| 1 |
// used pattern building blocks |
| 2 |
var Identifier = '[a-zA-Z\\$][a-zA-Z0-9\\$]*'; |
| 3 |
var pBase = "(?:\\d+)"; |
| 4 |
var pFloat = "(?:\\.\\d+|\\d+\\.\\d*|\\d+)"; |
| 5 |
var pFloatBase = "(?:\\.\\w+|\\w+\\.\\w*|\\w+)"; |
| 6 |
var pPrecision = "(?:`(?:`?"+pFloat+")?)"; |
| 7 |
|
| 8 |
// regular expressions |
| 9 |
var reBaseForm = new RegExp('(?:'+pBase+'(?:\\^\\^'+pFloatBase+pPrecision+'?(?:\\*\\^[+-]?\\d+)?))'); |
| 10 |
var reFloatForm = new RegExp('(?:' + pFloat + pPrecision + '?(?:\\*\\^[+-]?\\d+)?)'); |
| 11 |
var reIdInContext = new RegExp('(?:`?)(?:' + Identifier + ')(?:`(?:' + Identifier + '))*(?:`?)'); |
| 12 |
|
| 13 |
function tokenBase(stream, state) { |
| 14 |
var ch; |
| 15 |
|
| 16 |
// get next character |
| 17 |
ch = stream.next(); |
| 18 |
|
| 19 |
// string |
| 20 |
if (ch === '"') { |
| 21 |
state.tokenize = tokenString; |
| 22 |
return state.tokenize(stream, state); |
| 23 |
} |
| 24 |
|
| 25 |
// comment |
| 26 |
if (ch === '(') { |
| 27 |
if (stream.eat('*')) { |
| 28 |
state.commentLevel++; |
| 29 |
state.tokenize = tokenComment; |
| 30 |
return state.tokenize(stream, state); |
| 31 |
} |
| 32 |
} |
| 33 |
|
| 34 |
// go back one character |
| 35 |
stream.backUp(1); |
| 36 |
|
| 37 |
// look for numbers |
| 38 |
// Numbers in a baseform |
| 39 |
if (stream.match(reBaseForm, true, false)) { |
| 40 |
return 'number'; |
| 41 |
} |
| 42 |
|
| 43 |
// Mathematica numbers. Floats (1.2, .2, 1.) can have optionally a precision (`float) or an accuracy definition |
| 44 |
// (``float). Note: while 1.2` is possible 1.2`` is not. At the end an exponent (float*^+12) can follow. |
| 45 |
if (stream.match(reFloatForm, true, false)) { |
| 46 |
return 'number'; |
| 47 |
} |
| 48 |
|
| 49 |
/* In[23] and Out[34] */ |
| 50 |
if (stream.match(/(?:In|Out)\[[0-9]*\]/, true, false)) { |
| 51 |
return 'atom'; |
| 52 |
} |
| 53 |
|
| 54 |
// usage |
| 55 |
if (stream.match(/([a-zA-Z\$][a-zA-Z0-9\$]*(?:`[a-zA-Z0-9\$]+)*::usage)/, true, false)) { |
| 56 |
return 'meta'; |
| 57 |
} |
| 58 |
|
| 59 |
// message |
| 60 |
if (stream.match(/([a-zA-Z\$][a-zA-Z0-9\$]*(?:`[a-zA-Z0-9\$]+)*::[a-zA-Z\$][a-zA-Z0-9\$]*):?/, true, false)) { |
| 61 |
return 'string.special'; |
| 62 |
} |
| 63 |
|
| 64 |
// this makes a look-ahead match for something like variable:{_Integer} |
| 65 |
// the match is then forwarded to the mma-patterns tokenizer. |
| 66 |
if (stream.match(/([a-zA-Z\$][a-zA-Z0-9\$]*\s*:)(?:(?:[a-zA-Z\$][a-zA-Z0-9\$]*)|(?:[^:=>~@\^\&\*\)\[\]'\?,\|])).*/, true, false)) { |
| 67 |
return 'variableName.special'; |
| 68 |
} |
| 69 |
|
| 70 |
// catch variables which are used together with Blank (_), BlankSequence (__) or BlankNullSequence (___) |
| 71 |
// Cannot start with a number, but can have numbers at any other position. Examples |
| 72 |
// blub__Integer, a1_, b34_Integer32 |
| 73 |
if (stream.match(/[a-zA-Z\$][a-zA-Z0-9\$]*_+[a-zA-Z\$][a-zA-Z0-9\$]*/, true, false)) { |
| 74 |
return 'variableName.special'; |
| 75 |
} |
| 76 |
if (stream.match(/[a-zA-Z\$][a-zA-Z0-9\$]*_+/, true, false)) { |
| 77 |
return 'variableName.special'; |
| 78 |
} |
| 79 |
if (stream.match(/_+[a-zA-Z\$][a-zA-Z0-9\$]*/, true, false)) { |
| 80 |
return 'variableName.special'; |
| 81 |
} |
| 82 |
|
| 83 |
// Named characters in Mathematica, like \[Gamma]. |
| 84 |
if (stream.match(/\\\[[a-zA-Z\$][a-zA-Z0-9\$]*\]/, true, false)) { |
| 85 |
return 'character'; |
| 86 |
} |
| 87 |
|
| 88 |
// Match all braces separately |
| 89 |
if (stream.match(/(?:\[|\]|{|}|\(|\))/, true, false)) { |
| 90 |
return 'bracket'; |
| 91 |
} |
| 92 |
|
| 93 |
// Catch Slots (#, ##, #3, ##9 and the V10 named slots #name). I have never seen someone using more than one digit after #, so we match |
| 94 |
// only one. |
| 95 |
if (stream.match(/(?:#[a-zA-Z\$][a-zA-Z0-9\$]*|#+[0-9]?)/, true, false)) { |
| 96 |
return 'variableName.constant'; |
| 97 |
} |
| 98 |
|
| 99 |
// Literals like variables, keywords, functions |
| 100 |
if (stream.match(reIdInContext, true, false)) { |
| 101 |
return 'keyword'; |
| 102 |
} |
| 103 |
|
| 104 |
// operators. Note that operators like @@ or /; are matched separately for each symbol. |
| 105 |
if (stream.match(/(?:\\|\+|\-|\*|\/|,|;|\.|:|@|~|=|>|<|&|\||_|`|'|\^|\?|!|%)/, true, false)) { |
| 106 |
return 'operator'; |
| 107 |
} |
| 108 |
|
| 109 |
// everything else is an error |
| 110 |
stream.next(); // advance the stream. |
| 111 |
return 'error'; |
| 112 |
} |
| 113 |
|
| 114 |
function tokenString(stream, state) { |
| 115 |
var next, end = false, escaped = false; |
| 116 |
while ((next = stream.next()) != null) { |
| 117 |
if (next === '"' && !escaped) { |
| 118 |
end = true; |
| 119 |
break; |
| 120 |
} |
| 121 |
escaped = !escaped && next === '\\'; |
| 122 |
} |
| 123 |
if (end && !escaped) { |
| 124 |
state.tokenize = tokenBase; |
| 125 |
} |
| 126 |
return 'string'; |
| 127 |
}; |
| 128 |
|
| 129 |
function tokenComment(stream, state) { |
| 130 |
var prev, next; |
| 131 |
while(state.commentLevel > 0 && (next = stream.next()) != null) { |
| 132 |
if (prev === '(' && next === '*') state.commentLevel++; |
| 133 |
if (prev === '*' && next === ')') state.commentLevel--; |
| 134 |
prev = next; |
| 135 |
} |
| 136 |
if (state.commentLevel <= 0) { |
| 137 |
state.tokenize = tokenBase; |
| 138 |
} |
| 139 |
return 'comment'; |
| 140 |
} |
| 141 |
|
| 142 |
export const mathematica = { |
| 143 |
name: "mathematica", |
| 144 |
startState: function() {return {tokenize: tokenBase, commentLevel: 0};}, |
| 145 |
token: function(stream, state) { |
| 146 |
if (stream.eatSpace()) return null; |
| 147 |
return state.tokenize(stream, state); |
| 148 |
}, |
| 149 |
languageData: { |
| 150 |
commentTokens: {block: {open: "(*", close: "*)"}} |
| 151 |
} |
| 152 |
} |
| 153 |
|