hunt

Crates.iohunt
lib.rshunt
version
sourcesrc
created_at2022-01-09 23:21:45.171578
updated_at2024-11-13 16:33:40.532679
descriptionSimple command to quickly search a file/folder by name.
homepage
repositoryhttps://github.com/LyonSyonII/hunt-rs
max_upload_size
id511049
Cargo.toml error:TOML parse error at line 19, column 1 | 19 | autolib = false | ^^^^^^^ unknown field `autolib`, expected one of `name`, `version`, `edition`, `authors`, `description`, `readme`, `license`, `repository`, `homepage`, `documentation`, `build`, `resolver`, `links`, `default-run`, `default_dash_run`, `rust-version`, `rust_dash_version`, `rust_version`, `license-file`, `license_dash_file`, `license_file`, `licenseFile`, `license_capital_file`, `forced-target`, `forced_dash_target`, `autobins`, `autotests`, `autoexamples`, `autobenches`, `publish`, `metadata`, `keywords`, `categories`, `exclude`, `include`
size0
Liam Garriga (LyonSyonII)

documentation

https://github.com/LyonSyonII/hunt-rs

README

Crates.io
A highly-opinionated simplified Find command made with Rust.
By default it searches a file/folder in the working directory and divides the result between exact matches and ones that only contain the query.
Results will be sorted alphabetically.

For example, hunt SomeFile / will search "SomeFile" from the root directory, and the output could be:

Contains:
/SomeFileIsHere
/home/lyon/Downloads/abcdefgSomeFileeee
/mnt/Files/--SomeFile--

Exact:
/home/lyon/SomeFile

Check the Benchmarks for a comparison with other tools.

Usage

hunt [OPTIONS] [NAME] [SEARCH_IN_DIRS]...

By default, searches are case-insensitive, unless [NAME] contains an uppercase letter or the --case-sensitive flag is set.

Options

-f, --first
        Stop when first occurrence is found

-e, --exact
        Only search for exactly matching occurrences, any file only containing the query will be skipped
        
        e.g. if query is "SomeFile", "I'mSomeFile" will be skipped, as its name contains more letters than the search

-c, --canonicalize
        If enabled, all paths will be canonicalized

-C, --case-sensitive
        If enabled, the search will be case-sensitive
        
        Note that case-sensitivity will be activated automatically when the search query contains an uppercase letter

-v, --verbose
        Print verbose output
        
        It'll show all errors found: e.g. "Could not read /proc/81261/map_files"

-s, --simple...
        Prints without formatting (without "Contains:" and "Exact:")
        
        -ss Output is not sorted

-H, --hidden
        If enabled, it searches inside hidden directories
        
        If not enabled, hidden directories will be skipped

--select
        When the search is finished, choose one file between the results
        
        The selected file will be printed as if -ss was used

--multiselect
        When the search is finished, choose between the results
        
        The selected files will be printed one after the other, separated by spaces

-S, --starts <STARTS_WITH>
        Only files that start with this will be found

-E, --ends <ENDS_WITH>
        Only files that end with this will be found

-t, --type <FILE_TYPE>
        Specifies the type of the file
        
        'f' -> file | 'd' -> directory

-i, --ignore <IGNORE_DIRS>
        Ignores this directories. The format is:
        
        -i dir1,dir2,dir3,...

-h, --help
        Print help (see a summary with '-h')

-V, --version
        Print version
    

If the --first flag is set, the order in which the file will be searched is [current_dir, home_dir, root].
If you're already in one of these directories, current_dir will be skipped.

If the --hidden flag is not set, hidden files/directories will be skipped.

Args

[NAME]  Name of the file/folder to search
        By default, searches are case-insensitive, unless the query contains an uppercase letter.

[SEARCH_IN_DIRS]...
        Directories where you want to search
        If provided, hunt will only search there
        
        These directories are treated independently, so if one is nested into another the
        search will be done two times:  
        
        e.g. "hunt somefile /home/user /home/user/downloads" will search in the home
        directory, and because /home/user/downloads is inside it, /downloads will be
        traversed two times

Examples

  • Search for a specific file on the whole system (hunt will stop once found)

      hunt -f -e SomeFile
    
  • Search for files containing "SomeFile"

      hunt SomeFile
    
  • Search file in the home directory

      hunt -e SomeFile ~/
    
  • Search file in the downloads and pictures directories

      hunt -e SomeFile ~/downloads ~/pictures
    
  • Search all files that end with ".exe"

      hunt --ends .exe
    
  • Search all files that end with ".exe" in the wine directory

      hunt --ends .exe ~/.wine
    
  • Search all files that start with "." (all hidden files)

      hunt --starts .
    
  • Search all files that end with ".exe", start with "M" and contain "wind" in the wine directory

      hunt --starts=M --ends=.exe wind ~/.wine
    
  • Search a directory named "folder"

      hunt -t=d folder
    
  • Search a file named "notfolder"

      hunt -t=f notfolder
    
  • Remove all files named "SomeFile"

      hunt -s -e SomeFile | xargs rm -r
    

Motivation

Normally when I search for a file, I don't know the exact subdirectory where it is, so I end up searching in the whole $HOME directory.

Using the find command for this ended up being very slow, as it took a lot of time to traverse all the directories, and the output was also hard to read.

locate was faster, but it didn't always find the file I was looking for, as it only searches in its database which isn't updated in real time.

Seeing how find did not perform any parallelism at all, I decided to make a multithreaded version of it, and that's how Hunt was born.

Hunt is multithreaded, so it's a lot faster than find, and more reliable than locate (recent files cannot be found with it).

Installation

Packaging status

Precompiled binaries

Download the latest binary from releases.

Or install with cargo-binstall:

cargo binstall hunt

Compile from source

First check that you have Rust installed, then run

cargo install hunt

Benchmarks

Let's compare Hunt with some of the most used tools: the GNU locate and find and the very popular also written in rust, fd.

For benchmarking I'm using hyperfine, a tool developed by the fd dev.
These are done in a system with approximately 2,762,223 files, with a network drive and an external one.
Results on other systems may vary, so take this comparisons as a guide.

If you want to reproduce the benchmarks, you can do so by running the benchmarks.sh file from this repository.

Searching file in ~/

Find first occurrence of a heavily nested file in a hidden folder from the home directory. File is located in /home/user/.wine/drive_c/users/user/AppData/Local/mygame/User Data/Crashpad/reports/SomeFile.

Benchmark 1: hunt --hidden --first --exact SomeFile ~/
  Time (mean ± σ):     180.2 ms ±   7.4 ms    [User: 406.6 ms, System: 1135.9 ms]
  Range (min … max):   167.2 ms … 198.5 ms    16 runs
 
Benchmark 2: fd --hidden --no-ignore --glob --color=never --max-results=1 SomeFile ~/
  Time (mean ± σ):     913.6 ms ±  52.5 ms    [User: 2584.8 ms, System: 4628.6 ms]
  Range (min … max):   858.6 ms … 1018.6 ms    10 runs
 
Benchmark 3: find ~/ -name SomeFile -print -quit 2>/dev/null
  Time (mean ± σ):      2.219 s ±  0.071 s    [User: 0.587 s, System: 0.988 s]
  Range (min … max):    2.160 s …  2.395 s    10 runs
 
Benchmark 4: locate -n 1 -A SomeFile
  Time (mean ± σ):      1.244 s ±  0.015 s    [User: 1.231 s, System: 0.010 s]
  Range (min … max):    1.231 s …  1.281 s    10 runs
 
Summary
  'hunt --hidden --first --exact SomeFile ~/' ran
    5.07 ± 0.36 times faster than 'fd --hidden --no-ignore --glob --color=never --max-results=1 SomeFile ~/'
    6.90 ± 0.30 times faster than 'locate -n 1 -A SomeFile'
   12.31 ± 0.64 times faster than 'find ~/ -name SomeFile -print -quit 2>/dev/null'

Hunt

--hidden, search all files (it normally ignores hidden files and directories in the Ignore List).
--first, stop when first occurrence is found.
--exact, only search for files/folders named "SomeFile", names that only contain the pattern will be skipped.

Searching all files that contain "SomeFile"

Find all occurrences of "SomeFile" from the root directory (worst case scenario, checking all files in the system).

The occurrences in question are:

/home/lyon/Downloads/abcdefgSomeFileeee
/SomeFileIsHere
/mnt/Files/--SomeFile--
/home/lyon/.wine/drive_c/Program Files (x86)/Internet Explorer/SomeFile

For this benchmark I'll skip Locate. It's obviously faster because it doesn't traverse all the filesystem, as it is backed up by a database.
It must be noted though that the file in /mnt/Files was not found, as the database does not keep record of files in other drives.
For the curious, it scored a time of 486.8 ms, only 1.32 times faster than Hunt.

Hunt

Benchmark 1: hunt -H SomeFile /
  Time (mean ± σ):     633.6 ms ±  25.1 ms    [User: 2876.7 ms, System: 2507.5 ms]
  Range (min … max):   589.4 ms … 671.2 ms    10 runs

Fd

Benchmark 2: fd -HI -c never SomeFile /
  Time (mean ± σ):      1.452 s ±  0.014 s    [User: 4.116 s, System: 8.693 s]
  Range (min … max):    1.431 s …  1.474 s    10 runs

Find

Benchmark 3: find / -name "*SomeFile*"
  Time (mean ± σ):      3.473 s ±  0.144 s    [User: 1.234 s, System: 1.602 s]
  Range (min … max):    3.374 s …  3.874 s    10 runs

Summary

'hunt -H SomeFile /' ran
  2.29 ± 0.09 times faster than 'fd -HI -c never SomeFile /'
  5.48 ± 0.31 times faster than 'find / -name "*SomeFile*"'

Conclusion

Hunt is faster than other alternatives if you don't need a lot of features (like regex).
Think of it as a simple "where did I put that file?" solution.

Commit count: 242

cargo fmt