tree. (The files are text files with extension .hs.) The behaviour isn't what I expect.

$ ag aStar

finds no matches.

ag --haskell aStar

finds no matches.

ag --haskell --all-text aStar

finds the correct matches in the .hs files.

ag --all-text aStar

finds the correct matches in .hs files, plus a bunch in some temporary files I don't want searched. Any suggestions on what I'm doing wrong, and why the --haskell flag doesn't make ag search the Haskell files?

I'm using ag (The Silver Searcher ) version 0.31.0. I can easily look for a string in a bunch of files using: localhost:workspace davea$ ag 'ftp' . But what if I only want to scan files with certain extensions? I tried this: localhost:workspace davea$ ag 'ftp' .java ERR: Error stat()ing: .java ERR: Error opening directory .java: No such file or directory but got the errors you see above.

Are these related? Which is faster? Can either be limited to directory-names? https://github.com/BurntSushi/ripgrep https://github.com/ggreer/the_silver_searcher

Give an example file with the content: // find me val ignore me = "" // I should not be found // find me // find me I want to find all the lines with find me using ag silversearcher. Basically, it should match any line optionally starting with any white space and a double space //. I know that: ag ^// example only yields in the first line without spaces. For the beginning of the line I know I can use ^, for whitespace it should be \s, and my custom string should either be // or it may need to be to escaped to: \/\/ That's why I thought: ag ^\s*// example ag ^\s*\/\/ example Should do the trick, yet it finds also only the first line. How can I match all the lines starting with white space and my custom string?

Say for example I have a file with contents like this: # TODO: Optimize this. alias bash='start bash' # FIXME: Just fix this. alias fix='there is something wrong with this What I want is if I search for pattern TODO: it should only print result like this: # TODO: Optimize this. alias bash='start bash' Is this possible?

I want to find every file in from the current dir and downwards whose filename starts with foo using __[ag](https://github.com/ggreer/the_silver_searcher)__(the silver searcher). I have tried: ag -g '^foo' ag -g '\^foo' ag -g '\Afoo' no luck. But it should work as ag implements PCRE syntax, right? What am I missing here?

With two files with echo 'example.com/call_me?param' > file1.txt echo 'example.com/call_me_maybe?param' >> file1.txt echo 'example.com/call_me?' >> file1.txt echo 'example.com/call_me?param' > file2.txt echo 'example.com/call_me_maybe?' >> file2.txt $ ag -v '\?param' file2.txt 2:example.com/call_me_maybe? file1.txt 3:example.com/call_me? $ ag '\/call_me_maybe' -v '\?param' ERR: Error stat()ing: \?param ERR: Error opening directory \?param: No such file or directory $ ag '\/call_me_maybe' -v '\?param' file1.txt ERR: Error stat()ing: \?param ERR: Error opening directory \?param: No such file or directory file1.txt 1:example.com/call_me?param 3:example.com/call_me? I would like to search for one term and exclude the other pattern. Is it possible in one command? I'm using ag because it digging into log.gz files. So I can use zgrep or ag -z.

Given a directory structure such as: dev/project1/assets/** dev/project2/assets/** dev/project3/assets/** Is it possible to search, from within the dev directory all assets directories only. Given that the project directories will include other sub-directories. I don't want to have to explicitly name each project directory. But something like search on /*/*/assets. So x directories deep, with a given name.

I have a list of java reserved words, first letter capitalised. $ tail -n5 ~/reservedjava.txt Break While True False Null I'm trying to look through all my java source code to find methods that look like getWhile(). cat ~/reservedjava.txt | parallel 'ag "get{}\(\)$"' This shows me nothing. Now, I know that I have a method getBreak(): $ ag "getBreak\(\)$" src/main/java/Foo.java 154: public Break getBreak() Here's what a dry run looks like: $ cat ~/reservedjava.txt | parallel --dry-run 'ag "get{}\(\)$"' | tail -n5 ag "getBreak\(\)$" ag "getWhile\(\)$" ag "getTrue\(\)$" ag "getFalse\(\)$" ag "getNull\(\)$" I'm using gnu parallel (v. 20130722) and the silver searcher (ag) (v. 0.18.1). If it makes a difference, I'm on Fedora 19, but have compiled these utilities myself. I get the same result with ack (v. 2.12).

I am trying to search text in a directory and it turned out that the following syntaxes do not return any result ack -i "0xabcdef" ./ ack -i "0xabcdef" ack -i "0xabcdef" . while the following command works ack -i "0xabcdef" * Can someone explain why that is the case? What is the significance of *? I also noticed that the directory has symbolic links.

I am thinking methods to make the search faster and/or better which principally uses fgrep or ag. Code which searches the word and case-insensitively at $HOME, and redirects a list of matches to vim find -L $HOME -xtype f -name "*.tex" \ -exec fgrep -l -i "and" {} + 2>/dev/null | vim -R - It is faster with ag because of parallelism and ack find -L $HOME -xtype f -name "*.tex" \ -exec ag -l -i "and" {} + 2>/dev/null | vim -R - Statistics --- Small group average statistics with fgrep and ag by time fgrep ag terdon1 terdon2 terdon3 muru user 0.41s 0.32s 0.14s 0.22s 0.18s 0.12s sys 0.46s 0.44s 0.26s 0.28s 0.30s 0.32s Cases terdon1 and terdon3 can be equal fast. I get great fluctuations with those two. Some Ranking by sys time (not best criteria!) 1. terdon1 2. terdon2 3. terdon3 4. muru 5. ag 6. fgrep Abbreviations --- - terdon1 = terdon-many-find-grep - terdon2 = terdon-many-find-fgrep - terdon3 = terdon-many-find-ag (without F because not exists in ag) Other codes --- muru's proposal in comments grep -RFli "and" "$HOME" --include="*.tex" | vim -R - OS: Debian 8.5 Hardware: Asus Zenbook UX303UA

How do I get ag to list all files, including empty ones? +ravi@boxy:~$ mkdir new && cd new +ravi@boxy:~/new$ echo stuff > non-empty && touch empty +ravi@boxy:~/new$ ag -ul non-empty +ravi@boxy:~/new$ How do I get empty to appear in the example above?

I want to use [ag](https://github.com/ggreer/the_silver_searcher/blob/master/doc/ag.1.md) to print the classes and their methods in a python file. I thought that this would be easy using: ag --context=0 --nocolor -os '^\s*(def|class)\s+[_A-Za-z]*' prog.py but for reasons I don't understand this is also matching blank lines. For example, if you make prog.py the following class MyFavouriteClass def __init__ def __contains__ blah class MyNextFavouriteClass def _repr_ def __iter__ then it returns the full file, including the blank lines, *except* for the line containing blah. Of course, I can always pipe the output into something else to remove the blank lines but I'd rather get it right the first time. I suspect that the problem has nothing to do with the regular expression and, instead, that it's a feature of ag's --context, --after and --before flags but I can't find a combination of these that does what I want. Any ideas?