A C# implementation of Parsing Expression Grammars, intended to be used as a replacement for regular expressions. SharpPeg will dynamically compile PEGs to CIL, delivering performance similar to that of common regular expression engines.

The peg-match tool uses SharpPeg to implement a tool similar to grep, using PEGs instead of regular expressions.

The latest stable release of peg-match can be downloaded from the releases page. Extract the folder corresponding with your platform of choice to a directory, for example, for Windows you should extract the win\x64 folder. From the terminal, cd to this folder and run ./peg-match.

Note: on linux, you might still need to install some dependencies. For example, Ubuntu requires libunwind to be installed:

sudo apt install libunwind8

peg-match can be used by invoking it from the command line. For example:

./peg-match '"th" [a-z]+' input.txt

The command above will print all occurrences of words starting with 'th' in the file input.txt.

A part of the expression may be printed by using captures:

./peg-match -c word '"the " {word: [a-z]+}' input.txt

The command above will match all words prefixed by "the", but it will not print "the" in the output.

peg-match can optionally output structured data instead of plain text by using the -j switch:

./peg-match -j File::Csv::CsvFile data.csv

The command above will parse the csv file, and output the parsed data as a CSV file. Any grammar containing one or more captures can be converted to JSON. For example, a very simple expression to match key=value pairs might look like this:

./peg-match -j '{key: [A-Za-z0-9]+}"="{value: [A-Za-z0-9]+}' data.txt

Which will output a JSON object that looks something like this:

[
  {
    "key": "username",
    "value": "jos"
  },
  {
    "key": "password",
    "value": "hunter1"
  }
]

This work is licensed under the LGPL license. Please see the LICENSE file for more info.