Michael Winter 555e2c196e | 5 years ago | |
---|---|---|
.. | ||
node_modules/is-number | 5 years ago | |
LICENSE | 5 years ago | |
README.md | 5 years ago | |
index.js | 5 years ago | |
package.json | 5 years ago |
README.md
to-regex-range
Pass two numbers, get a regex-compatible source string for matching ranges. Validated against more than 2.78 million test assertions.
Install
Install with npm:
$ npm install --save to-regex-range
Install with yarn:
$ yarn add to-regex-range
What does this do?
This libary generates the source
string to be passed to new RegExp()
for matching a range of numbers.
Example
var toRegexRange = require('to-regex-range');
var regex = new RegExp(toRegexRange('15', '95'));
A string is returned so that you can do whatever you need with it before passing it to new RegExp()
(like adding ^
or $
boundaries, defining flags, or combining it another string).
Why use this library?
Convenience
Creating regular expressions for matching numbers gets deceptively complicated pretty fast.
For example, let's say you need a validation regex for matching part of a user-id, postal code, social security number, tax id, etc:
- regex for matching
1
=>/1/
(easy enough) - regex for matching
1
through5
=>/[1-5]/
(not bad...) - regex for matching
1
or5
=>/(1|5)/
(still easy...) - regex for matching
1
through50
=>/([1-9]|[1-4][0-9]|50)/
(uh-oh...) - regex for matching
1
through55
=>/([1-9]|[1-4][0-9]|5[0-5])/
(no prob, I can do this...) - regex for matching
1
through555
=>/([1-9]|[1-9][0-9]|[1-4][0-9]{2}|5[0-4][0-9]|55[0-5])/
(maybe not...) - regex for matching
0001
through5555
=>/(0{3}[1-9]|0{2}[1-9][0-9]|0[1-9][0-9]{2}|[1-4][0-9]{3}|5[0-4][0-9]{2}|55[0-4][0-9]|555[0-5])/
(okay, I get the point!)
The numbers are contrived, but they're also really basic. In the real world you might need to generate a regex on-the-fly for validation.
Learn more
If you're interested in learning more about character classes and other regex features, I personally have always found regular-expressions.info to be pretty useful.
Heavily tested
As of April 27, 2017, this library runs 2,783,483 test assertions against generated regex-ranges to provide brute-force verification that results are indeed correct.
Tests run in ~870ms on my MacBook Pro, 2.5 GHz Intel Core i7.
Highly optimized
Generated regular expressions are highly optimized:
- duplicate sequences and character classes are reduced using quantifiers
- smart enough to use
?
conditionals when number(s) or range(s) can be positive or negative - uses fragment caching to avoid processing the same exact string more than once
Usage
Add this library to your javascript application with the following line of code
var toRegexRange = require('to-regex-range');
The main export is a function that takes two integers: the min
value and max
value (formatted as strings or numbers).
var source = toRegexRange('15', '95');
//=> 1[5-9]|[2-8][0-9]|9[0-5]
var re = new RegExp('^' + source + '$');
console.log(re.test('14')); //=> false
console.log(re.test('50')); //=> true
console.log(re.test('94')); //=> true
console.log(re.test('96')); //=> false
Options
options.capture
Type: boolean
Deafault: undefined
Wrap the returned value in parentheses when there is more than one regex condition. Useful when you're dynamically generating ranges.
console.log(toRegexRange('-10', '10'));
//=> -[1-9]|-?10|[0-9]
console.log(toRegexRange('-10', '10', {capture: true}));
//=> (-[1-9]|-?10|[0-9])
options.shorthand
Type: boolean
Deafault: undefined
Use the regex shorthand for [0-9]
:
console.log(toRegexRange('0', '999999'));
//=> [0-9]|[1-9][0-9]{1,5}
console.log(toRegexRange('0', '999999', {shorthand: true}));
//=> \d|[1-9]\d{1,5}
options.relaxZeros
Type: boolean
Default: true
This option only applies to negative zero-padded ranges. By default, when a negative zero-padded range is defined, the number of leading zeros is relaxed using -0*
.
console.log(toRegexRange('-001', '100'));
//=> -0*1|0{2}[0-9]|0[1-9][0-9]|100
console.log(toRegexRange('-001', '100', {relaxZeros: false}));
//=> -0{2}1|0{2}[0-9]|0[1-9][0-9]|100
Why are zeros relaxed for negative zero-padded ranges by default?
Consider the following.
var regex = toRegexRange('-001', '100');
Note that -001
and 100
are both three digits long.
In most zero-padding implementations, only a single leading zero is enough to indicate that zero-padding should be applied. Thus, the leading zeros would be "corrected" on the negative range in the example to -01
, instead of -001
, to make total length of each string no greater than the length of the largest number in the range (in other words, -001
is 4 digits, but 100
is only three digits).
If zeros were not relaxed by default, you might expect the resulting regex of the above pattern to match -001
- given that it's defined that way in the arguments - but it wouldn't. It would, however, match -01
. This gets even more ambiguous with large ranges, like -01
to 1000000
.
Thus, we relax zeros by default to provide a more predictable experience for users.
Examples
Range | Result | Compile time |
---|---|---|
toRegexRange('5, 5') |
5 |
33μs |
toRegexRange('5, 6') |
5|6 |
53μs |
toRegexRange('29, 51') |
29|[34][0-9]|5[01] |
699μs |
toRegexRange('31, 877') |
3[1-9]|[4-9][0-9]|[1-7][0-9]{2}|8[0-6][0-9]|87[0-7] |
711μs |
toRegexRange('111, 555') |
11[1-9]|1[2-9][0-9]|[2-4][0-9]{2}|5[0-4][0-9]|55[0-5] |
62μs |
toRegexRange('-10, 10') |
-[1-9]|-?10|[0-9] |
74μs |
toRegexRange('-100, -10') |
-1[0-9]|-[2-9][0-9]|-100 |
49μs |
toRegexRange('-100, 100') |
-[1-9]|-?[1-9][0-9]|-?100|[0-9] |
45μs |
toRegexRange('001, 100') |
0{2}[1-9]|0[1-9][0-9]|100 |
158μs |
toRegexRange('0010, 1000') |
0{2}1[0-9]|0{2}[2-9][0-9]|0[1-9][0-9]{2}|1000 |
61μs |
toRegexRange('1, 2') |
1|2 |
10μs |
toRegexRange('1, 5') |
[1-5] |
24μs |
toRegexRange('1, 10') |
[1-9]|10 |
23μs |
toRegexRange('1, 100') |
[1-9]|[1-9][0-9]|100 |
30μs |
toRegexRange('1, 1000') |
[1-9]|[1-9][0-9]{1,2}|1000 |
52μs |
toRegexRange('1, 10000') |
[1-9]|[1-9][0-9]{1,3}|10000 |
47μs |
toRegexRange('1, 100000') |
[1-9]|[1-9][0-9]{1,4}|100000 |
44μs |
toRegexRange('1, 1000000') |
[1-9]|[1-9][0-9]{1,5}|1000000 |
49μs |
toRegexRange('1, 10000000') |
[1-9]|[1-9][0-9]{1,6}|10000000 |
63μs |
Heads up!
Order of arguments
When the min
is larger than the max
, values will be flipped to create a valid range:
toRegexRange('51', '29');
Is effectively flipped to:
toRegexRange('29', '51');
//=> 29|[3-4][0-9]|5[0-1]
Steps / increments
This library does not support steps (increments). A pr to add support would be welcome.
History
v2.0.0 - 2017-04-21
New features
Adds support for zero-padding!
v1.0.0
Optimizations
Repeating ranges are now grouped using quantifiers. rocessing time is roughly the same, but the generated regex is much smaller, which should result in faster matching.
Attribution
Inspired by the python library range-regex.
About
Related projects
- expand-range: Fast, bash-like range expansion. Expand a range of numbers or letters, uppercase or lowercase. See… more | homepage
- fill-range: Fill in a range of numbers or letters, optionally passing an increment or
step
to… more | homepage - micromatch: Glob matching for javascript/node.js. A drop-in replacement and faster alternative to minimatch and multimatch. | homepage
- repeat-element: Create an array by repeating the given value n times. | homepage
- repeat-string: Repeat the given string n times. Fastest implementation for repeating a string. | homepage
Contributing
Pull requests and stars are always welcome. For bugs and feature requests, please create an issue.
Building docs
(This project's readme.md is generated by verb, please don't edit the readme directly. Any changes to the readme must be made in the .verb.md readme template.)
To generate the readme, run the following command:
$ npm install -g verbose/verb#dev verb-generate-readme && verb
Running tests
Running and reviewing unit tests is a great way to get familiarized with a library and its API. You can install dependencies and run tests with the following command:
$ npm install && npm test
Author
Jon Schlinkert
License
Copyright © 2017, Jon Schlinkert. Released under the MIT License.
This file was generated by verb-generate-readme, v0.6.0, on April 27, 2017.