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