rewrite-pattern.js 9.58 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338
'use strict';

const generate = require('regjsgen').generate;
const parse = require('regjsparser').parse;
const regenerate = require('regenerate');
const unicodeMatchProperty = require('unicode-match-property-ecmascript');
const unicodeMatchPropertyValue = require('unicode-match-property-value-ecmascript');
const iuMappings = require('./data/iu-mappings.js');
const ESCAPE_SETS = require('./data/character-class-escape-sets.js');

// Prepare a Regenerate set containing all code points, used for negative
// character classes (if any).
const UNICODE_SET = regenerate().addRange(0x0, 0x10FFFF);
// Without the `u` flag, the range stops at 0xFFFF.
// https://mths.be/es6#sec-pattern-semantics
const BMP_SET = regenerate().addRange(0x0, 0xFFFF);

// Prepare a Regenerate set containing all code points that are supposed to be
// matched by `/./u`. https://mths.be/es6#sec-atom
const DOT_SET_UNICODE = UNICODE_SET.clone() // all Unicode code points
	.remove(
		// minus `LineTerminator`s (https://mths.be/es6#sec-line-terminators):
		0x000A, // Line Feed <LF>
		0x000D, // Carriage Return <CR>
		0x2028, // Line Separator <LS>
		0x2029  // Paragraph Separator <PS>
	);

const getCharacterClassEscapeSet = (character, unicode, ignoreCase) => {
	if (unicode) {
		if (ignoreCase) {
			return ESCAPE_SETS.UNICODE_IGNORE_CASE.get(character);
		}
		return ESCAPE_SETS.UNICODE.get(character);
	}
	return ESCAPE_SETS.REGULAR.get(character);
};

const getUnicodeDotSet = (dotAll) => {
	return dotAll ? UNICODE_SET : DOT_SET_UNICODE;
};

const getUnicodePropertyValueSet = (property, value) => {
	const path = value ?
		`${ property }/${ value }` :
		`Binary_Property/${ property }`;
	try {
		return require(`regenerate-unicode-properties/${ path }.js`);
	} catch (exception) {
		throw new Error(
			`Failed to recognize value \`${ value }\` for property ` +
			`\`${ property }\`.`
		);
	}
};

const handleLoneUnicodePropertyNameOrValue = (value) => {
	// It could be a `General_Category` value or a binary property.
	// Note: `unicodeMatchPropertyValue` throws on invalid values.
	try {
		const property = 'General_Category';
		const category = unicodeMatchPropertyValue(property, value);
		return getUnicodePropertyValueSet(property, category);
	} catch (exception) {}
	// It’s not a `General_Category` value, so check if it’s a binary
	// property. Note: `unicodeMatchProperty` throws on invalid properties.
	const property = unicodeMatchProperty(value);
	return getUnicodePropertyValueSet(property);
};

const getUnicodePropertyEscapeSet = (value, isNegative) => {
	const parts = value.split('=');
	const firstPart = parts[0];
	let set;
	if (parts.length == 1) {
		set = handleLoneUnicodePropertyNameOrValue(firstPart);
	} else {
		// The pattern consists of two parts, i.e. `Property=Value`.
		const property = unicodeMatchProperty(firstPart);
		const value = unicodeMatchPropertyValue(property, parts[1]);
		set = getUnicodePropertyValueSet(property, value);
	}
	if (isNegative) {
		return UNICODE_SET.clone().remove(set);
	}
	return set.clone();
};

// Given a range of code points, add any case-folded code points in that range
// to a set.
regenerate.prototype.iuAddRange = function(min, max) {
	const $this = this;
	do {
		const folded = caseFold(min);
		if (folded) {
			$this.add(folded);
		}
	} while (++min <= max);
	return $this;
};

const update = (item, pattern) => {
	let tree = parse(pattern, config.useUnicodeFlag ? 'u' : '');
	switch (tree.type) {
		case 'characterClass':
		case 'group':
		case 'value':
			// No wrapping needed.
			break;
		default:
			// Wrap the pattern in a non-capturing group.
			tree = wrap(tree, pattern);
	}
	Object.assign(item, tree);
};

const wrap = (tree, pattern) => {
	// Wrap the pattern in a non-capturing group.
	return {
		'type': 'group',
		'behavior': 'ignore',
		'body': [tree],
		'raw': `(?:${ pattern })`
	};
};

const caseFold = (codePoint) => {
	return iuMappings.get(codePoint) || false;
};

const processCharacterClass = (characterClassItem, regenerateOptions) => {
	let set = regenerate();
	for (const item of characterClassItem.body) {
		switch (item.type) {
			case 'value':
				set.add(item.codePoint);
				if (config.ignoreCase && config.unicode && !config.useUnicodeFlag) {
					const folded = caseFold(item.codePoint);
					if (folded) {
						set.add(folded);
					}
				}
				break;
			case 'characterClassRange':
				const min = item.min.codePoint;
				const max = item.max.codePoint;
				set.addRange(min, max);
				if (config.ignoreCase && config.unicode && !config.useUnicodeFlag) {
					set.iuAddRange(min, max);
				}
				break;
			case 'characterClassEscape':
				set.add(getCharacterClassEscapeSet(
					item.value,
					config.unicode,
					config.ignoreCase
				));
				break;
			case 'unicodePropertyEscape':
				set.add(getUnicodePropertyEscapeSet(item.value, item.negative));
				break;
			// The `default` clause is only here as a safeguard; it should never be
			// reached. Code coverage tools should ignore it.
			/* istanbul ignore next */
			default:
				throw new Error(`Unknown term type: ${ item.type }`);
		}
	}
	if (characterClassItem.negative) {
		set = (config.unicode ? UNICODE_SET : BMP_SET).clone().remove(set);
	}
	update(characterClassItem, set.toString(regenerateOptions));
	return characterClassItem;
};

const updateNamedReference = (item, index) => {
	delete item.name;
	item.matchIndex = index;
};

const assertNoUnmatchedReferences = (groups) => {
	const unmatchedReferencesNames = Object.keys(groups.unmatchedReferences);
	if (unmatchedReferencesNames.length > 0) {
		throw new Error(`Unknown group names: ${unmatchedReferencesNames}`);
	}
};

const processTerm = (item, regenerateOptions, groups) => {
	switch (item.type) {
		case 'dot':
			if (config.unicode) {
				update(
					item,
					getUnicodeDotSet(config.dotAll).toString(regenerateOptions)
				);
			} else if (config.dotAll) {
				// TODO: consider changing this at the regenerate level.
				update(item, '[\\s\\S]');
			}
			break;
		case 'characterClass':
			item = processCharacterClass(item, regenerateOptions);
			break;
		case 'unicodePropertyEscape':
			if (config.unicodePropertyEscape) {
				update(
					item,
					getUnicodePropertyEscapeSet(item.value, item.negative)
						.toString(regenerateOptions)
				);
			}
			break;
		case 'characterClassEscape':
			update(
				item,
				getCharacterClassEscapeSet(
					item.value,
					config.unicode,
					config.ignoreCase
				).toString(regenerateOptions)
			);
			break;
		case 'group':
			if (item.behavior == 'normal') {
				groups.lastIndex++;
			}
			if (item.name && config.namedGroup) {
				const name = item.name.value;

				if (groups.names[name]) {
					throw new Error(
						`Multiple groups with the same name (${ name }) are not allowed.`
					);
				}

				const index = groups.lastIndex;
				delete item.name;

				groups.names[name] = index;
				if (groups.onNamedGroup) {
					groups.onNamedGroup.call(null, name, index);
				}

				if (groups.unmatchedReferences[name]) {
					groups.unmatchedReferences[name].forEach(reference => {
						updateNamedReference(reference, index);
					});
					delete groups.unmatchedReferences[name];
				}
			}
			/* falls through */
		case 'alternative':
		case 'disjunction':
		case 'quantifier':
			item.body = item.body.map(term => {
				return processTerm(term, regenerateOptions, groups);
			});
			break;
		case 'value':
			const codePoint = item.codePoint;
			const set = regenerate(codePoint);
			if (config.ignoreCase && config.unicode && !config.useUnicodeFlag) {
				const folded = caseFold(codePoint);
				if (folded) {
					set.add(folded);
				}
			}
			update(item, set.toString(regenerateOptions));
			break;
		case 'reference':
			if (item.name) {
				const name = item.name.value;
				const index = groups.names[name];
				if (index) {
					updateNamedReference(item, index);
					break;
				}

				if (!groups.unmatchedReferences[name]) {
					groups.unmatchedReferences[name] = [];
				}
				// Keep track of references used before the corresponding group.
				groups.unmatchedReferences[name].push(item);
			}
			break;
		case 'anchor':
		case 'empty':
		case 'group':
			// Nothing to do here.
			break;
		// The `default` clause is only here as a safeguard; it should never be
		// reached. Code coverage tools should ignore it.
		/* istanbul ignore next */
		default:
			throw new Error(`Unknown term type: ${ item.type }`);
	}
	return item;
};

const config = {
	'ignoreCase': false,
	'unicode': false,
	'dotAll': false,
	'useUnicodeFlag': false,
	'unicodePropertyEscape': false,
	'namedGroup': false
};
const rewritePattern = (pattern, flags, options) => {
	config.unicode = flags && flags.includes('u');
	const regjsparserFeatures = {
		'unicodePropertyEscape': config.unicode,
		'namedGroups': true,
		'lookbehind': options && options.lookbehind
	};
	config.ignoreCase = flags && flags.includes('i');
	const supportDotAllFlag = options && options.dotAllFlag;
	config.dotAll = supportDotAllFlag && flags && flags.includes('s');
	config.namedGroup = options && options.namedGroup;
	config.useUnicodeFlag = options && options.useUnicodeFlag;
	config.unicodePropertyEscape = options && options.unicodePropertyEscape;
	const regenerateOptions = {
		'hasUnicodeFlag': config.useUnicodeFlag,
		'bmpOnly': !config.unicode
	};
	const groups = {
		'onNamedGroup': options && options.onNamedGroup,
		'lastIndex': 0,
		'names': Object.create(null), // { [name]: index }
		'unmatchedReferences': Object.create(null) // { [name]: Array<reference> }
	};
	const tree = parse(pattern, flags, regjsparserFeatures);
	// Note: `processTerm` mutates `tree` and `groups`.
	processTerm(tree, regenerateOptions, groups);
	assertNoUnmatchedReferences(groups);
	return generate(tree);
};

module.exports = rewritePattern;