123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447 |
- // CodeMirror, copyright (c) by Marijn Haverbeke and others
- // Distributed under an MIT license: http://codemirror.net/LICENSE
-
- (function(mod) {
- if (typeof exports == "object" && typeof module == "object") // CommonJS
- mod(require("../../lib/codemirror"));
- else if (typeof define == "function" && define.amd) // AMD
- define(["../../lib/codemirror"], mod);
- else // Plain browser env
- mod(CodeMirror);
- })(function(CodeMirror) {
- "use strict";
-
- CodeMirror.defineMode("xquery", function() {
-
- // The keywords object is set to the result of this self executing
- // function. Each keyword is a property of the keywords object whose
- // value is {type: atype, style: astyle}
- var keywords = function(){
- // conveinence functions used to build keywords object
- function kw(type) {return {type: type, style: "keyword"};}
- var A = kw("keyword a")
- , B = kw("keyword b")
- , C = kw("keyword c")
- , operator = kw("operator")
- , atom = {type: "atom", style: "atom"}
- , punctuation = {type: "punctuation", style: null}
- , qualifier = {type: "axis_specifier", style: "qualifier"};
-
- // kwObj is what is return from this function at the end
- var kwObj = {
- 'if': A, 'switch': A, 'while': A, 'for': A,
- 'else': B, 'then': B, 'try': B, 'finally': B, 'catch': B,
- 'element': C, 'attribute': C, 'let': C, 'implements': C, 'import': C, 'module': C, 'namespace': C,
- 'return': C, 'super': C, 'this': C, 'throws': C, 'where': C, 'private': C,
- ',': punctuation,
- 'null': atom, 'fn:false()': atom, 'fn:true()': atom
- };
-
- // a list of 'basic' keywords. For each add a property to kwObj with the value of
- // {type: basic[i], style: "keyword"} e.g. 'after' --> {type: "after", style: "keyword"}
- var basic = ['after','ancestor','ancestor-or-self','and','as','ascending','assert','attribute','before',
- 'by','case','cast','child','comment','declare','default','define','descendant','descendant-or-self',
- 'descending','document','document-node','element','else','eq','every','except','external','following',
- 'following-sibling','follows','for','function','if','import','in','instance','intersect','item',
- 'let','module','namespace','node','node','of','only','or','order','parent','precedes','preceding',
- 'preceding-sibling','processing-instruction','ref','return','returns','satisfies','schema','schema-element',
- 'self','some','sortby','stable','text','then','to','treat','typeswitch','union','variable','version','where',
- 'xquery', 'empty-sequence'];
- for(var i=0, l=basic.length; i < l; i++) { kwObj[basic[i]] = kw(basic[i]);};
-
- // a list of types. For each add a property to kwObj with the value of
- // {type: "atom", style: "atom"}
- var types = ['xs:string', 'xs:float', 'xs:decimal', 'xs:double', 'xs:integer', 'xs:boolean', 'xs:date', 'xs:dateTime',
- 'xs:time', 'xs:duration', 'xs:dayTimeDuration', 'xs:time', 'xs:yearMonthDuration', 'numeric', 'xs:hexBinary',
- 'xs:base64Binary', 'xs:anyURI', 'xs:QName', 'xs:byte','xs:boolean','xs:anyURI','xf:yearMonthDuration'];
- for(var i=0, l=types.length; i < l; i++) { kwObj[types[i]] = atom;};
-
- // each operator will add a property to kwObj with value of {type: "operator", style: "keyword"}
- var operators = ['eq', 'ne', 'lt', 'le', 'gt', 'ge', ':=', '=', '>', '>=', '<', '<=', '.', '|', '?', 'and', 'or', 'div', 'idiv', 'mod', '*', '/', '+', '-'];
- for(var i=0, l=operators.length; i < l; i++) { kwObj[operators[i]] = operator;};
-
- // each axis_specifiers will add a property to kwObj with value of {type: "axis_specifier", style: "qualifier"}
- var axis_specifiers = ["self::", "attribute::", "child::", "descendant::", "descendant-or-self::", "parent::",
- "ancestor::", "ancestor-or-self::", "following::", "preceding::", "following-sibling::", "preceding-sibling::"];
- for(var i=0, l=axis_specifiers.length; i < l; i++) { kwObj[axis_specifiers[i]] = qualifier; };
-
- return kwObj;
- }();
-
- // Used as scratch variables to communicate multiple values without
- // consing up tons of objects.
- var type, content;
-
- function ret(tp, style, cont) {
- type = tp; content = cont;
- return style;
- }
-
- function chain(stream, state, f) {
- state.tokenize = f;
- return f(stream, state);
- }
-
- // the primary mode tokenizer
- function tokenBase(stream, state) {
- var ch = stream.next(),
- mightBeFunction = false,
- isEQName = isEQNameAhead(stream);
-
- // an XML tag (if not in some sub, chained tokenizer)
- if (ch == "<") {
- if(stream.match("!--", true))
- return chain(stream, state, tokenXMLComment);
-
- if(stream.match("![CDATA", false)) {
- state.tokenize = tokenCDATA;
- return ret("tag", "tag");
- }
-
- if(stream.match("?", false)) {
- return chain(stream, state, tokenPreProcessing);
- }
-
- var isclose = stream.eat("/");
- stream.eatSpace();
- var tagName = "", c;
- while ((c = stream.eat(/[^\s\u00a0=<>\"\'\/?]/))) tagName += c;
-
- return chain(stream, state, tokenTag(tagName, isclose));
- }
- // start code block
- else if(ch == "{") {
- pushStateStack(state,{ type: "codeblock"});
- return ret("", null);
- }
- // end code block
- else if(ch == "}") {
- popStateStack(state);
- return ret("", null);
- }
- // if we're in an XML block
- else if(isInXmlBlock(state)) {
- if(ch == ">")
- return ret("tag", "tag");
- else if(ch == "/" && stream.eat(">")) {
- popStateStack(state);
- return ret("tag", "tag");
- }
- else
- return ret("word", "variable");
- }
- // if a number
- else if (/\d/.test(ch)) {
- stream.match(/^\d*(?:\.\d*)?(?:E[+\-]?\d+)?/);
- return ret("number", "atom");
- }
- // comment start
- else if (ch === "(" && stream.eat(":")) {
- pushStateStack(state, { type: "comment"});
- return chain(stream, state, tokenComment);
- }
- // quoted string
- else if ( !isEQName && (ch === '"' || ch === "'"))
- return chain(stream, state, tokenString(ch));
- // variable
- else if(ch === "$") {
- return chain(stream, state, tokenVariable);
- }
- // assignment
- else if(ch ===":" && stream.eat("=")) {
- return ret("operator", "keyword");
- }
- // open paren
- else if(ch === "(") {
- pushStateStack(state, { type: "paren"});
- return ret("", null);
- }
- // close paren
- else if(ch === ")") {
- popStateStack(state);
- return ret("", null);
- }
- // open paren
- else if(ch === "[") {
- pushStateStack(state, { type: "bracket"});
- return ret("", null);
- }
- // close paren
- else if(ch === "]") {
- popStateStack(state);
- return ret("", null);
- }
- else {
- var known = keywords.propertyIsEnumerable(ch) && keywords[ch];
-
- // if there's a EQName ahead, consume the rest of the string portion, it's likely a function
- if(isEQName && ch === '\"') while(stream.next() !== '"'){}
- if(isEQName && ch === '\'') while(stream.next() !== '\''){}
-
- // gobble up a word if the character is not known
- if(!known) stream.eatWhile(/[\w\$_-]/);
-
- // gobble a colon in the case that is a lib func type call fn:doc
- var foundColon = stream.eat(":");
-
- // if there's not a second colon, gobble another word. Otherwise, it's probably an axis specifier
- // which should get matched as a keyword
- if(!stream.eat(":") && foundColon) {
- stream.eatWhile(/[\w\$_-]/);
- }
- // if the next non whitespace character is an open paren, this is probably a function (if not a keyword of other sort)
- if(stream.match(/^[ \t]*\(/, false)) {
- mightBeFunction = true;
- }
- // is the word a keyword?
- var word = stream.current();
- known = keywords.propertyIsEnumerable(word) && keywords[word];
-
- // if we think it's a function call but not yet known,
- // set style to variable for now for lack of something better
- if(mightBeFunction && !known) known = {type: "function_call", style: "variable def"};
-
- // if the previous word was element, attribute, axis specifier, this word should be the name of that
- if(isInXmlConstructor(state)) {
- popStateStack(state);
- return ret("word", "variable", word);
- }
- // as previously checked, if the word is element,attribute, axis specifier, call it an "xmlconstructor" and
- // push the stack so we know to look for it on the next word
- if(word == "element" || word == "attribute" || known.type == "axis_specifier") pushStateStack(state, {type: "xmlconstructor"});
-
- // if the word is known, return the details of that else just call this a generic 'word'
- return known ? ret(known.type, known.style, word) :
- ret("word", "variable", word);
- }
- }
-
- // handle comments, including nested
- function tokenComment(stream, state) {
- var maybeEnd = false, maybeNested = false, nestedCount = 0, ch;
- while (ch = stream.next()) {
- if (ch == ")" && maybeEnd) {
- if(nestedCount > 0)
- nestedCount--;
- else {
- popStateStack(state);
- break;
- }
- }
- else if(ch == ":" && maybeNested) {
- nestedCount++;
- }
- maybeEnd = (ch == ":");
- maybeNested = (ch == "(");
- }
-
- return ret("comment", "comment");
- }
-
- // tokenizer for string literals
- // optionally pass a tokenizer function to set state.tokenize back to when finished
- function tokenString(quote, f) {
- return function(stream, state) {
- var ch;
-
- if(isInString(state) && stream.current() == quote) {
- popStateStack(state);
- if(f) state.tokenize = f;
- return ret("string", "string");
- }
-
- pushStateStack(state, { type: "string", name: quote, tokenize: tokenString(quote, f) });
-
- // if we're in a string and in an XML block, allow an embedded code block
- if(stream.match("{", false) && isInXmlAttributeBlock(state)) {
- state.tokenize = tokenBase;
- return ret("string", "string");
- }
-
-
- while (ch = stream.next()) {
- if (ch == quote) {
- popStateStack(state);
- if(f) state.tokenize = f;
- break;
- }
- else {
- // if we're in a string and in an XML block, allow an embedded code block in an attribute
- if(stream.match("{", false) && isInXmlAttributeBlock(state)) {
- state.tokenize = tokenBase;
- return ret("string", "string");
- }
-
- }
- }
-
- return ret("string", "string");
- };
- }
-
- // tokenizer for variables
- function tokenVariable(stream, state) {
- var isVariableChar = /[\w\$_-]/;
-
- // a variable may start with a quoted EQName so if the next character is quote, consume to the next quote
- if(stream.eat("\"")) {
- while(stream.next() !== '\"'){};
- stream.eat(":");
- } else {
- stream.eatWhile(isVariableChar);
- if(!stream.match(":=", false)) stream.eat(":");
- }
- stream.eatWhile(isVariableChar);
- state.tokenize = tokenBase;
- return ret("variable", "variable");
- }
-
- // tokenizer for XML tags
- function tokenTag(name, isclose) {
- return function(stream, state) {
- stream.eatSpace();
- if(isclose && stream.eat(">")) {
- popStateStack(state);
- state.tokenize = tokenBase;
- return ret("tag", "tag");
- }
- // self closing tag without attributes?
- if(!stream.eat("/"))
- pushStateStack(state, { type: "tag", name: name, tokenize: tokenBase});
- if(!stream.eat(">")) {
- state.tokenize = tokenAttribute;
- return ret("tag", "tag");
- }
- else {
- state.tokenize = tokenBase;
- }
- return ret("tag", "tag");
- };
- }
-
- // tokenizer for XML attributes
- function tokenAttribute(stream, state) {
- var ch = stream.next();
-
- if(ch == "/" && stream.eat(">")) {
- if(isInXmlAttributeBlock(state)) popStateStack(state);
- if(isInXmlBlock(state)) popStateStack(state);
- return ret("tag", "tag");
- }
- if(ch == ">") {
- if(isInXmlAttributeBlock(state)) popStateStack(state);
- return ret("tag", "tag");
- }
- if(ch == "=")
- return ret("", null);
- // quoted string
- if (ch == '"' || ch == "'")
- return chain(stream, state, tokenString(ch, tokenAttribute));
-
- if(!isInXmlAttributeBlock(state))
- pushStateStack(state, { type: "attribute", tokenize: tokenAttribute});
-
- stream.eat(/[a-zA-Z_:]/);
- stream.eatWhile(/[-a-zA-Z0-9_:.]/);
- stream.eatSpace();
-
- // the case where the attribute has not value and the tag was closed
- if(stream.match(">", false) || stream.match("/", false)) {
- popStateStack(state);
- state.tokenize = tokenBase;
- }
-
- return ret("attribute", "attribute");
- }
-
- // handle comments, including nested
- function tokenXMLComment(stream, state) {
- var ch;
- while (ch = stream.next()) {
- if (ch == "-" && stream.match("->", true)) {
- state.tokenize = tokenBase;
- return ret("comment", "comment");
- }
- }
- }
-
-
- // handle CDATA
- function tokenCDATA(stream, state) {
- var ch;
- while (ch = stream.next()) {
- if (ch == "]" && stream.match("]", true)) {
- state.tokenize = tokenBase;
- return ret("comment", "comment");
- }
- }
- }
-
- // handle preprocessing instructions
- function tokenPreProcessing(stream, state) {
- var ch;
- while (ch = stream.next()) {
- if (ch == "?" && stream.match(">", true)) {
- state.tokenize = tokenBase;
- return ret("comment", "comment meta");
- }
- }
- }
-
-
- // functions to test the current context of the state
- function isInXmlBlock(state) { return isIn(state, "tag"); }
- function isInXmlAttributeBlock(state) { return isIn(state, "attribute"); }
- function isInXmlConstructor(state) { return isIn(state, "xmlconstructor"); }
- function isInString(state) { return isIn(state, "string"); }
-
- function isEQNameAhead(stream) {
- // assume we've already eaten a quote (")
- if(stream.current() === '"')
- return stream.match(/^[^\"]+\"\:/, false);
- else if(stream.current() === '\'')
- return stream.match(/^[^\"]+\'\:/, false);
- else
- return false;
- }
-
- function isIn(state, type) {
- return (state.stack.length && state.stack[state.stack.length - 1].type == type);
- }
-
- function pushStateStack(state, newState) {
- state.stack.push(newState);
- }
-
- function popStateStack(state) {
- state.stack.pop();
- var reinstateTokenize = state.stack.length && state.stack[state.stack.length-1].tokenize;
- state.tokenize = reinstateTokenize || tokenBase;
- }
-
- // the interface for the mode API
- return {
- startState: function() {
- return {
- tokenize: tokenBase,
- cc: [],
- stack: []
- };
- },
-
- token: function(stream, state) {
- if (stream.eatSpace()) return null;
- var style = state.tokenize(stream, state);
- return style;
- },
-
- blockCommentStart: "(:",
- blockCommentEnd: ":)"
-
- };
-
- });
-
- CodeMirror.defineMIME("application/xquery", "xquery");
-
- });
|