ret 0.1.4

A pure-python command-line regular expression tool for stream filtering, extracting, and parsing.

A pure-python command-line regular expression tool for stream filtering, and extracting, designed to be minimal with an intuitive command-line interface.

Installation

You can install this via

$ python3 -m pip install ret
✨🍰✨

or using pipx

$ pipx install ret
✨🍰✨

Ret is pure python (tested on python 3.6+) with no dependencies.

That way, you can get a clean uninstall.

Note

If you want to install the bleeding edge version of ret, right when it gets pushed to master, see here for instructions.

Usage

Example

You can use ret to extract text via regex capture groups:

$ git branch
* master
$ git branch | ret "\* (\w+)" s --group 1
master

…finding all occurrences of a pattern:

$ ls | ret ".*\.py" findall
foo.py
bar.py

and even all occurrences of a pattern with capture groups:

$ ls | ret "(.*)\.py" findall --group 1
foo
bar

While those may seem untypical use cases, I have found myself using Ret countless times.

Here’s a one-liner for uninstalling unnecessary stuff for pip:

$ pip list --not-required | ret ".+\n.+\n((?:\n|.)+)" f -g 1 | ret "([^\s]+)\s+.+\n" f -g 1 | xargs pip uninstall --yes

Another case

Imagine this: you have just downloaded a bunch of tarballs, and have ran

for x in $(grep ".+\.tar\.gz"); do tar -xzf $x; done

Now you just want to cd into all of the extracted files, run ./configure && make && make install.

You could use ret to get the names of the extracted files, just from the tarballs’ names. Like this:

$ ls | grep ".+\.tar\.gz"
foo.tar.gz
bar.tar.gz
foobar.tar.gz
extractme.tar.gz


$ ls | ret "(.+)\.tar\.gz" f -g 1
foo
bar
foobar
extractme

and with that combined, we can do

$ for x in (ls | ret "(.+)\.tar\.gz" f -g 1); do {
   cd $x &&
   ./configure && make && make install &&
   cd -}; done
✨🍰✨

A life saver.

---

And remember, this is python regex: a very powerful regular expression engine.

The possibilities of usage are endless.

Demonstration

Background

I love grep. But grep isn’t really for text extraction.

For example, you cannot extract regexes via capture groups.

Since I wanted that functionality, I decided to build this, Ret.

Why the name?

Ret is an acronym for regular expression tool.

Why it can’t replace grep

Ret originally was designed to provide some features grep lacks. It never intended to replace good ol’ grep.

Grep is great for searching directories while ret (currently) can only read from a file or stdin.

Furthermore, you cannot guarantee that ret is installed on the machine.

Also, Ret relies on the (slow) python regex engine.

What about sed?

Sed is for find-and-replacing streams. Ret’s different. But together, you can do some powerful things

Feel free to contribute!