a regexp wrapper, less pain and more fun with regexp :)
- compose the regexp by string, regexp, rewrap object
- compositional method
- refer to the predefined regexp
- get match by named path
npm install rewrap
dist/rewrap.js and dist/rewrap.min.js can be used in the browser
const rewrap = require('rewrap');
const {re, wrap} = rewrap; // wrap is the same as rewrap;
// This npm package contains some frequently used regexp.
// It can be used as the registry for rewrap.
const reBuiltins = require('regexp-frequent');
rewrap.registry(reBuiltins);
re(/^/) // /^/
.ref('digits') // /^[0-9]+/
.and(/abc/) /^[0-9]+abc/
.or(/bc/, /cd/) // /^[0-9]+abc|bc|cd/
// because the line below is leaded by re,
// so @digits will not refer to registry,
// and will generate regexp as is.
.and(/@digits/) // /^[0-9]+abc|bc|cd@digits/
.wrap // now switch to rewrap style
// @digits will refer to registry, will be replaced by [0-9]+
// necessary () is added automatically for or expression
.and(/@digits/) // /(?:^[0-9]+abc|bc|cd)@digits[0-9]+/
// &a indicate that a will become (a)
// and will be saved as result.$$rewrapData.a while matching
.and(/a&a/) // /^[0-9]+abc|bc|cd@digits[0-9]+(a)/
.and(
// generate (xy)
// will be saved as result.$$rewrapData.xy while matching
// same as wrap(/(xy)&xy/)
wrap('x').and('y').save('xy') // /(xy)/
) // /(?:^[0-9]+abc|bc|cd)@digits[0-9]+(a)(xy)/
// the whole above will be saved as result.$$rewrapData. part1 while matching
.save('part1')
// add current rewrap to registry as sample
.register('sample')
// refer to sample
.rewrap(/@sample/) // /((?:^[0-9]+abc|bc|cd)@digits[0-9]+(a)(xy))((?:^[0-9]+abc|bc|cd)@digits[0-9]+(a)(xy))/- rewrap, regexp, or, and, paren, not, lookAhead, optional, any, some, repeat, times, timesMore, headTail, head, tail
they have no side effects, i.e. they always generate a new Rewrap object.
- not, lookAhead, optional, any, some, repeat, times, timesMore
These methods can with/without an item argument, If without an item argument, it will operate on itself(this) and return a new Rewrap object.
someRewrap.register(registry)
someRewrap.register(refPath, value)
someRewrap.register(refPath): value is someRewrap itself
someRewrap.register({refPath:value, ...}): register multiple refPath:value pairs
someRewrap.ref(refPath): use the predefined rewrap in the registry(someRewrap._registry)
someRewrap.and(/@refPath%/): use the predefined rewrap in the registry(someRewrap._registry)
refPath is like x, y, a.b.c, _a, a0.b1._c
% can be omit if the following char is not identifier chars [_A-Za-z0-9]
NB. $ is not allowed in the refPath or dataPath.
someRewrap.save(dataPath): similar to /@someRewrap&dataPath%/
dataPath is like x, y, a.b.c, _a, a0.b1._c
% can be omit if the following char is not identifier chars [_A-Za-z0-9]
If necessary, () will be added around someRewrap to generate a sub match.
flags, registry
someRewrap.re: lead the regexp style chain. In regexp style, @refPath% and @dataPath% will be treated as raw regexp string.
someRewrap.wrap: lead the rewrap style chain, @refPath% and @dataPath% will be treated like ref(refPath) and save(dataPath).
Please dive into the code for more information. The code is simple and intuitive. Look the tests for more examples.
The npm package regexp-frequent contains some frequently used regexp. It can be used as the registry for rewrap.
The npm package rewrap-patch does monkey patch on String.prototype, make the methods(match, search, split, replace) works better with rewrap. Most of time you do NOT need rewrap-patch, except that you need the match method support someString.match(aRewrap, data).
These two npm packages have similar part fearture with rewrap: save sub match with name. But rewrap is more convenient and powerful than them(no parens is necessary for name, and dot separated path is supported):
named-regexp: https://www.npmjs.com/package/named-regexp
named-js-regexp: https://www.npmjs.com/package/named-js-regexp
MIT, see LICENSE