README.md

package grammar is designed to make long and convoluted regular expressions easier to handle.

Features

• Enable the use of whitespaces within regexes to make them less dense.
• Allow spreading regexes across multiple lines to allow easier grouping.
• Allow inline comments in regexes.
• Interpolation of regexes: Allows a regex to be composed of multiple sub-rules.

Usage

package grammar // import "github.com/gaissmai/grammar"

Whitespaces / Newlines

Example regex containing white-spaces and newlines to make it more readable. This regex matches number of the form 1.2e+42, .3, 11, 42., 3.1415,....

   `                       // NUMBER: 1.2e+42, .3, 11, 42., 3.1415, ...
    [+-]?                  // first, match an optional sign
    (?:                    // then match f.p mantissas or integer:
         \d+ \. \d+        // mantissa of the form a.b
       | \d+ \.            // or mantissa of the form a.
       |     \. \d+        // or mantissa of the form .b
       | \d+               // or integer of the form a
    )
    (?: [eE] [+-]? \d+ )?  // finally, optionally match an exponent
    `

This yields [+-]?(\d+\.\d+|\d+\.|\.\d+|\d+)([eE][+-]?\d+)? but is much more readable.

Regex Interpolation

Complex rules can be comprised of simpler subrules using string interpolation. For example: Using the above regex as ${NUMBER}, you can easily assemble a rule that matches many numbers using the following snippet:

     ^                  // 1.23   3.1415 0.5E3 ...
       ${NUMBER}        // start with number
       (?:
         \s+ ${NUMBER}  // followed by one ore more numbers, separated by whitespace(s)
       )+
    $

Any number of rules can be added to a grammar, dependent or independent, as long as there are no cyclic dependencies.

The package has a very thin API, please see the examples in the documentation.

ATTENTION

The package is already in production use, but the API may still change.