source: imaps-frontend/node_modules/eslint/lib/rules/no-misleading-character-class.js

main
Last change on this file was d565449, checked in by stefan toskovski <stefantoska84@…>, 4 weeks ago

Update repo after prototype presentation
  • Property mode set to 100644
File size: 9.9 KB
Line 
1/**
2 * @author Toru Nagashima <https://github.com/mysticatea>
3 */
4"use strict";
5
6const { CALL, CONSTRUCT, ReferenceTracker, getStringIfConstant } = require("@eslint-community/eslint-utils");
7const { RegExpParser, visitRegExpAST } = require("@eslint-community/regexpp");
8const { isCombiningCharacter, isEmojiModifier, isRegionalIndicatorSymbol, isSurrogatePair } = require("./utils/unicode");
9const astUtils = require("./utils/ast-utils.js");
10const { isValidWithUnicodeFlag } = require("./utils/regular-expressions");
11
12//------------------------------------------------------------------------------
13// Helpers
14//------------------------------------------------------------------------------
15
16/**
17 * @typedef {import('@eslint-community/regexpp').AST.Character} Character
18 * @typedef {import('@eslint-community/regexpp').AST.CharacterClassElement} CharacterClassElement
19 */
20
21/**
22 * Iterate character sequences of a given nodes.
23 *
24 * CharacterClassRange syntax can steal a part of character sequence,
25 * so this function reverts CharacterClassRange syntax and restore the sequence.
26 * @param {CharacterClassElement[]} nodes The node list to iterate character sequences.
27 * @returns {IterableIterator<Character[]>} The list of character sequences.
28 */
29function *iterateCharacterSequence(nodes) {
30
31 /** @type {Character[]} */
32 let seq = [];
33
34 for (const node of nodes) {
35 switch (node.type) {
36 case "Character":
37 seq.push(node);
38 break;
39
40 case "CharacterClassRange":
41 seq.push(node.min);
42 yield seq;
43 seq = [node.max];
44 break;
45
46 case "CharacterSet":
47 case "CharacterClass": // [[]] nesting character class
48 case "ClassStringDisjunction": // \q{...}
49 case "ExpressionCharacterClass": // [A--B]
50 if (seq.length > 0) {
51 yield seq;
52 seq = [];
53 }
54 break;
55
56 // no default
57 }
58 }
59
60 if (seq.length > 0) {
61 yield seq;
62 }
63}
64
65
66/**
67 * Checks whether the given character node is a Unicode code point escape or not.
68 * @param {Character} char the character node to check.
69 * @returns {boolean} `true` if the character node is a Unicode code point escape.
70 */
71function isUnicodeCodePointEscape(char) {
72 return /^\\u\{[\da-f]+\}$/iu.test(char.raw);
73}
74
75/**
76 * Each function returns `true` if it detects that kind of problem.
77 * @type {Record<string, (chars: Character[]) => boolean>}
78 */
79const hasCharacterSequence = {
80 surrogatePairWithoutUFlag(chars) {
81 return chars.some((c, i) => {
82 if (i === 0) {
83 return false;
84 }
85 const c1 = chars[i - 1];
86
87 return (
88 isSurrogatePair(c1.value, c.value) &&
89 !isUnicodeCodePointEscape(c1) &&
90 !isUnicodeCodePointEscape(c)
91 );
92 });
93 },
94
95 surrogatePair(chars) {
96 return chars.some((c, i) => {
97 if (i === 0) {
98 return false;
99 }
100 const c1 = chars[i - 1];
101
102 return (
103 isSurrogatePair(c1.value, c.value) &&
104 (
105 isUnicodeCodePointEscape(c1) ||
106 isUnicodeCodePointEscape(c)
107 )
108 );
109 });
110 },
111
112 combiningClass(chars) {
113 return chars.some((c, i) => (
114 i !== 0 &&
115 isCombiningCharacter(c.value) &&
116 !isCombiningCharacter(chars[i - 1].value)
117 ));
118 },
119
120 emojiModifier(chars) {
121 return chars.some((c, i) => (
122 i !== 0 &&
123 isEmojiModifier(c.value) &&
124 !isEmojiModifier(chars[i - 1].value)
125 ));
126 },
127
128 regionalIndicatorSymbol(chars) {
129 return chars.some((c, i) => (
130 i !== 0 &&
131 isRegionalIndicatorSymbol(c.value) &&
132 isRegionalIndicatorSymbol(chars[i - 1].value)
133 ));
134 },
135
136 zwj(chars) {
137 const lastIndex = chars.length - 1;
138
139 return chars.some((c, i) => (
140 i !== 0 &&
141 i !== lastIndex &&
142 c.value === 0x200d &&
143 chars[i - 1].value !== 0x200d &&
144 chars[i + 1].value !== 0x200d
145 ));
146 }
147};
148
149const kinds = Object.keys(hasCharacterSequence);
150
151//------------------------------------------------------------------------------
152// Rule Definition
153//------------------------------------------------------------------------------
154
155/** @type {import('../shared/types').Rule} */
156module.exports = {
157 meta: {
158 type: "problem",
159
160 docs: {
161 description: "Disallow characters which are made with multiple code points in character class syntax",
162 recommended: true,
163 url: "https://eslint.org/docs/latest/rules/no-misleading-character-class"
164 },
165
166 hasSuggestions: true,
167
168 schema: [],
169
170 messages: {
171 surrogatePairWithoutUFlag: "Unexpected surrogate pair in character class. Use 'u' flag.",
172 surrogatePair: "Unexpected surrogate pair in character class.",
173 combiningClass: "Unexpected combined character in character class.",
174 emojiModifier: "Unexpected modified Emoji in character class.",
175 regionalIndicatorSymbol: "Unexpected national flag in character class.",
176 zwj: "Unexpected joined character sequence in character class.",
177 suggestUnicodeFlag: "Add unicode 'u' flag to regex."
178 }
179 },
180 create(context) {
181 const sourceCode = context.sourceCode;
182 const parser = new RegExpParser();
183
184 /**
185 * Verify a given regular expression.
186 * @param {Node} node The node to report.
187 * @param {string} pattern The regular expression pattern to verify.
188 * @param {string} flags The flags of the regular expression.
189 * @param {Function} unicodeFixer Fixer for missing "u" flag.
190 * @returns {void}
191 */
192 function verify(node, pattern, flags, unicodeFixer) {
193 let patternNode;
194
195 try {
196 patternNode = parser.parsePattern(
197 pattern,
198 0,
199 pattern.length,
200 {
201 unicode: flags.includes("u"),
202 unicodeSets: flags.includes("v")
203 }
204 );
205 } catch {
206
207 // Ignore regular expressions with syntax errors
208 return;
209 }
210
211 const foundKinds = new Set();
212
213 visitRegExpAST(patternNode, {
214 onCharacterClassEnter(ccNode) {
215 for (const chars of iterateCharacterSequence(ccNode.elements)) {
216 for (const kind of kinds) {
217 if (hasCharacterSequence[kind](chars)) {
218 foundKinds.add(kind);
219 }
220 }
221 }
222 }
223 });
224
225 for (const kind of foundKinds) {
226 let suggest;
227
228 if (kind === "surrogatePairWithoutUFlag") {
229 suggest = [{
230 messageId: "suggestUnicodeFlag",
231 fix: unicodeFixer
232 }];
233 }
234
235 context.report({
236 node,
237 messageId: kind,
238 suggest
239 });
240 }
241 }
242
243 return {
244 "Literal[regex]"(node) {
245 verify(node, node.regex.pattern, node.regex.flags, fixer => {
246 if (!isValidWithUnicodeFlag(context.languageOptions.ecmaVersion, node.regex.pattern)) {
247 return null;
248 }
249
250 return fixer.insertTextAfter(node, "u");
251 });
252 },
253 "Program"(node) {
254 const scope = sourceCode.getScope(node);
255 const tracker = new ReferenceTracker(scope);
256
257 /*
258 * Iterate calls of RegExp.
259 * E.g., `new RegExp()`, `RegExp()`, `new window.RegExp()`,
260 * `const {RegExp: a} = window; new a()`, etc...
261 */
262 for (const { node: refNode } of tracker.iterateGlobalReferences({
263 RegExp: { [CALL]: true, [CONSTRUCT]: true }
264 })) {
265 const [patternNode, flagsNode] = refNode.arguments;
266 const pattern = getStringIfConstant(patternNode, scope);
267 const flags = getStringIfConstant(flagsNode, scope);
268
269 if (typeof pattern === "string") {
270 verify(refNode, pattern, flags || "", fixer => {
271
272 if (!isValidWithUnicodeFlag(context.languageOptions.ecmaVersion, pattern)) {
273 return null;
274 }
275
276 if (refNode.arguments.length === 1) {
277 const penultimateToken = sourceCode.getLastToken(refNode, { skip: 1 }); // skip closing parenthesis
278
279 return fixer.insertTextAfter(
280 penultimateToken,
281 astUtils.isCommaToken(penultimateToken)
282 ? ' "u",'
283 : ', "u"'
284 );
285 }
286
287 if ((flagsNode.type === "Literal" && typeof flagsNode.value === "string") || flagsNode.type === "TemplateLiteral") {
288 const range = [flagsNode.range[0], flagsNode.range[1] - 1];
289
290 return fixer.insertTextAfterRange(range, "u");
291 }
292
293 return null;
294 });
295 }
296 }
297 }
298 };
299 }
300};
Note: See TracBrowser for help on using the repository browser.