joepie91
/
pegjs
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
redux
master
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from '570658756a'
${ noResults }
pegjs/lib/compiler.js

1084 lines
36 KiB
JavaScript
Raw Blame History

/* PEG.js compiler. */
(function() {
/* ===== PEG ===== */
/* no var */ PEG = {};
/*
* Generates a parser from a specified grammar and start rule and returns it.
*
* The grammar must be a string in the format described by the metagramar in the
* metagrammar.pegjs file. The start rule may be unspecified, in which case
* "start" is used.
*
* Throws |PEG.grammarParser.SyntaxError| if the grammar contains a syntax error
* or |PEG.GrammarError| if it contains a semantic error. Note that not all
* errors are detected during the generation and some may protrude to the
* generated parser and cause its malfunction.
*/
PEG.buildParser = function(grammar, startRule) {
startRule = startRule || "start";
return PEG.Compiler.compileParser(
PEG.grammarParser.parse(grammar),
startRule
);
};
/* ===== PEG.GrammarError ===== */
/* Thrown when the grammar contains an error. */
PEG.GrammarError = function(message) {
this.name = "PEG.GrammarError";
this.message = message;
};
PEG.GrammarError.prototype = Error.prototype;
/* ===== PEG.ArrayUtils ===== */
/* Array manipulation utility functions. */
PEG.ArrayUtils = {
/* Like Python's |range|, but without |step|. */
range: function(start, stop) {
if (typeof(stop) === "undefined") {
stop = start;
start = 0;
}
var result = new Array(Math.max(0, stop - start));
for (var i = 0, j = start; j < stop; i++, j++) {
result[i] = j;
}
return result;
},
/*
* The code needs to be in sync with the code template in the compilation
* function for "action" nodes.
*/
contains: function(array, value) {
/*
* Stupid IE does not have Array.prototype.indexOf, otherwise this function
* would be a one-liner.
*/
var length = array.length;
for (var i = 0; i < length; i++) {
if (array[i] === value) {
return true;
}
}
return false;
},
each: function(array, callback) {
var length = array.length;
for (var i = 0; i < length; i++) {
callback(array[i]);
}
},
map: function(array, callback) {
var result = [];
var length = array.length;
for (var i = 0; i < length; i++) {
result[i] = callback(array[i]);
}
return result;
}
};
/* ===== PEG.StringUtils ===== */
/* String manipulation utility functions. */
PEG.StringUtils = {
/*
* Surrounds the string with quotes and escapes characters inside so that the
* result is a valid JavaScript string.
*
* The code needs to be in sync with th code template in the compilation
* function for "action" nodes.
*/
quote: function(s) {
/*
* ECMA-262, 5th ed., 7.8.4: All characters may appear literally in a string
* literal except for the closing quote character, backslash, carriage
* return, line separator, paragraph separator, and line feed. Any character
* may appear in the form of an escape sequence.
*/
return '"' + s
.replace(/\\/g, '\\\\') // backslash
.replace(/"/g, '\\"') // closing quote character
.replace(/\r/g, '\\r') // carriage return
.replace(/\u2028/g, '\\u2028') // line separator
.replace(/\u2029/g, '\\u2029') // paragraph separator
.replace(/\n/g, '\\n') // line feed
+ '"';
}
};
/* ===== PEG.RegExpUtils ===== */
/* RegExp manipulation utility functions. */
PEG.RegExpUtils = {
/*
* Escapes characters inside the string so that it can be used as a list of
* characters in a character class of a regular expression.
*/
quoteForClass: function(s) {
/* Based on ECMA-262, 5th ed., 7.8.5 & 15.10.1. */
return s
.replace(/\\/g, '\\\\') // backslash
.replace(/\0/g, '\\0') // null, IE needs this
.replace(/\//g, '\\/') // closing slash
.replace(/]/g, '\\]') // closing bracket
.replace(/-/g, '\\-') // dash
.replace(/\r/g, '\\r') // carriage return
.replace(/\u2028/g, '\\u2028') // line separator
.replace(/\u2029/g, '\\u2029') // paragraph separator
.replace(/\n/g, '\\n') // line feed
}
};
/* ===== PEG.Compiler ===== */
PEG.Compiler = {
/*
* Takes parts of code, interpolates variables inside them and joins them with
* a newline.
*
* Variables are delimited with "${" and "}" and their names must be valid
* identifiers (i.e. they must match [a-zA-Z_][a-zA-Z0-9_]*). Variable values
* are specified as properties of the last parameter (if this is an object,
* otherwise empty variable set is assumed). Undefined variables result in
* throwing |Error|.
*
* There can be a filter specified after the variable name, prefixed with "|".
* The filter name must be a valid identifier. The only recognized filter
* right now is "string", which quotes the variable value as a JavaScript
* string. Unrecognized filters result in throwing |Error|.
*
* If any part has multiple lines and the first line is indented by some
* amount of whitespace (as defined by the /\s+/ JavaScript regular
* expression), second to last lines are indented by the same amount of
* whitespace. This results in nicely indented multiline code in variables
* without making the templates look ugly.
*
* Examples:
*
* PEG.Compiler.formatCode("foo", "bar"); // "foo\nbar"
* PEG.Compiler.formatCode(
* "foo", "${bar}",
* { bar: "baz" }
* ); // "foo\nbaz"
* PEG.Compiler.formatCode("foo", "${bar}"); // throws Error
* PEG.Compiler.formatCode(
* "foo", "${bar|string}",
* { bar: "baz" }
* ); // "foo\n\"baz\""
* PEG.Compiler.formatCode(
* "foo", "${bar|eeek}",
* { bar: "baz" }
* ); // throws Error
* PEG.Compiler.formatCode(
* "foo", "${bar}",
* { bar: " baz\nqux" }
* ); // "foo\n baz\n qux"
*/
formatCode: function() {
function interpolateVariablesInParts(parts) {
return PEG.ArrayUtils.map(parts, function(part) {
return part.replace(
/\$\{([a-zA-Z_][a-zA-Z0-9_]*)(\|([a-zA-Z_][a-zA-Z0-9_]*))?\}/g,
function(match, name, dummy, filter) {
var value = vars[name];
if (typeof(value) === "undefined") {
throw new Error("Undefined variable: \"" + name + "\".");
}
if (typeof(filter) !== "undefined" && filter != "") { // JavaScript engines differ here.
if (filter === "string") {
return PEG.StringUtils.quote(value);
} else {
throw new Error("Unrecognized filter: \"" + filter + "\".");
}
} else {
return value;
}
}
);
});
}
function indentMultilineParts(parts) {
return PEG.ArrayUtils.map(parts, function(part) {
if (!/\n/.test(part)) { return part; }
var firstLineWhitespacePrefix = part.match(/^\s*/)[0];
var lines = part.split("\n");
var linesIndented = [lines[0]].concat(
PEG.ArrayUtils.map(lines.slice(1), function(line) {
return firstLineWhitespacePrefix + line;
})
);
return linesIndented.join("\n");
});
}
var args = Array.prototype.slice.call(arguments);
var vars = args[args.length - 1] instanceof Object ? args.pop() : {};
return indentMultilineParts(interpolateVariablesInParts(args)).join("\n");
},
_uniqueIdentifierCounters: {},
_resetUniqueIdentifierCounters: function() {
this._uniqueIdentifierCounters = {};
},
/* Generates a unique identifier with specified prefix. */
generateUniqueIdentifier: function(prefix) {
this._uniqueIdentifierCounters[prefix]
= this._uniqueIdentifierCounters[prefix] || 0;
return prefix + this._uniqueIdentifierCounters[prefix]++;
},
/*
* Checks made on the grammar AST before compilation. Each check is a function
* that is passed the AST and start rule and does not return anything. If the
* check passes, the function does not do anything special, otherwise it
* throws |PEG.GrammarError|. The checks are run in sequence in order of their
* definition.
*/
_checks: [
/* Checks that all referenced rules exist. */
function(ast, startRule) {
function nop() {}
function checkExpression(node) { check(node.expression); }
function checkSubnodes(propertyName) {
return function(node) {
PEG.ArrayUtils.each(node[propertyName], check);
};
}
var checkFunctions = {
rule: checkExpression,
choice: checkSubnodes("alternatives"),
sequence: checkSubnodes("elements"),
and_predicate: checkExpression,
not_predicate: checkExpression,
optional: checkExpression,
zero_or_more: checkExpression,
one_or_more: checkExpression,
action: checkExpression,
rule_ref:
function(node) {
if (typeof(ast[node.name]) === "undefined") {
throw new PEG.GrammarError(
"Referenced rule \"" + node.name + "\" does not exist."
);
}
},
literal: nop,
any: nop,
"class": nop
};
function check(node) { checkFunctions[node.type](node); }
for (var rule in ast) {
check(ast[rule]);
}
},
/* Checks that the start rule is defined. */
function(ast, startRule) {
if (typeof(ast[startRule]) === "undefined") {
throw new PEG.GrammarError(
"Missing \"" + startRule + "\" rule."
);
}
},
/* Checks that no left recursion is present. */
function(ast, startRule) {
function nop() {}
function checkExpression(node, appliedRules) {
check(node.expression, appliedRules);
}
var checkFunctions = {
rule:
function(node, appliedRules) {
check(node.expression, appliedRules.concat(node.name));
},
choice:
function(node, appliedRules) {
PEG.ArrayUtils.each(node.alternatives, function(alternative) {
check(alternative, appliedRules);
});
},
sequence:
function(node, appliedRules) {
if (node.elements.length > 0) {
check(node.elements[0], appliedRules);
}
},
and_predicate: checkExpression,
not_predicate: checkExpression,
optional: checkExpression,
zero_or_more: checkExpression,
one_or_more: checkExpression,
action: checkExpression,
rule_ref:
function(node, appliedRules) {
if (PEG.ArrayUtils.contains(appliedRules, node.name)) {
throw new PEG.GrammarError(
"Left recursion detected for rule \"" + node.name + "\"."
);
}
check(ast[node.name], appliedRules);
},
literal: nop,
any: nop,
"class": nop
};
function check(node, appliedRules) {
checkFunctions[node.type](node, appliedRules);
}
for (var rule in ast) {
check(ast[rule], []);
}
}
],
/*
* Optimalization passes made on the grammar AST before compilation. Each pass
* is a function that is passed the AST and start rule and returns a new AST
* and start rule. The AST can be modified in-place by the pass. The passes
* are run in sequence in order of their definition.
*/
_passes: [
/*
* Removes proxy rules -- that is, rules that only delegate to other rule.
*/
function(ast, startRule) {
function isProxyRule(node) {
return node.type === "rule" && node.expression.type === "rule_ref";
}
function replaceRuleRefs(ast, from, to) {
function nop() {}
function replaceInExpression(node, from, to) {
replace(node.expression, from, to);
}
function replaceInSubnodes(propertyName) {
return function(node, from, to) {
PEG.ArrayUtils.each(node[propertyName], function(node) {
replace(node, from, to);
});
};
}
var replaceFunctions = {
rule: replaceInExpression,
choice: replaceInSubnodes("alternatives"),
sequence: replaceInSubnodes("elements"),
and_predicate: replaceInExpression,
not_predicate: replaceInExpression,
optional: replaceInExpression,
zero_or_more: replaceInExpression,
one_or_more: replaceInExpression,
action: replaceInExpression,
rule_ref:
function(node, from, to) {
if (node.name === from) {
node.name = to;
}
},
literal: nop,
any: nop,
"class": nop
};
function replace(node, from, to) {
replaceFunctions[node.type](node, from, to);
}
for (var rule in ast) {
replace(ast[rule], from, to);
}
}
for (var rule in ast) {
if (isProxyRule(ast[rule])) {
replaceRuleRefs(ast, ast[rule].name, ast[rule].expression.name);
if (rule === startRule) {
startRule = ast[rule].expression.name;
}
delete ast[rule];
}
}
return [ast, startRule];
}
],
_compileFunctions: {
rule: function(node) {
var resultVar = PEG.Compiler.generateUniqueIdentifier("result");
if (node.displayName !== null) {
var setReportMatchFailuresCode = PEG.Compiler.formatCode(
"var savedReportMatchFailures = context.reportMatchFailures;",
"context.reportMatchFailures = false;"
);
var restoreReportMatchFailuresCode = PEG.Compiler.formatCode(
"context.reportMatchFailures = savedReportMatchFailures;"
);
var reportMatchFailureCode = PEG.Compiler.formatCode(
"if (context.reportMatchFailures && ${resultVar} === null) {",
" this._matchFailed(${displayName|string});",
"}",
{
displayName: node.displayName,
resultVar: resultVar
}
);
} else {
var setReportMatchFailuresCode = "";
var restoreReportMatchFailuresCode = "";
var reportMatchFailureCode = "";
}
return PEG.Compiler.formatCode(
"_parse_${name}: function(context) {",
" var cacheKey = ${name|string} + '@' + this._pos;",
" var cachedResult = this._cache[cacheKey];",
" if (cachedResult) {",
" this._pos = cachedResult.nextPos;",
" return cachedResult.result;",
" }",
" ",
" var pos = this._pos;",
" ",
" ${setReportMatchFailuresCode}",
" ${code}",
" ${restoreReportMatchFailuresCode}",
" ${reportMatchFailureCode}",
" ",
" this._cache[cacheKey] = {",
" nextPos: this._pos,",
" result: ${resultVar}",
" };",
" return ${resultVar};",
"},",
{
name: node.name,
setReportMatchFailuresCode: setReportMatchFailuresCode,
restoreReportMatchFailuresCode: restoreReportMatchFailuresCode,
reportMatchFailureCode: reportMatchFailureCode,
code: PEG.Compiler.compileNode(node.expression, resultVar),
resultVar: resultVar
}
);
},
/*
* The contract for all code fragments generated by the following functions
* is as follows:
*
* * The code fragment should try to match a part of the input starting with
* the position indicated in |this._pos|. That position may point past the
* end of the input.
*
* * If the code fragment matches the input, it advances |this._pos| after
* the matched part of the input and sets variable with a name stored in
* |resultVar| to appropriate value, which is always non-null.
*
* * If the code fragment does not match the input, it does not change
* |this._pos| and it sets a variable with a name stored in |resultVar| to
* |null|.
*/
choice: function(node, resultVar) {
var code = PEG.Compiler.formatCode(
"var ${resultVar} = null;",
{ resultVar: resultVar }
);
for (var i = node.alternatives.length - 1; i >= 0; i--) {
var alternativeResultVar = PEG.Compiler.generateUniqueIdentifier("result");
code = PEG.Compiler.formatCode(
"${alternativeCode}",
"if (${alternativeResultVar} !== null) {",
" var ${resultVar} = ${alternativeResultVar};",
"} else {",
" ${code};",
"}",
{
alternativeCode: PEG.Compiler.compileNode(node.alternatives[i], alternativeResultVar),
alternativeResultVar: alternativeResultVar,
code: code,
resultVar: resultVar
}
);
}
return code;
},
sequence: function(node, resultVar) {
var savedPosVar = PEG.Compiler.generateUniqueIdentifier("savedPos");
var elementResultVars = PEG.ArrayUtils.map(node.elements, function() {
return PEG.Compiler.generateUniqueIdentifier("result")
});
var code = PEG.Compiler.formatCode(
"var ${resultVar} = ${elementResultVarArray};",
{
resultVar: resultVar,
elementResultVarArray: "[" + elementResultVars.join(", ") + "]"
}
);
for (var i = node.elements.length - 1; i >= 0; i--) {
code = PEG.Compiler.formatCode(
"${elementCode}",
"if (${elementResultVar} !== null) {",
" ${code}",
"} else {",
" var ${resultVar} = null;",
" this._pos = ${savedPosVar};",
"}",
{
elementCode: PEG.Compiler.compileNode(node.elements[i], elementResultVars[i]),
elementResultVar: elementResultVars[i],
code: code,
savedPosVar: savedPosVar,
resultVar: resultVar
}
);
}
return PEG.Compiler.formatCode(
"var ${savedPosVar} = this._pos;",
"${code}",
{
code: code,
savedPosVar: savedPosVar
}
);
},
and_predicate: function(node, resultVar) {
var savedPosVar = PEG.Compiler.generateUniqueIdentifier("savedPos");
var savedReportMatchFailuresVar = PEG.Compiler.generateUniqueIdentifier("savedReportMatchFailuresVar");
var expressionResultVar = PEG.Compiler.generateUniqueIdentifier("result");
return PEG.Compiler.formatCode(
"var ${savedPosVar} = this._pos;",
"var ${savedReportMatchFailuresVar} = context.reportMatchFailures;",
"context.reportMatchFailures = false;",
"${expressionCode}",
"context.reportMatchFailures = ${savedReportMatchFailuresVar};",
"if (${expressionResultVar} !== null) {",
" var ${resultVar} = '';",
" this._pos = ${savedPosVar};",
"} else {",
" var ${resultVar} = null;",
"}",
{
expressionCode: PEG.Compiler.compileNode(node.expression, expressionResultVar),
expressionResultVar: expressionResultVar,
savedPosVar: savedPosVar,
savedReportMatchFailuresVar: savedReportMatchFailuresVar,
resultVar: resultVar
}
);
},
not_predicate: function(node, resultVar) {
var savedPosVar = PEG.Compiler.generateUniqueIdentifier("savedPos");
var savedReportMatchFailuresVar = PEG.Compiler.generateUniqueIdentifier("savedReportMatchFailuresVar");
var expressionResultVar = PEG.Compiler.generateUniqueIdentifier("result");
return PEG.Compiler.formatCode(
"var ${savedPosVar} = this._pos;",
"var ${savedReportMatchFailuresVar} = context.reportMatchFailures;",
"context.reportMatchFailures = false;",
"${expressionCode}",
"context.reportMatchFailures = ${savedReportMatchFailuresVar};",
"if (${expressionResultVar} === null) {",
" var ${resultVar} = '';",
"} else {",
" var ${resultVar} = null;",
" this._pos = ${savedPosVar};",
"}",
{
expressionCode: PEG.Compiler.compileNode(node.expression, expressionResultVar),
expressionResultVar: expressionResultVar,
savedPosVar: savedPosVar,
savedReportMatchFailuresVar: savedReportMatchFailuresVar,
resultVar: resultVar
}
);
},
optional: function(node, resultVar) {
var expressionResultVar = PEG.Compiler.generateUniqueIdentifier("result");
return PEG.Compiler.formatCode(
"${expressionCode}",
"var ${resultVar} = ${expressionResultVar} !== null ? ${expressionResultVar} : '';",
{
expressionCode: PEG.Compiler.compileNode(node.expression, expressionResultVar),
expressionResultVar: expressionResultVar,
resultVar: resultVar
}
);
},
zero_or_more: function(node, resultVar) {
var expressionResultVar = PEG.Compiler.generateUniqueIdentifier("result");
return PEG.Compiler.formatCode(
"var ${resultVar} = [];",
"${expressionCode}",
"while (${expressionResultVar} !== null) {",
" ${resultVar}.push(${expressionResultVar});",
" ${expressionCode}",
"}",
{
expressionCode: PEG.Compiler.compileNode(node.expression, expressionResultVar),
expressionResultVar: expressionResultVar,
resultVar: resultVar
}
);
},
one_or_more: function(node, resultVar) {
var expressionResultVar = PEG.Compiler.generateUniqueIdentifier("result");
return PEG.Compiler.formatCode(
"${expressionCode}",
"if (${expressionResultVar} !== null) {",
" var ${resultVar} = [];",
" while (${expressionResultVar} !== null) {",
" ${resultVar}.push(${expressionResultVar});",
" ${expressionCode}",
" }",
"} else {",
" var ${resultVar} = null;",
"}",
{
expressionCode: PEG.Compiler.compileNode(node.expression, expressionResultVar),
expressionResultVar: expressionResultVar,
resultVar: resultVar
}
);
},
action: function(node, resultVar) {
/*
* In case of sequences, we splat their elements into function arguments
* one by one. Example:
*
* start: "a" "b" "c" { alert(arguments.length) } // => 3
*
* This behavior is reflected in this function.
*/
var expressionResultVar = PEG.Compiler.generateUniqueIdentifier("result");
if (node.expression.type === "sequence") {
var params = PEG.ArrayUtils.map(
PEG.ArrayUtils.range(1, node.expression.elements.length + 1),
function(n) { return "$" + n; }
).join(", ");
var invocationCode = PEG.Compiler.formatCode(
"(function(${params}) { ${action} }).apply(null, ${expressionResultVar})",
{
params: params,
action: node.action,
expressionResultVar: expressionResultVar
}
);
} else {
var invocationCode = PEG.Compiler.formatCode(
"(function($1) { ${action} })(${expressionResultVar})",
{
action: node.action,
expressionResultVar: expressionResultVar
}
);
}
return PEG.Compiler.formatCode(
"${expressionCode}",
"var ${resultVar} = ${expressionResultVar} !== null",
" ? ${invocationCode}",
" : null;",
{
expressionCode: PEG.Compiler.compileNode(node.expression, expressionResultVar),
expressionResultVar: expressionResultVar,
invocationCode: invocationCode,
resultVar: resultVar
}
);
},
rule_ref: function(node, resultVar) {
return PEG.Compiler.formatCode(
"var ${resultVar} = this.${ruleMethod}(context);",
{
ruleMethod: "_parse_" + node.name,
resultVar: resultVar
}
);
},
literal: function(node, resultVar) {
return PEG.Compiler.formatCode(
"if (this._input.substr(this._pos, ${length}) === ${value|string}) {",
" var ${resultVar} = ${value|string};",
" this._pos += ${length};",
"} else {",
" var ${resultVar} = null;",
" if (context.reportMatchFailures) {",
" this._matchFailed(this._quoteString(${value|string}));",
" }",
"}",
{
value: node.value,
length: node.value.length,
resultVar: resultVar
}
);
},
any: function(node, resultVar) {
return PEG.Compiler.formatCode(
"if (this._input.length > this._pos) {",
" var ${resultVar} = this._input.charAt(this._pos);",
" this._pos++;",
"} else {",
" var ${resultVar} = null;",
" if (context.reportMatchFailures) {",
" this._matchFailed('any character');",
" }",
"}",
{ resultVar: resultVar }
);
},
"class": function(node, resultVar) {
if (node.parts.length > 0) {
var regexp = "/^["
+ (node.inverted ? "^" : "")
+ PEG.ArrayUtils.map(node.parts, function(part) {
return part instanceof Array
? PEG.RegExpUtils.quoteForClass(part[0])
+ "-"
+ PEG.RegExpUtils.quoteForClass(part[1])
: PEG.RegExpUtils.quoteForClass(part);
}).join("")
+ "]/";
} else {
/*
* Stupid IE considers regexps /[]/ and /[^]/ syntactically invalid, so
* we translate them into euqivalents it can handle.
*/
var regexp = node.inverted ? "/^[\\S\\s]/" : "/^(?!)/";
}
return PEG.Compiler.formatCode(
"if (this._input.substr(this._pos).match(${regexp}) !== null) {",
" var ${resultVar} = this._input.charAt(this._pos);",
" this._pos++;",
"} else {",
" var ${resultVar} = null;",
" if (context.reportMatchFailures) {",
" this._matchFailed(${rawText|string});",
" }",
"}",
{
regexp: regexp,
rawText: node.rawText,
resultVar: resultVar
}
);
}
},
/*
* Compiles an AST node and returns the generated code. The |resultVar|
* parameter contains a name of variable in which the match result will be
* stored in the generated code.
*/
compileNode: function(node, resultVar) {
return this._compileFunctions[node.type](node, resultVar);
},
/*
* Generates a parser from a specified grammar AST and start rule. Throws
* |PEG.GrammarError| if the AST contains a semantic error. Note that not all
* errors are detected during the generation and some may protrude to the
* generated parser and cause its malfunction.
*/
compileParser: function(ast, startRule) {
/*
* This ensures that the same grammar and start rule always generate exactly
* the same parser.
*/
this._resetUniqueIdentifierCounters();
for (var i = 0; i < this._checks.length; i++) {
this._checks[i](ast, startRule);
}
for (var i = 0; i < this._passes.length; i++) {
var newAstNadStartRule = this._passes[i](ast, startRule);
ast = newAstNadStartRule[0];
startRule = newAstNadStartRule[1];
}
var parseFunctionDefinitions = [];
for (var rule in ast) {
parseFunctionDefinitions.push(this.compileNode(ast[rule]));
}
var source = this.formatCode(
"(function(){",
" /* Generated by PEG.js (http://pegjs.majda.cz/). */",
" ",
" var result = {",
" _startRule: ${startRule|string},",
" ",
/* This needs to be in sync with PEG.StringUtils.quote. */
" _quoteString: function(s) {",
" /*",
" * ECMA-262, 5th ed., 7.8.4: All characters may appear literally in a string",
" * literal except for the closing quote character, backslash, carriage",
" * return, line separator, paragraph separator, and line feed. Any character",
" * may appear in the form of an escape sequence.",
" */",
" return '\"' + s",
" .replace(/\\\\/g, '\\\\\\\\') // backslash",
" .replace(/\"/g, '\\\\\"') // closing quote character",
" .replace(/\\r/g, '\\\\r') // carriage return",
" .replace(/\\u2028/g, '\\\\u2028') // line separator",
" .replace(/\\u2029/g, '\\\\u2029') // paragraph separator",
" .replace(/\\n/g, '\\\\n') // line feed",
" + '\"';",
" },",
" ",
/* This needs to be in sync with PEG.ArrayUtils.contains. */
" _arrayContains: function(array, value) {",
" /*",
" * Stupid IE does not have Array.prototype.indexOf, otherwise this function",
" * would be a one-liner.",
" */",
" var length = array.length;",
" for (var i = 0; i < length; i++) {",
" if (array[i] === value) {",
" return true;",
" }",
" }",
" return false;",
" },",
" ",
" _matchFailed: function(failure) {",
" if (this._pos < this._rightmostMatchFailuresPos) {",
" return;",
" }",
" ",
" if (this._pos > this._rightmostMatchFailuresPos) {",
" this._rightmostMatchFailuresPos = this._pos;",
" this._rightmostMatchFailuresExpected = [];",
" }",
" ",
" if (!this._arrayContains(this._rightmostMatchFailuresExpected, failure)) {",
" this._rightmostMatchFailuresExpected.push(failure);",
" }",
" },",
" ",
" ${parseFunctionDefinitions}",
" ",
" /*",
" * Parses the input with a generated parser. If the parsing is successfull,",
" * returns a value explicitly or implicitly specified by the grammar from",
" * which the parser was generated (see |PEG.buildParser|). If the parsing is",
" * unsuccessful, throws |PEG.grammarParser.SyntaxError| describing the error.",
" */",
" parse: function(input) {",
" var that = this;",
" ",
" function initialize() {",
" that._input = input;",
" that._pos = 0;",
" that._rightmostMatchFailuresPos = 0;",
" that._rightmostMatchFailuresExpected = [];",
" that._cache = {};",
" }",
" ",
" function buildErrorMessage() {",
" function buildExpected(failuresExpected) {",
" switch (failuresExpected.length) {",
" case 0:",
" return 'end of input';",
" case 1:",
" return failuresExpected[0];",
" default:",
" failuresExpected.sort();",
" return failuresExpected.slice(0, failuresExpected.length - 1).join(', ')",
" + ' or '",
" + failuresExpected[failuresExpected.length - 1];",
" }",
" }",
" ",
" var expected = buildExpected(that._rightmostMatchFailuresExpected);",
" var pos = Math.max(that._pos, that._rightmostMatchFailuresPos);",
" var actual = pos < that._input.length",
" ? that._quoteString(that._input.charAt(pos))",
" : 'end of input';",
" ",
" return 'Expected ' + expected + ' but ' + actual + ' found.';",
" }",
" ",
" function computeErrorPosition() {",
" /*",
" * The first idea was to use |String.split| to break the input up to the",
" * error position along newlines and derive the line and column from",
" * there. However IE's |split| implementation is so broken that it was",
" * enough to prevent it.",
" */",
" ",
" var input = that._input;",
" var pos = that._rightmostMatchFailuresPos;",
" var line = 1;",
" var column = 1;",
" var seenCR = false;",
" ",
" for (var i = 0; i < pos; i++) {",
" var ch = input.charAt(i);",
" if (ch === '\\n') {",
" if (!seenCR) { line++; }",
" column = 1;",
" seenCR = false;",
" } else if (ch === '\\r' | ch === '\\u2028' || ch === '\\u2029') {",
" line++;",
" column = 1;",
" seenCR = true;",
" } else {",
" column++;",
" seenCR = false;",
" }",
" }",
" ",
" return { line: line, column: column };",
" }",
" ",
" initialize();",
" ",
" var initialContext = {",
" reportMatchFailures: true",
" };",
" ",
" var result = this['_parse_' + this._startRule](initialContext);",
" ",
" /*",
" * The parser is now in one of the following three states:",
" *",
" * 1. The parser successfully parsed the whole input.",
" *",
" * - |result !== null|",
" * - |that._pos === input.length|",
" * - |that._rightmostMatchFailuresExpected.length| may or may not contain",
" * something",
" *",
" * 2. The parser successfully parsed only a part of the input.",
" *",
" * - |result !== null|",
" * - |that._pos < input.length|",
" * - |that._rightmostMatchFailuresExpected.length| may or may not contain",
" * something",
" *",
" * 3. The parser did not successfully parse any part of the input.",
" *",
" * - |result === null|",
" * - |that._pos === 0|",
" * - |that._rightmostMatchFailuresExpected.length| contains at least one failure",
" *",
" * All code following this comment (including called functions) must",
" * handle these states.",
" */",
" if (result === null || this._pos !== input.length) {",
" var errorPosition = computeErrorPosition();",
" throw new this.SyntaxError(",
" buildErrorMessage(),",
" errorPosition.line,",
" errorPosition.column",
" );",
" }",
" ",
" return result;",
" },",
" ",
" /* Returns the parser source code. */",
" toSource: function() { return this._source; }",
" };",
" ",
" /* Thrown when a parser encounters a syntax error. */",
" ",
" result.SyntaxError = function(message, line, column) {",
" this.name = 'SyntaxError';",
" this.message = message;",
" this.line = line;",
" this.column = column;",
" };",
" ",
" result.SyntaxError.prototype = Error.prototype;",
" ",
" return result;",
"})()",
{
parseFunctionDefinitions: parseFunctionDefinitions.join("\n\n"),
startRule: startRule
}
);
var result = eval(source);
result._source = source;
return result;
}
};
})();
Reference in New Issue View Git Blame Copy Permalink