| [d24f17c] | 1 | (function (Prism) {
|
|---|
| 2 |
|
|---|
| 3 | var templateString = Prism.languages.javascript['template-string'];
|
|---|
| 4 |
|
|---|
| 5 | // see the pattern in prism-javascript.js
|
|---|
| 6 | var templateLiteralPattern = templateString.pattern.source;
|
|---|
| 7 | var interpolationObject = templateString.inside['interpolation'];
|
|---|
| 8 | var interpolationPunctuationObject = interpolationObject.inside['interpolation-punctuation'];
|
|---|
| 9 | var interpolationPattern = interpolationObject.pattern.source;
|
|---|
| 10 |
|
|---|
| 11 |
|
|---|
| 12 | /**
|
|---|
| 13 | * Creates a new pattern to match a template string with a special tag.
|
|---|
| 14 | *
|
|---|
| 15 | * This will return `undefined` if there is no grammar with the given language id.
|
|---|
| 16 | *
|
|---|
| 17 | * @param {string} language The language id of the embedded language. E.g. `markdown`.
|
|---|
| 18 | * @param {string} tag The regex pattern to match the tag.
|
|---|
| 19 | * @returns {object | undefined}
|
|---|
| 20 | * @example
|
|---|
| 21 | * createTemplate('css', /\bcss/.source);
|
|---|
| 22 | */
|
|---|
| 23 | function createTemplate(language, tag) {
|
|---|
| 24 | if (!Prism.languages[language]) {
|
|---|
| 25 | return undefined;
|
|---|
| 26 | }
|
|---|
| 27 |
|
|---|
| 28 | return {
|
|---|
| 29 | pattern: RegExp('((?:' + tag + ')\\s*)' + templateLiteralPattern),
|
|---|
| 30 | lookbehind: true,
|
|---|
| 31 | greedy: true,
|
|---|
| 32 | inside: {
|
|---|
| 33 | 'template-punctuation': {
|
|---|
| 34 | pattern: /^`|`$/,
|
|---|
| 35 | alias: 'string'
|
|---|
| 36 | },
|
|---|
| 37 | 'embedded-code': {
|
|---|
| 38 | pattern: /[\s\S]+/,
|
|---|
| 39 | alias: language
|
|---|
| 40 | }
|
|---|
| 41 | }
|
|---|
| 42 | };
|
|---|
| 43 | }
|
|---|
| 44 |
|
|---|
| 45 |
|
|---|
| 46 | Prism.languages.javascript['template-string'] = [
|
|---|
| 47 | // styled-jsx:
|
|---|
| 48 | // css`a { color: #25F; }`
|
|---|
| 49 | // styled-components:
|
|---|
| 50 | // styled.h1`color: red;`
|
|---|
| 51 | createTemplate('css', /\b(?:styled(?:\([^)]*\))?(?:\s*\.\s*\w+(?:\([^)]*\))*)*|css(?:\s*\.\s*(?:global|resolve))?|createGlobalStyle|keyframes)/.source),
|
|---|
| 52 |
|
|---|
| 53 | // html`<p></p>`
|
|---|
| 54 | // div.innerHTML = `<p></p>`
|
|---|
| 55 | createTemplate('html', /\bhtml|\.\s*(?:inner|outer)HTML\s*\+?=/.source),
|
|---|
| 56 |
|
|---|
| 57 | // svg`<path fill="#fff" d="M55.37 ..."/>`
|
|---|
| 58 | createTemplate('svg', /\bsvg/.source),
|
|---|
| 59 |
|
|---|
| 60 | // md`# h1`, markdown`## h2`
|
|---|
| 61 | createTemplate('markdown', /\b(?:markdown|md)/.source),
|
|---|
| 62 |
|
|---|
| 63 | // gql`...`, graphql`...`, graphql.experimental`...`
|
|---|
| 64 | createTemplate('graphql', /\b(?:gql|graphql(?:\s*\.\s*experimental)?)/.source),
|
|---|
| 65 |
|
|---|
| 66 | // sql`...`
|
|---|
| 67 | createTemplate('sql', /\bsql/.source),
|
|---|
| 68 |
|
|---|
| 69 | // vanilla template string
|
|---|
| 70 | templateString
|
|---|
| 71 | ].filter(Boolean);
|
|---|
| 72 |
|
|---|
| 73 |
|
|---|
| 74 | /**
|
|---|
| 75 | * Returns a specific placeholder literal for the given language.
|
|---|
| 76 | *
|
|---|
| 77 | * @param {number} counter
|
|---|
| 78 | * @param {string} language
|
|---|
| 79 | * @returns {string}
|
|---|
| 80 | */
|
|---|
| 81 | function getPlaceholder(counter, language) {
|
|---|
| 82 | return '___' + language.toUpperCase() + '_' + counter + '___';
|
|---|
| 83 | }
|
|---|
| 84 |
|
|---|
| 85 | /**
|
|---|
| 86 | * Returns the tokens of `Prism.tokenize` but also runs the `before-tokenize` and `after-tokenize` hooks.
|
|---|
| 87 | *
|
|---|
| 88 | * @param {string} code
|
|---|
| 89 | * @param {any} grammar
|
|---|
| 90 | * @param {string} language
|
|---|
| 91 | * @returns {(string|Token)[]}
|
|---|
| 92 | */
|
|---|
| 93 | function tokenizeWithHooks(code, grammar, language) {
|
|---|
| 94 | var env = {
|
|---|
| 95 | code: code,
|
|---|
| 96 | grammar: grammar,
|
|---|
| 97 | language: language
|
|---|
| 98 | };
|
|---|
| 99 | Prism.hooks.run('before-tokenize', env);
|
|---|
| 100 | env.tokens = Prism.tokenize(env.code, env.grammar);
|
|---|
| 101 | Prism.hooks.run('after-tokenize', env);
|
|---|
| 102 | return env.tokens;
|
|---|
| 103 | }
|
|---|
| 104 |
|
|---|
| 105 | /**
|
|---|
| 106 | * Returns the token of the given JavaScript interpolation expression.
|
|---|
| 107 | *
|
|---|
| 108 | * @param {string} expression The code of the expression. E.g. `"${42}"`
|
|---|
| 109 | * @returns {Token}
|
|---|
| 110 | */
|
|---|
| 111 | function tokenizeInterpolationExpression(expression) {
|
|---|
| 112 | var tempGrammar = {};
|
|---|
| 113 | tempGrammar['interpolation-punctuation'] = interpolationPunctuationObject;
|
|---|
| 114 |
|
|---|
| 115 | /** @type {Array} */
|
|---|
| 116 | var tokens = Prism.tokenize(expression, tempGrammar);
|
|---|
| 117 | if (tokens.length === 3) {
|
|---|
| 118 | /**
|
|---|
| 119 | * The token array will look like this
|
|---|
| 120 | * [
|
|---|
| 121 | * ["interpolation-punctuation", "${"]
|
|---|
| 122 | * "..." // JavaScript expression of the interpolation
|
|---|
| 123 | * ["interpolation-punctuation", "}"]
|
|---|
| 124 | * ]
|
|---|
| 125 | */
|
|---|
| 126 |
|
|---|
| 127 | var args = [1, 1];
|
|---|
| 128 | args.push.apply(args, tokenizeWithHooks(tokens[1], Prism.languages.javascript, 'javascript'));
|
|---|
| 129 |
|
|---|
| 130 | tokens.splice.apply(tokens, args);
|
|---|
| 131 | }
|
|---|
| 132 |
|
|---|
| 133 | return new Prism.Token('interpolation', tokens, interpolationObject.alias, expression);
|
|---|
| 134 | }
|
|---|
| 135 |
|
|---|
| 136 | /**
|
|---|
| 137 | * Tokenizes the given code with support for JavaScript interpolation expressions mixed in.
|
|---|
| 138 | *
|
|---|
| 139 | * This function has 3 phases:
|
|---|
| 140 | *
|
|---|
| 141 | * 1. Replace all JavaScript interpolation expression with a placeholder.
|
|---|
| 142 | * The placeholder will have the syntax of a identify of the target language.
|
|---|
| 143 | * 2. Tokenize the code with placeholders.
|
|---|
| 144 | * 3. Tokenize the interpolation expressions and re-insert them into the tokenize code.
|
|---|
| 145 | * The insertion only works if a placeholder hasn't been "ripped apart" meaning that the placeholder has been
|
|---|
| 146 | * tokenized as two tokens by the grammar of the embedded language.
|
|---|
| 147 | *
|
|---|
| 148 | * @param {string} code
|
|---|
| 149 | * @param {object} grammar
|
|---|
| 150 | * @param {string} language
|
|---|
| 151 | * @returns {Token}
|
|---|
| 152 | */
|
|---|
| 153 | function tokenizeEmbedded(code, grammar, language) {
|
|---|
| 154 | // 1. First filter out all interpolations
|
|---|
| 155 |
|
|---|
| 156 | // because they might be escaped, we need a lookbehind, so we use Prism
|
|---|
| 157 | /** @type {(Token|string)[]} */
|
|---|
| 158 | var _tokens = Prism.tokenize(code, {
|
|---|
| 159 | 'interpolation': {
|
|---|
| 160 | pattern: RegExp(interpolationPattern),
|
|---|
| 161 | lookbehind: true
|
|---|
| 162 | }
|
|---|
| 163 | });
|
|---|
| 164 |
|
|---|
| 165 | // replace all interpolations with a placeholder which is not in the code already
|
|---|
| 166 | var placeholderCounter = 0;
|
|---|
| 167 | /** @type {Object<string, string>} */
|
|---|
| 168 | var placeholderMap = {};
|
|---|
| 169 | var embeddedCode = _tokens.map(function (token) {
|
|---|
| 170 | if (typeof token === 'string') {
|
|---|
| 171 | return token;
|
|---|
| 172 | } else {
|
|---|
| 173 | var interpolationExpression = token.content;
|
|---|
| 174 |
|
|---|
| 175 | var placeholder;
|
|---|
| 176 | while (code.indexOf(placeholder = getPlaceholder(placeholderCounter++, language)) !== -1) { /* noop */ }
|
|---|
| 177 | placeholderMap[placeholder] = interpolationExpression;
|
|---|
| 178 | return placeholder;
|
|---|
| 179 | }
|
|---|
| 180 | }).join('');
|
|---|
| 181 |
|
|---|
| 182 |
|
|---|
| 183 | // 2. Tokenize the embedded code
|
|---|
| 184 |
|
|---|
| 185 | var embeddedTokens = tokenizeWithHooks(embeddedCode, grammar, language);
|
|---|
| 186 |
|
|---|
| 187 |
|
|---|
| 188 | // 3. Re-insert the interpolation
|
|---|
| 189 |
|
|---|
| 190 | var placeholders = Object.keys(placeholderMap);
|
|---|
| 191 | placeholderCounter = 0;
|
|---|
| 192 |
|
|---|
| 193 | /**
|
|---|
| 194 | *
|
|---|
| 195 | * @param {(Token|string)[]} tokens
|
|---|
| 196 | * @returns {void}
|
|---|
| 197 | */
|
|---|
| 198 | function walkTokens(tokens) {
|
|---|
| 199 | for (var i = 0; i < tokens.length; i++) {
|
|---|
| 200 | if (placeholderCounter >= placeholders.length) {
|
|---|
| 201 | return;
|
|---|
| 202 | }
|
|---|
| 203 |
|
|---|
| 204 | var token = tokens[i];
|
|---|
| 205 |
|
|---|
| 206 | if (typeof token === 'string' || typeof token.content === 'string') {
|
|---|
| 207 | var placeholder = placeholders[placeholderCounter];
|
|---|
| 208 | var s = typeof token === 'string' ? token : /** @type {string} */ (token.content);
|
|---|
| 209 |
|
|---|
| 210 | var index = s.indexOf(placeholder);
|
|---|
| 211 | if (index !== -1) {
|
|---|
| 212 | ++placeholderCounter;
|
|---|
| 213 |
|
|---|
| 214 | var before = s.substring(0, index);
|
|---|
| 215 | var middle = tokenizeInterpolationExpression(placeholderMap[placeholder]);
|
|---|
| 216 | var after = s.substring(index + placeholder.length);
|
|---|
| 217 |
|
|---|
| 218 | var replacement = [];
|
|---|
| 219 | if (before) {
|
|---|
| 220 | replacement.push(before);
|
|---|
| 221 | }
|
|---|
| 222 | replacement.push(middle);
|
|---|
| 223 | if (after) {
|
|---|
| 224 | var afterTokens = [after];
|
|---|
| 225 | walkTokens(afterTokens);
|
|---|
| 226 | replacement.push.apply(replacement, afterTokens);
|
|---|
| 227 | }
|
|---|
| 228 |
|
|---|
| 229 | if (typeof token === 'string') {
|
|---|
| 230 | tokens.splice.apply(tokens, [i, 1].concat(replacement));
|
|---|
| 231 | i += replacement.length - 1;
|
|---|
| 232 | } else {
|
|---|
| 233 | token.content = replacement;
|
|---|
| 234 | }
|
|---|
| 235 | }
|
|---|
| 236 | } else {
|
|---|
| 237 | var content = token.content;
|
|---|
| 238 | if (Array.isArray(content)) {
|
|---|
| 239 | walkTokens(content);
|
|---|
| 240 | } else {
|
|---|
| 241 | walkTokens([content]);
|
|---|
| 242 | }
|
|---|
| 243 | }
|
|---|
| 244 | }
|
|---|
| 245 | }
|
|---|
| 246 | walkTokens(embeddedTokens);
|
|---|
| 247 |
|
|---|
| 248 | return new Prism.Token(language, embeddedTokens, 'language-' + language, code);
|
|---|
| 249 | }
|
|---|
| 250 |
|
|---|
| 251 | /**
|
|---|
| 252 | * The languages for which JS templating will handle tagged template literals.
|
|---|
| 253 | *
|
|---|
| 254 | * JS templating isn't active for only JavaScript but also related languages like TypeScript, JSX, and TSX.
|
|---|
| 255 | */
|
|---|
| 256 | var supportedLanguages = {
|
|---|
| 257 | 'javascript': true,
|
|---|
| 258 | 'js': true,
|
|---|
| 259 | 'typescript': true,
|
|---|
| 260 | 'ts': true,
|
|---|
| 261 | 'jsx': true,
|
|---|
| 262 | 'tsx': true,
|
|---|
| 263 | };
|
|---|
| 264 | Prism.hooks.add('after-tokenize', function (env) {
|
|---|
| 265 | if (!(env.language in supportedLanguages)) {
|
|---|
| 266 | return;
|
|---|
| 267 | }
|
|---|
| 268 |
|
|---|
| 269 | /**
|
|---|
| 270 | * Finds and tokenizes all template strings with an embedded languages.
|
|---|
| 271 | *
|
|---|
| 272 | * @param {(Token | string)[]} tokens
|
|---|
| 273 | * @returns {void}
|
|---|
| 274 | */
|
|---|
| 275 | function findTemplateStrings(tokens) {
|
|---|
| 276 | for (var i = 0, l = tokens.length; i < l; i++) {
|
|---|
| 277 | var token = tokens[i];
|
|---|
| 278 |
|
|---|
| 279 | if (typeof token === 'string') {
|
|---|
| 280 | continue;
|
|---|
| 281 | }
|
|---|
| 282 |
|
|---|
| 283 | var content = token.content;
|
|---|
| 284 | if (!Array.isArray(content)) {
|
|---|
| 285 | if (typeof content !== 'string') {
|
|---|
| 286 | findTemplateStrings([content]);
|
|---|
| 287 | }
|
|---|
| 288 | continue;
|
|---|
| 289 | }
|
|---|
| 290 |
|
|---|
| 291 | if (token.type === 'template-string') {
|
|---|
| 292 | /**
|
|---|
| 293 | * A JavaScript template-string token will look like this:
|
|---|
| 294 | *
|
|---|
| 295 | * ["template-string", [
|
|---|
| 296 | * ["template-punctuation", "`"],
|
|---|
| 297 | * (
|
|---|
| 298 | * An array of "string" and "interpolation" tokens. This is the simple string case.
|
|---|
| 299 | * or
|
|---|
| 300 | * ["embedded-code", "..."] This is the token containing the embedded code.
|
|---|
| 301 | * It also has an alias which is the language of the embedded code.
|
|---|
| 302 | * ),
|
|---|
| 303 | * ["template-punctuation", "`"]
|
|---|
| 304 | * ]]
|
|---|
| 305 | */
|
|---|
| 306 |
|
|---|
| 307 | var embedded = content[1];
|
|---|
| 308 | if (content.length === 3 && typeof embedded !== 'string' && embedded.type === 'embedded-code') {
|
|---|
| 309 | // get string content
|
|---|
| 310 | var code = stringContent(embedded);
|
|---|
| 311 |
|
|---|
| 312 | var alias = embedded.alias;
|
|---|
| 313 | var language = Array.isArray(alias) ? alias[0] : alias;
|
|---|
| 314 |
|
|---|
| 315 | var grammar = Prism.languages[language];
|
|---|
| 316 | if (!grammar) {
|
|---|
| 317 | // the embedded language isn't registered.
|
|---|
| 318 | continue;
|
|---|
| 319 | }
|
|---|
| 320 |
|
|---|
| 321 | content[1] = tokenizeEmbedded(code, grammar, language);
|
|---|
| 322 | }
|
|---|
| 323 | } else {
|
|---|
| 324 | findTemplateStrings(content);
|
|---|
| 325 | }
|
|---|
| 326 | }
|
|---|
| 327 | }
|
|---|
| 328 |
|
|---|
| 329 | findTemplateStrings(env.tokens);
|
|---|
| 330 | });
|
|---|
| 331 |
|
|---|
| 332 |
|
|---|
| 333 | /**
|
|---|
| 334 | * Returns the string content of a token or token stream.
|
|---|
| 335 | *
|
|---|
| 336 | * @param {string | Token | (string | Token)[]} value
|
|---|
| 337 | * @returns {string}
|
|---|
| 338 | */
|
|---|
| 339 | function stringContent(value) {
|
|---|
| 340 | if (typeof value === 'string') {
|
|---|
| 341 | return value;
|
|---|
| 342 | } else if (Array.isArray(value)) {
|
|---|
| 343 | return value.map(stringContent).join('');
|
|---|
| 344 | } else {
|
|---|
| 345 | return stringContent(value.content);
|
|---|
| 346 | }
|
|---|
| 347 | }
|
|---|
| 348 |
|
|---|
| 349 | }(Prism));
|
|---|
