| 1 |
// The keywords object is set to the result of this self executing |
| 2 |
// function. Each keyword is a property of the keywords object whose |
| 3 |
// value is {type: atype, style: astyle} |
| 4 |
var keywords = function(){ |
| 5 |
// convenience functions used to build keywords object |
| 6 |
function kw(type) {return {type: type, style: "keyword"};} |
| 7 |
var operator = kw("operator") |
| 8 |
, atom = {type: "atom", style: "atom"} |
| 9 |
, punctuation = {type: "punctuation", style: null} |
| 10 |
, qualifier = {type: "axis_specifier", style: "qualifier"}; |
| 11 |
|
| 12 |
// kwObj is what is return from this function at the end |
| 13 |
var kwObj = { |
| 14 |
',': punctuation |
| 15 |
}; |
| 16 |
|
| 17 |
// a list of 'basic' keywords. For each add a property to kwObj with the value of |
| 18 |
// {type: basic[i], style: "keyword"} e.g. 'after' --> {type: "after", style: "keyword"} |
| 19 |
var basic = ['after', 'all', 'allowing', 'ancestor', 'ancestor-or-self', 'any', 'array', 'as', |
| 20 |
'ascending', 'at', 'attribute', 'base-uri', 'before', 'boundary-space', 'by', 'case', 'cast', |
| 21 |
'castable', 'catch', 'child', 'collation', 'comment', 'construction', 'contains', 'content', |
| 22 |
'context', 'copy', 'copy-namespaces', 'count', 'decimal-format', 'declare', 'default', 'delete', |
| 23 |
'descendant', 'descendant-or-self', 'descending', 'diacritics', 'different', 'distance', |
| 24 |
'document', 'document-node', 'element', 'else', 'empty', 'empty-sequence', 'encoding', 'end', |
| 25 |
'entire', 'every', 'exactly', 'except', 'external', 'first', 'following', 'following-sibling', |
| 26 |
'for', 'from', 'ftand', 'ftnot', 'ft-option', 'ftor', 'function', 'fuzzy', 'greatest', 'group', |
| 27 |
'if', 'import', 'in', 'inherit', 'insensitive', 'insert', 'instance', 'intersect', 'into', |
| 28 |
'invoke', 'is', 'item', 'language', 'last', 'lax', 'least', 'let', 'levels', 'lowercase', 'map', |
| 29 |
'modify', 'module', 'most', 'namespace', 'next', 'no', 'node', 'nodes', 'no-inherit', |
| 30 |
'no-preserve', 'not', 'occurs', 'of', 'only', 'option', 'order', 'ordered', 'ordering', |
| 31 |
'paragraph', 'paragraphs', 'parent', 'phrase', 'preceding', 'preceding-sibling', 'preserve', |
| 32 |
'previous', 'processing-instruction', 'relationship', 'rename', 'replace', 'return', |
| 33 |
'revalidation', 'same', 'satisfies', 'schema', 'schema-attribute', 'schema-element', 'score', |
| 34 |
'self', 'sensitive', 'sentence', 'sentences', 'sequence', 'skip', 'sliding', 'some', 'stable', |
| 35 |
'start', 'stemming', 'stop', 'strict', 'strip', 'switch', 'text', 'then', 'thesaurus', 'times', |
| 36 |
'to', 'transform', 'treat', 'try', 'tumbling', 'type', 'typeswitch', 'union', 'unordered', |
| 37 |
'update', 'updating', 'uppercase', 'using', 'validate', 'value', 'variable', 'version', |
| 38 |
'weight', 'when', 'where', 'wildcards', 'window', 'with', 'without', 'word', 'words', 'xquery']; |
| 39 |
for(var i=0, l=basic.length; i < l; i++) { kwObj[basic[i]] = kw(basic[i]);}; |
| 40 |
|
| 41 |
// a list of types. For each add a property to kwObj with the value of |
| 42 |
// {type: "atom", style: "atom"} |
| 43 |
var types = ['xs:anyAtomicType', 'xs:anySimpleType', 'xs:anyType', 'xs:anyURI', |
| 44 |
'xs:base64Binary', 'xs:boolean', 'xs:byte', 'xs:date', 'xs:dateTime', 'xs:dateTimeStamp', |
| 45 |
'xs:dayTimeDuration', 'xs:decimal', 'xs:double', 'xs:duration', 'xs:ENTITIES', 'xs:ENTITY', |
| 46 |
'xs:float', 'xs:gDay', 'xs:gMonth', 'xs:gMonthDay', 'xs:gYear', 'xs:gYearMonth', 'xs:hexBinary', |
| 47 |
'xs:ID', 'xs:IDREF', 'xs:IDREFS', 'xs:int', 'xs:integer', 'xs:item', 'xs:java', 'xs:language', |
| 48 |
'xs:long', 'xs:Name', 'xs:NCName', 'xs:negativeInteger', 'xs:NMTOKEN', 'xs:NMTOKENS', |
| 49 |
'xs:nonNegativeInteger', 'xs:nonPositiveInteger', 'xs:normalizedString', 'xs:NOTATION', |
| 50 |
'xs:numeric', 'xs:positiveInteger', 'xs:precisionDecimal', 'xs:QName', 'xs:short', 'xs:string', |
| 51 |
'xs:time', 'xs:token', 'xs:unsignedByte', 'xs:unsignedInt', 'xs:unsignedLong', |
| 52 |
'xs:unsignedShort', 'xs:untyped', 'xs:untypedAtomic', 'xs:yearMonthDuration']; |
| 53 |
for(var i=0, l=types.length; i < l; i++) { kwObj[types[i]] = atom;}; |
| 54 |
|
| 55 |
// each operator will add a property to kwObj with value of {type: "operator", style: "keyword"} |
| 56 |
var operators = ['eq', 'ne', 'lt', 'le', 'gt', 'ge', ':=', '=', '>', '>=', '<', '<=', '.', '|', '?', 'and', 'or', 'div', 'idiv', 'mod', '*', '/', '+', '-']; |
| 57 |
for(var i=0, l=operators.length; i < l; i++) { kwObj[operators[i]] = operator;}; |
| 58 |
|
| 59 |
// each axis_specifiers will add a property to kwObj with value of {type: "axis_specifier", style: "qualifier"} |
| 60 |
var axis_specifiers = ["self::", "attribute::", "child::", "descendant::", "descendant-or-self::", "parent::", |
| 61 |
"ancestor::", "ancestor-or-self::", "following::", "preceding::", "following-sibling::", "preceding-sibling::"]; |
| 62 |
for(var i=0, l=axis_specifiers.length; i < l; i++) { kwObj[axis_specifiers[i]] = qualifier; }; |
| 63 |
|
| 64 |
return kwObj; |
| 65 |
}(); |
| 66 |
|
| 67 |
function chain(stream, state, f) { |
| 68 |
state.tokenize = f; |
| 69 |
return f(stream, state); |
| 70 |
} |
| 71 |
|
| 72 |
// the primary mode tokenizer |
| 73 |
function tokenBase(stream, state) { |
| 74 |
var ch = stream.next(), |
| 75 |
mightBeFunction = false, |
| 76 |
isEQName = isEQNameAhead(stream); |
| 77 |
|
| 78 |
// an XML tag (if not in some sub, chained tokenizer) |
| 79 |
if (ch == "<") { |
| 80 |
if(stream.match("!--", true)) |
| 81 |
return chain(stream, state, tokenXMLComment); |
| 82 |
|
| 83 |
if(stream.match("![CDATA", false)) { |
| 84 |
state.tokenize = tokenCDATA; |
| 85 |
return "tag"; |
| 86 |
} |
| 87 |
|
| 88 |
if(stream.match("?", false)) { |
| 89 |
return chain(stream, state, tokenPreProcessing); |
| 90 |
} |
| 91 |
|
| 92 |
var isclose = stream.eat("/"); |
| 93 |
stream.eatSpace(); |
| 94 |
var tagName = "", c; |
| 95 |
while ((c = stream.eat(/[^\s\u00a0=<>\"\'\/?]/))) tagName += c; |
| 96 |
|
| 97 |
return chain(stream, state, tokenTag(tagName, isclose)); |
| 98 |
} |
| 99 |
// start code block |
| 100 |
else if(ch == "{") { |
| 101 |
pushStateStack(state, { type: "codeblock"}); |
| 102 |
return null; |
| 103 |
} |
| 104 |
// end code block |
| 105 |
else if(ch == "}") { |
| 106 |
popStateStack(state); |
| 107 |
return null; |
| 108 |
} |
| 109 |
// if we're in an XML block |
| 110 |
else if(isInXmlBlock(state)) { |
| 111 |
if(ch == ">") |
| 112 |
return "tag"; |
| 113 |
else if(ch == "/" && stream.eat(">")) { |
| 114 |
popStateStack(state); |
| 115 |
return "tag"; |
| 116 |
} |
| 117 |
else |
| 118 |
return "variable"; |
| 119 |
} |
| 120 |
// if a number |
| 121 |
else if (/\d/.test(ch)) { |
| 122 |
stream.match(/^\d*(?:\.\d*)?(?:E[+\-]?\d+)?/); |
| 123 |
return "atom"; |
| 124 |
} |
| 125 |
// comment start |
| 126 |
else if (ch === "(" && stream.eat(":")) { |
| 127 |
pushStateStack(state, { type: "comment"}); |
| 128 |
return chain(stream, state, tokenComment); |
| 129 |
} |
| 130 |
// quoted string |
| 131 |
else if (!isEQName && (ch === '"' || ch === "'")) |
| 132 |
return startString(stream, state, ch); |
| 133 |
// variable |
| 134 |
else if(ch === "$") { |
| 135 |
return chain(stream, state, tokenVariable); |
| 136 |
} |
| 137 |
// assignment |
| 138 |
else if(ch ===":" && stream.eat("=")) { |
| 139 |
return "keyword"; |
| 140 |
} |
| 141 |
// open paren |
| 142 |
else if(ch === "(") { |
| 143 |
pushStateStack(state, { type: "paren"}); |
| 144 |
return null; |
| 145 |
} |
| 146 |
// close paren |
| 147 |
else if(ch === ")") { |
| 148 |
popStateStack(state); |
| 149 |
return null; |
| 150 |
} |
| 151 |
// open paren |
| 152 |
else if(ch === "[") { |
| 153 |
pushStateStack(state, { type: "bracket"}); |
| 154 |
return null; |
| 155 |
} |
| 156 |
// close paren |
| 157 |
else if(ch === "]") { |
| 158 |
popStateStack(state); |
| 159 |
return null; |
| 160 |
} |
| 161 |
else { |
| 162 |
var known = keywords.propertyIsEnumerable(ch) && keywords[ch]; |
| 163 |
|
| 164 |
// if there's a EQName ahead, consume the rest of the string portion, it's likely a function |
| 165 |
if(isEQName && ch === '\"') while(stream.next() !== '"'){} |
| 166 |
if(isEQName && ch === '\'') while(stream.next() !== '\''){} |
| 167 |
|
| 168 |
// gobble up a word if the character is not known |
| 169 |
if(!known) stream.eatWhile(/[\w\$_-]/); |
| 170 |
|
| 171 |
// gobble a colon in the case that is a lib func type call fn:doc |
| 172 |
var foundColon = stream.eat(":"); |
| 173 |
|
| 174 |
// if there's not a second colon, gobble another word. Otherwise, it's probably an axis specifier |
| 175 |
// which should get matched as a keyword |
| 176 |
if(!stream.eat(":") && foundColon) { |
| 177 |
stream.eatWhile(/[\w\$_-]/); |
| 178 |
} |
| 179 |
// if the next non whitespace character is an open paren, this is probably a function (if not a keyword of other sort) |
| 180 |
if(stream.match(/^[ \t]*\(/, false)) { |
| 181 |
mightBeFunction = true; |
| 182 |
} |
| 183 |
// is the word a keyword? |
| 184 |
var word = stream.current(); |
| 185 |
known = keywords.propertyIsEnumerable(word) && keywords[word]; |
| 186 |
|
| 187 |
// if we think it's a function call but not yet known, |
| 188 |
// set style to variable for now for lack of something better |
| 189 |
if(mightBeFunction && !known) known = {type: "function_call", style: "def"}; |
| 190 |
|
| 191 |
// if the previous word was element, attribute, axis specifier, this word should be the name of that |
| 192 |
if(isInXmlConstructor(state)) { |
| 193 |
popStateStack(state); |
| 194 |
return "variable"; |
| 195 |
} |
| 196 |
// as previously checked, if the word is element,attribute, axis specifier, call it an "xmlconstructor" and |
| 197 |
// push the stack so we know to look for it on the next word |
| 198 |
if(word == "element" || word == "attribute" || known.type == "axis_specifier") pushStateStack(state, {type: "xmlconstructor"}); |
| 199 |
|
| 200 |
// if the word is known, return the details of that else just call this a generic 'word' |
| 201 |
return known ? known.style : "variable"; |
| 202 |
} |
| 203 |
} |
| 204 |
|
| 205 |
// handle comments, including nested |
| 206 |
function tokenComment(stream, state) { |
| 207 |
var maybeEnd = false, maybeNested = false, nestedCount = 0, ch; |
| 208 |
while (ch = stream.next()) { |
| 209 |
if (ch == ")" && maybeEnd) { |
| 210 |
if(nestedCount > 0) |
| 211 |
nestedCount--; |
| 212 |
else { |
| 213 |
popStateStack(state); |
| 214 |
break; |
| 215 |
} |
| 216 |
} |
| 217 |
else if(ch == ":" && maybeNested) { |
| 218 |
nestedCount++; |
| 219 |
} |
| 220 |
maybeEnd = (ch == ":"); |
| 221 |
maybeNested = (ch == "("); |
| 222 |
} |
| 223 |
|
| 224 |
return "comment"; |
| 225 |
} |
| 226 |
|
| 227 |
// tokenizer for string literals |
| 228 |
// optionally pass a tokenizer function to set state.tokenize back to when finished |
| 229 |
function tokenString(quote, f) { |
| 230 |
return function(stream, state) { |
| 231 |
var ch; |
| 232 |
while (ch = stream.next()) { |
| 233 |
if (ch == quote) { |
| 234 |
popStateStack(state); |
| 235 |
if (f) state.tokenize = f; |
| 236 |
break; |
| 237 |
} else if (stream.match("{", false) && isInXmlAttributeBlock(state)) { |
| 238 |
// if we're in a string and in an XML block, allow an embedded code block in an attribute |
| 239 |
pushStateStack(state, { type: "codeblock"}); |
| 240 |
state.tokenize = tokenBase; |
| 241 |
return "string"; |
| 242 |
} |
| 243 |
} |
| 244 |
|
| 245 |
return "string"; |
| 246 |
}; |
| 247 |
} |
| 248 |
|
| 249 |
function startString(stream, state, quote, f) { |
| 250 |
let tokenize = tokenString(quote, f); |
| 251 |
pushStateStack(state, { type: "string", name: quote, tokenize }); |
| 252 |
return chain(stream, state, tokenize); |
| 253 |
} |
| 254 |
|
| 255 |
// tokenizer for variables |
| 256 |
function tokenVariable(stream, state) { |
| 257 |
var isVariableChar = /[\w\$_-]/; |
| 258 |
|
| 259 |
// a variable may start with a quoted EQName so if the next character is quote, consume to the next quote |
| 260 |
if(stream.eat("\"")) { |
| 261 |
while(stream.next() !== '\"'){}; |
| 262 |
stream.eat(":"); |
| 263 |
} else { |
| 264 |
stream.eatWhile(isVariableChar); |
| 265 |
if(!stream.match(":=", false)) stream.eat(":"); |
| 266 |
} |
| 267 |
stream.eatWhile(isVariableChar); |
| 268 |
state.tokenize = tokenBase; |
| 269 |
return "variable"; |
| 270 |
} |
| 271 |
|
| 272 |
// tokenizer for XML tags |
| 273 |
function tokenTag(name, isclose) { |
| 274 |
return function(stream, state) { |
| 275 |
stream.eatSpace(); |
| 276 |
if(isclose && stream.eat(">")) { |
| 277 |
popStateStack(state); |
| 278 |
state.tokenize = tokenBase; |
| 279 |
return "tag"; |
| 280 |
} |
| 281 |
// self closing tag without attributes? |
| 282 |
if(!stream.eat("/")) |
| 283 |
pushStateStack(state, { type: "tag", name: name, tokenize: tokenBase}); |
| 284 |
if(!stream.eat(">")) { |
| 285 |
state.tokenize = tokenAttribute; |
| 286 |
return "tag"; |
| 287 |
} |
| 288 |
else { |
| 289 |
state.tokenize = tokenBase; |
| 290 |
} |
| 291 |
return "tag"; |
| 292 |
}; |
| 293 |
} |
| 294 |
|
| 295 |
// tokenizer for XML attributes |
| 296 |
function tokenAttribute(stream, state) { |
| 297 |
var ch = stream.next(); |
| 298 |
|
| 299 |
if(ch == "/" && stream.eat(">")) { |
| 300 |
if(isInXmlAttributeBlock(state)) popStateStack(state); |
| 301 |
if(isInXmlBlock(state)) popStateStack(state); |
| 302 |
return "tag"; |
| 303 |
} |
| 304 |
if(ch == ">") { |
| 305 |
if(isInXmlAttributeBlock(state)) popStateStack(state); |
| 306 |
return "tag"; |
| 307 |
} |
| 308 |
if(ch == "=") |
| 309 |
return null; |
| 310 |
// quoted string |
| 311 |
if (ch == '"' || ch == "'") |
| 312 |
return startString(stream, state, ch, tokenAttribute); |
| 313 |
|
| 314 |
if(!isInXmlAttributeBlock(state)) |
| 315 |
pushStateStack(state, { type: "attribute", tokenize: tokenAttribute}); |
| 316 |
|
| 317 |
stream.eat(/[a-zA-Z_:]/); |
| 318 |
stream.eatWhile(/[-a-zA-Z0-9_:.]/); |
| 319 |
stream.eatSpace(); |
| 320 |
|
| 321 |
// the case where the attribute has not value and the tag was closed |
| 322 |
if(stream.match(">", false) || stream.match("/", false)) { |
| 323 |
popStateStack(state); |
| 324 |
state.tokenize = tokenBase; |
| 325 |
} |
| 326 |
|
| 327 |
return "attribute"; |
| 328 |
} |
| 329 |
|
| 330 |
// handle comments, including nested |
| 331 |
function tokenXMLComment(stream, state) { |
| 332 |
var ch; |
| 333 |
while (ch = stream.next()) { |
| 334 |
if (ch == "-" && stream.match("->", true)) { |
| 335 |
state.tokenize = tokenBase; |
| 336 |
return "comment"; |
| 337 |
} |
| 338 |
} |
| 339 |
} |
| 340 |
|
| 341 |
|
| 342 |
// handle CDATA |
| 343 |
function tokenCDATA(stream, state) { |
| 344 |
var ch; |
| 345 |
while (ch = stream.next()) { |
| 346 |
if (ch == "]" && stream.match("]", true)) { |
| 347 |
state.tokenize = tokenBase; |
| 348 |
return "comment"; |
| 349 |
} |
| 350 |
} |
| 351 |
} |
| 352 |
|
| 353 |
// handle preprocessing instructions |
| 354 |
function tokenPreProcessing(stream, state) { |
| 355 |
var ch; |
| 356 |
while (ch = stream.next()) { |
| 357 |
if (ch == "?" && stream.match(">", true)) { |
| 358 |
state.tokenize = tokenBase; |
| 359 |
return "processingInstruction"; |
| 360 |
} |
| 361 |
} |
| 362 |
} |
| 363 |
|
| 364 |
|
| 365 |
// functions to test the current context of the state |
| 366 |
function isInXmlBlock(state) { return isIn(state, "tag"); } |
| 367 |
function isInXmlAttributeBlock(state) { return isIn(state, "attribute"); } |
| 368 |
function isInXmlConstructor(state) { return isIn(state, "xmlconstructor"); } |
| 369 |
function isInString(state) { return isIn(state, "string"); } |
| 370 |
|
| 371 |
function isEQNameAhead(stream) { |
| 372 |
// assume we've already eaten a quote (") |
| 373 |
if(stream.current() === '"') |
| 374 |
return stream.match(/^[^\"]+\"\:/, false); |
| 375 |
else if(stream.current() === '\'') |
| 376 |
return stream.match(/^[^\"]+\'\:/, false); |
| 377 |
else |
| 378 |
return false; |
| 379 |
} |
| 380 |
|
| 381 |
function isIn(state, type) { |
| 382 |
return (state.stack.length && state.stack[state.stack.length - 1].type == type); |
| 383 |
} |
| 384 |
|
| 385 |
function pushStateStack(state, newState) { |
| 386 |
state.stack.push(newState); |
| 387 |
} |
| 388 |
|
| 389 |
function popStateStack(state) { |
| 390 |
state.stack.pop(); |
| 391 |
var reinstateTokenize = state.stack.length && state.stack[state.stack.length-1].tokenize; |
| 392 |
state.tokenize = reinstateTokenize || tokenBase; |
| 393 |
} |
| 394 |
|
| 395 |
// the interface for the mode API |
| 396 |
export const xQuery = { |
| 397 |
name: "xquery", |
| 398 |
startState: function() { |
| 399 |
return { |
| 400 |
tokenize: tokenBase, |
| 401 |
cc: [], |
| 402 |
stack: [] |
| 403 |
}; |
| 404 |
}, |
| 405 |
|
| 406 |
token: function(stream, state) { |
| 407 |
if (stream.eatSpace()) return null; |
| 408 |
var style = state.tokenize(stream, state); |
| 409 |
return style; |
| 410 |
}, |
| 411 |
|
| 412 |
languageData: { |
| 413 |
commentTokens: {block: {open: "(:", close: ":)"}} |
| 414 |
} |
| 415 |
}; |