srch

Text Search For Humans

srch is a cli to run text expressions and perform basic text operations such as filtering, ignoring and replacing on the command line. There are many great tools that do this job. But most other tools have one in common: They are hard to memorize if you dont use them regularly. srch tries to solve this issue by providing a super simple cli & expression language which can be easily memorized and is well documented.

Quickstart

$ srch for 'equals "foobar"' -m word                # matches all occurences `foobar` in the text
$ srch for 'length 20'                              # matches all lines with 20 chars
$ srch not 'numeric or special'                     # ignores all lines which contain only numbers and special chars
$ srch replace 'numeric and length 5' 12345 -m word # replaces all 5 digit numbers with `12345`

Common tasks where srch excels grep in readability

Task	srch	grep
Find all words containing a string	srch for 'contains "substr"' -m word	grep -oh "\w*substr\w*"
Find all lines in a file with a specific length	srch for 'length 10'	grep -x '.\{10\}'
Ignore all lines containing a string	srch not 'contains "hide me"'	grep -v "hide me"
Replacing all words following a specific pattern	srch replace 'numeric and length 5' 12345 -m word	grep itself cant replace, you need to use sed for that (which gets even more complicated).
Replacing all email addresses in a file with your email	srch replace 'contains "@" and contains ".com"' [email protected] -m word	Same as above.

When to use other tools

As said earlier: srch is no direct competitor to grep, awk, etc.! If you find yourself reaching the limits of the text expression language, you probably want to use more advanced tools.

Installing

At the moment srch can be installed only via cargo using:

$ cargo install srch

Documentation

There are the following global options:

• -m / --mode, sets the operation mode, can be either line or word, defaults to line

And there are the following global flags:

• -f / --first`, print only the first match if available
• -l / --last`, print only the last match if available
• --skip n, skip the first n matches
• --limit n, show at most n matches

srch for [FLAGS] [OPTIONS] <EXPRESSION> [FILE]
srch not [FLAGS] [OPTIONS] <EXPRESSION> [FILE]
srch replace [FLAGS] [OPTIONS] <EXPRESSION> <REPLACEMENT> [FILE]

If no file is provided srch tries to read from stdin.

Examples

$ docker ps | srch for 'alphanumeric and length 12' -m word # prints all docker container ids

The Text Expression Language

This is a super simple format of writing readable and easy to memorize text processing expressions - there are many great and far more advanced languages and tools to process text on the commandline out there but all of them have one problem in common - they're unreadable and hard to memorize if not used often.

The Text Expression Languages provides only 9 Attributes to query by. These attributes indicate the format of a string which gets tested against it.

Attribute	Resolve to true if the tested string
starts <str>	starts with the given string
ends <str>	ends with the given string
contains <str>	contains a substring equal to the given string
equals <str>	exactly equals the given string
length <int>	has the given length
numeric	contains only numeric chars
alpha	contains only alphabetic chars
alphanumeric	contains only alphanumeric chars
special	contains only special chars

Currently there are only two binary logical operations: and and or

Operator	Boolean Algebra
and	Conjunction
or	Disjunction

Attributes can be concattenated by logical operators.

Examples

starts "FOO" and ends "BAR"
contains "@" and contains ".com"
length 5 and length 10
numeric and length 8

Limitations

This Syntax might not cover all use cases. It's not meant to do that. If you find yourself reaching the limits of this language you might want to use more advanced tools (such as awk, grep, sed..)