Модуль:Pegex/doc
Это страница документации Модуль:Pegex.
Pegex — модуль, преобразующий регулярное выражение в стиле Perl (с ограничениями) в LPeg.
Автор — Chris Emerson [email protected], 2014.
Источник — https://github.com/jugglerchris/ta-regex/blob/master/pegex.lua.
Алгоритм разработан Роберту Иерусалимски, Марсело Оикава и Анной Лусией де Моура (Marcelo Oikawa, Roberto Ierusalimschy, Anna Lúcia de Moura «Converting regexes to Parsing Expression Grammars» // Departamento de Informática, PUC-Rio.).
Порядок вызова из модулей «Традиции»:
local pat = require 'Module:pegex'.compile '(?:foo|bar)+' -- это грамматика LPEG.
local result1 = pat:match 'asdfoo' -- возвратит { _start=4, _end=6}.
local result2 = pat:match 'asdf' -- возвратит nil (образец не найден).
ta-regex/Pegex (оригинальная страница справки)
Pegex is a regular expression (regexp) implementation built on top of LPeg.
The original motivation was to add regular expression search support for the Textadept editor; however the underlying engine is generic.
This module replaces the default text search with one which uses regular expressions.
Currently the full regular expressions are supported (not including eg Perl extensions, though some are planned); this is more than the subset supported natively in Textadept (which eg don't include "|").
| Syntax | Matches |
|---|---|
| . | Any character except newline |
| [abA-Z] | The characters a,b, or any capital letter |
| \< | Zero-length, matches just before the start of a word |
| \> | Zero-length, matches just after the end of a word |
| bar | Match foo or bar |
| (pat) | Match the same as pat (subgroup) |
| (?:pat) | Match the same as pat (subgroup), non-capturing |
| x* | Match zero or more x |
| x+ | match one or more x |
| x? | match zero or one x |
| \x | . : match the character x |
| \w | Any "word" character [a-zA-Z_] |
| \W | Any non-"word" character [^a-zA-Z_] |
| \d | Any digit character [0-9] |
| \D | Any non-digit character [^0-9] |
| \s | Any whitespace character [ \t\n\v\r] |
| \S | Any non-whitespace character [^\t\n\v\r] |
| \1 … \9 | Back reference to Nth (subgroup) |
Installation
To install Pegex standalone, use "luarocks install pegex". Example usage:
local pegex = require('pegex')
pat = pegex.compile('(?:foo|bar)+')
result = pat:match("asdfoo") -- returns { _start=4, _end=6}
result = pat:match("asdf") -- returns nil (not found)
See the tests for examples using captures and backreferences.
To use with Textadept to replace the default search method:
Add the ta-regex directory to ~/.textadept/modules/
Add the following line to ~/.textadept/init.lua
local ta_regex = require 'ta-regex'
ta_regex.install()
Internal details
The module adds a handler for events.FIND to intercept searches. Regular expressions are converted to equivalent LPEG patterns, which are then used for searching the text.
The regex-to-LPEG conversion can be used independently.