Son CV dans un terminal web en Javascript! https://terminal-cv.gregandev.fr
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

592 lines
23 KiB

var TokenStream = require('../common/TokenStream');
var adoptBuffer = require('../common/adopt-buffer');
var constants = require('./const');
var TYPE = constants.TYPE;
var charCodeDefinitions = require('./char-code-definitions');
var isNewline = charCodeDefinitions.isNewline;
var isName = charCodeDefinitions.isName;
var isValidEscape = charCodeDefinitions.isValidEscape;
var isNumberStart = charCodeDefinitions.isNumberStart;
var isIdentifierStart = charCodeDefinitions.isIdentifierStart;
var charCodeCategory = charCodeDefinitions.charCodeCategory;
var isBOM = charCodeDefinitions.isBOM;
var utils = require('./utils');
var cmpStr = utils.cmpStr;
var getNewlineLength = utils.getNewlineLength;
var findWhiteSpaceEnd = utils.findWhiteSpaceEnd;
var consumeEscaped = utils.consumeEscaped;
var consumeName = utils.consumeName;
var consumeNumber = utils.consumeNumber;
var consumeBadUrlRemnants = utils.consumeBadUrlRemnants;
var OFFSET_MASK = 0x00FFFFFF;
var TYPE_SHIFT = 24;
function tokenize(source, stream) {
function getCharCode(offset) {
return offset < sourceLength ? source.charCodeAt(offset) : 0;
}
// ยง 4.3.3. Consume a numeric token
function consumeNumericToken() {
// Consume a number and let number be the result.
offset = consumeNumber(source, offset);
// If the next 3 input code points would start an identifier, then:
if (isIdentifierStart(getCharCode(offset), getCharCode(offset + 1), getCharCode(offset + 2))) {
// Create a <dimension-token> with the same value and type flag as number, and a unit set initially to the empty string.
// Consume a name. Set the <dimension-token>โ€™s unit to the returned value.
// Return the <dimension-token>.
type = TYPE.Dimension;
offset = consumeName(source, offset);
return;
}
// Otherwise, if the next input code point is U+0025 PERCENTAGE SIGN (%), consume it.
if (getCharCode(offset) === 0x0025) {
// Create a <percentage-token> with the same value as number, and return it.
type = TYPE.Percentage;
offset++;
return;
}
// Otherwise, create a <number-token> with the same value and type flag as number, and return it.
type = TYPE.Number;
}
// ยง 4.3.4. Consume an ident-like token
function consumeIdentLikeToken() {
const nameStartOffset = offset;
// Consume a name, and let string be the result.
offset = consumeName(source, offset);
// If stringโ€™s value is an ASCII case-insensitive match for "url",
// and the next input code point is U+0028 LEFT PARENTHESIS ((), consume it.
if (cmpStr(source, nameStartOffset, offset, 'url') && getCharCode(offset) === 0x0028) {
// While the next two input code points are whitespace, consume the next input code point.
offset = findWhiteSpaceEnd(source, offset + 1);
// If the next one or two input code points are U+0022 QUOTATION MARK ("), U+0027 APOSTROPHE ('),
// or whitespace followed by U+0022 QUOTATION MARK (") or U+0027 APOSTROPHE ('),
// then create a <function-token> with its value set to string and return it.
if (getCharCode(offset) === 0x0022 ||
getCharCode(offset) === 0x0027) {
type = TYPE.Function;
offset = nameStartOffset + 4;
return;
}
// Otherwise, consume a url token, and return it.
consumeUrlToken();
return;
}
// Otherwise, if the next input code point is U+0028 LEFT PARENTHESIS ((), consume it.
// Create a <function-token> with its value set to string and return it.
if (getCharCode(offset) === 0x0028) {
type = TYPE.Function;
offset++;
return;
}
// Otherwise, create an <ident-token> with its value set to string and return it.
type = TYPE.Ident;
}
// ยง 4.3.5. Consume a string token
function consumeStringToken(endingCodePoint) {
// This algorithm may be called with an ending code point, which denotes the code point
// that ends the string. If an ending code point is not specified,
// the current input code point is used.
if (!endingCodePoint) {
endingCodePoint = getCharCode(offset++);
}
// Initially create a <string-token> with its value set to the empty string.
type = TYPE.String;
// Repeatedly consume the next input code point from the stream:
for (; offset < source.length; offset++) {
var code = source.charCodeAt(offset);
switch (charCodeCategory(code)) {
// ending code point
case endingCodePoint:
// Return the <string-token>.
offset++;
return;
// EOF
case charCodeCategory.Eof:
// This is a parse error. Return the <string-token>.
return;
// newline
case charCodeCategory.WhiteSpace:
if (isNewline(code)) {
// This is a parse error. Reconsume the current input code point,
// create a <bad-string-token>, and return it.
offset += getNewlineLength(source, offset, code);
type = TYPE.BadString;
return;
}
break;
// U+005C REVERSE SOLIDUS (\)
case 0x005C:
// If the next input code point is EOF, do nothing.
if (offset === source.length - 1) {
break;
}
var nextCode = getCharCode(offset + 1);
// Otherwise, if the next input code point is a newline, consume it.
if (isNewline(nextCode)) {
offset += getNewlineLength(source, offset + 1, nextCode);
} else if (isValidEscape(code, nextCode)) {
// Otherwise, (the stream starts with a valid escape) consume
// an escaped code point and append the returned code point to
// the <string-token>โ€™s value.
offset = consumeEscaped(source, offset) - 1;
}
break;
// anything else
// Append the current input code point to the <string-token>โ€™s value.
}
}
}
// ยง 4.3.6. Consume a url token
// Note: This algorithm assumes that the initial "url(" has already been consumed.
// This algorithm also assumes that itโ€™s being called to consume an "unquoted" value, like url(foo).
// A quoted value, like url("foo"), is parsed as a <function-token>. Consume an ident-like token
// automatically handles this distinction; this algorithm shouldnโ€™t be called directly otherwise.
function consumeUrlToken() {
// Initially create a <url-token> with its value set to the empty string.
type = TYPE.Url;
// Consume as much whitespace as possible.
offset = findWhiteSpaceEnd(source, offset);
// Repeatedly consume the next input code point from the stream:
for (; offset < source.length; offset++) {
var code = source.charCodeAt(offset);
switch (charCodeCategory(code)) {
// U+0029 RIGHT PARENTHESIS ())
case 0x0029:
// Return the <url-token>.
offset++;
return;
// EOF
case charCodeCategory.Eof:
// This is a parse error. Return the <url-token>.
return;
// whitespace
case charCodeCategory.WhiteSpace:
// Consume as much whitespace as possible.
offset = findWhiteSpaceEnd(source, offset);
// If the next input code point is U+0029 RIGHT PARENTHESIS ()) or EOF,
// consume it and return the <url-token>
// (if EOF was encountered, this is a parse error);
if (getCharCode(offset) === 0x0029 || offset >= source.length) {
if (offset < source.length) {
offset++;
}
return;
}
// otherwise, consume the remnants of a bad url, create a <bad-url-token>,
// and return it.
offset = consumeBadUrlRemnants(source, offset);
type = TYPE.BadUrl;
return;
// U+0022 QUOTATION MARK (")
// U+0027 APOSTROPHE (')
// U+0028 LEFT PARENTHESIS (()
// non-printable code point
case 0x0022:
case 0x0027:
case 0x0028:
case charCodeCategory.NonPrintable:
// This is a parse error. Consume the remnants of a bad url,
// create a <bad-url-token>, and return it.
offset = consumeBadUrlRemnants(source, offset);
type = TYPE.BadUrl;
return;
// U+005C REVERSE SOLIDUS (\)
case 0x005C:
// If the stream starts with a valid escape, consume an escaped code point and
// append the returned code point to the <url-token>โ€™s value.
if (isValidEscape(code, getCharCode(offset + 1))) {
offset = consumeEscaped(source, offset) - 1;
break;
}
// Otherwise, this is a parse error. Consume the remnants of a bad url,
// create a <bad-url-token>, and return it.
offset = consumeBadUrlRemnants(source, offset);
type = TYPE.BadUrl;
return;
// anything else
// Append the current input code point to the <url-token>โ€™s value.
}
}
}
if (!stream) {
stream = new TokenStream();
}
// ensure source is a string
source = String(source || '');
var sourceLength = source.length;
var offsetAndType = adoptBuffer(stream.offsetAndType, sourceLength + 1); // +1 because of eof-token
var balance = adoptBuffer(stream.balance, sourceLength + 1);
var tokenCount = 0;
var start = isBOM(getCharCode(0));
var offset = start;
var balanceCloseType = 0;
var balanceStart = 0;
var balancePrev = 0;
// https://drafts.csswg.org/css-syntax-3/#consume-token
// ยง 4.3.1. Consume a token
while (offset < sourceLength) {
var code = source.charCodeAt(offset);
var type = 0;
balance[tokenCount] = sourceLength;
switch (charCodeCategory(code)) {
// whitespace
case charCodeCategory.WhiteSpace:
// Consume as much whitespace as possible. Return a <whitespace-token>.
type = TYPE.WhiteSpace;
offset = findWhiteSpaceEnd(source, offset + 1);
break;
// U+0022 QUOTATION MARK (")
case 0x0022:
// Consume a string token and return it.
consumeStringToken();
break;
// U+0023 NUMBER SIGN (#)
case 0x0023:
// If the next input code point is a name code point or the next two input code points are a valid escape, then:
if (isName(getCharCode(offset + 1)) || isValidEscape(getCharCode(offset + 1), getCharCode(offset + 2))) {
// Create a <hash-token>.
type = TYPE.Hash;
// If the next 3 input code points would start an identifier, set the <hash-token>โ€™s type flag to "id".
// if (isIdentifierStart(getCharCode(offset + 1), getCharCode(offset + 2), getCharCode(offset + 3))) {
// // TODO: set id flag
// }
// Consume a name, and set the <hash-token>โ€™s value to the returned string.
offset = consumeName(source, offset + 1);
// Return the <hash-token>.
} else {
// Otherwise, return a <delim-token> with its value set to the current input code point.
type = TYPE.Delim;
offset++;
}
break;
// U+0027 APOSTROPHE (')
case 0x0027:
// Consume a string token and return it.
consumeStringToken();
break;
// U+0028 LEFT PARENTHESIS (()
case 0x0028:
// Return a <(-token>.
type = TYPE.LeftParenthesis;
offset++;
break;
// U+0029 RIGHT PARENTHESIS ())
case 0x0029:
// Return a <)-token>.
type = TYPE.RightParenthesis;
offset++;
break;
// U+002B PLUS SIGN (+)
case 0x002B:
// If the input stream starts with a number, ...
if (isNumberStart(code, getCharCode(offset + 1), getCharCode(offset + 2))) {
// ... reconsume the current input code point, consume a numeric token, and return it.
consumeNumericToken();
} else {
// Otherwise, return a <delim-token> with its value set to the current input code point.
type = TYPE.Delim;
offset++;
}
break;
// U+002C COMMA (,)
case 0x002C:
// Return a <comma-token>.
type = TYPE.Comma;
offset++;
break;
// U+002D HYPHEN-MINUS (-)
case 0x002D:
// If the input stream starts with a number, reconsume the current input code point, consume a numeric token, and return it.
if (isNumberStart(code, getCharCode(offset + 1), getCharCode(offset + 2))) {
consumeNumericToken();
} else {
// Otherwise, if the next 2 input code points are U+002D HYPHEN-MINUS U+003E GREATER-THAN SIGN (->), consume them and return a <CDC-token>.
if (getCharCode(offset + 1) === 0x002D &&
getCharCode(offset + 2) === 0x003E) {
type = TYPE.CDC;
offset = offset + 3;
} else {
// Otherwise, if the input stream starts with an identifier, ...
if (isIdentifierStart(code, getCharCode(offset + 1), getCharCode(offset + 2))) {
// ... reconsume the current input code point, consume an ident-like token, and return it.
consumeIdentLikeToken();
} else {
// Otherwise, return a <delim-token> with its value set to the current input code point.
type = TYPE.Delim;
offset++;
}
}
}
break;
// U+002E FULL STOP (.)
case 0x002E:
// If the input stream starts with a number, ...
if (isNumberStart(code, getCharCode(offset + 1), getCharCode(offset + 2))) {
// ... reconsume the current input code point, consume a numeric token, and return it.
consumeNumericToken();
} else {
// Otherwise, return a <delim-token> with its value set to the current input code point.
type = TYPE.Delim;
offset++;
}
break;
// U+002F SOLIDUS (/)
case 0x002F:
// If the next two input code point are U+002F SOLIDUS (/) followed by a U+002A ASTERISK (*),
if (getCharCode(offset + 1) === 0x002A) {
// ... consume them and all following code points up to and including the first U+002A ASTERISK (*)
// followed by a U+002F SOLIDUS (/), or up to an EOF code point.
type = TYPE.Comment;
offset = source.indexOf('*/', offset + 2) + 2;
if (offset === 1) {
offset = source.length;
}
} else {
type = TYPE.Delim;
offset++;
}
break;
// U+003A COLON (:)
case 0x003A:
// Return a <colon-token>.
type = TYPE.Colon;
offset++;
break;
// U+003B SEMICOLON (;)
case 0x003B:
// Return a <semicolon-token>.
type = TYPE.Semicolon;
offset++;
break;
// U+003C LESS-THAN SIGN (<)
case 0x003C:
// If the next 3 input code points are U+0021 EXCLAMATION MARK U+002D HYPHEN-MINUS U+002D HYPHEN-MINUS (!--), ...
if (getCharCode(offset + 1) === 0x0021 &&
getCharCode(offset + 2) === 0x002D &&
getCharCode(offset + 3) === 0x002D) {
// ... consume them and return a <CDO-token>.
type = TYPE.CDO;
offset = offset + 4;
} else {
// Otherwise, return a <delim-token> with its value set to the current input code point.
type = TYPE.Delim;
offset++;
}
break;
// U+0040 COMMERCIAL AT (@)
case 0x0040:
// If the next 3 input code points would start an identifier, ...
if (isIdentifierStart(getCharCode(offset + 1), getCharCode(offset + 2), getCharCode(offset + 3))) {
// ... consume a name, create an <at-keyword-token> with its value set to the returned value, and return it.
type = TYPE.AtKeyword;
offset = consumeName(source, offset + 1);
} else {
// Otherwise, return a <delim-token> with its value set to the current input code point.
type = TYPE.Delim;
offset++;
}
break;
// U+005B LEFT SQUARE BRACKET ([)
case 0x005B:
// Return a <[-token>.
type = TYPE.LeftSquareBracket;
offset++;
break;
// U+005C REVERSE SOLIDUS (\)
case 0x005C:
// If the input stream starts with a valid escape, ...
if (isValidEscape(code, getCharCode(offset + 1))) {
// ... reconsume the current input code point, consume an ident-like token, and return it.
consumeIdentLikeToken();
} else {
// Otherwise, this is a parse error. Return a <delim-token> with its value set to the current input code point.
type = TYPE.Delim;
offset++;
}
break;
// U+005D RIGHT SQUARE BRACKET (])
case 0x005D:
// Return a <]-token>.
type = TYPE.RightSquareBracket;
offset++;
break;
// U+007B LEFT CURLY BRACKET ({)
case 0x007B:
// Return a <{-token>.
type = TYPE.LeftCurlyBracket;
offset++;
break;
// U+007D RIGHT CURLY BRACKET (})
case 0x007D:
// Return a <}-token>.
type = TYPE.RightCurlyBracket;
offset++;
break;
// digit
case charCodeCategory.Digit:
// Reconsume the current input code point, consume a numeric token, and return it.
consumeNumericToken();
break;
// name-start code point
case charCodeCategory.NameStart:
// Reconsume the current input code point, consume an ident-like token, and return it.
consumeIdentLikeToken();
break;
// EOF
case charCodeCategory.Eof:
// Return an <EOF-token>.
break;
// anything else
default:
// Return a <delim-token> with its value set to the current input code point.
type = TYPE.Delim;
offset++;
}
switch (type) {
case balanceCloseType:
balancePrev = balanceStart & OFFSET_MASK;
balanceStart = balance[balancePrev];
balanceCloseType = balanceStart >> TYPE_SHIFT;
balance[tokenCount] = balancePrev;
balance[balancePrev++] = tokenCount;
for (; balancePrev < tokenCount; balancePrev++) {
if (balance[balancePrev] === sourceLength) {
balance[balancePrev] = tokenCount;
}
}
break;
case TYPE.LeftParenthesis:
case TYPE.Function:
balance[tokenCount] = balanceStart;
balanceCloseType = TYPE.RightParenthesis;
balanceStart = (balanceCloseType << TYPE_SHIFT) | tokenCount;
break;
case TYPE.LeftSquareBracket:
balance[tokenCount] = balanceStart;
balanceCloseType = TYPE.RightSquareBracket;
balanceStart = (balanceCloseType << TYPE_SHIFT) | tokenCount;
break;
case TYPE.LeftCurlyBracket:
balance[tokenCount] = balanceStart;
balanceCloseType = TYPE.RightCurlyBracket;
balanceStart = (balanceCloseType << TYPE_SHIFT) | tokenCount;
break;
}
offsetAndType[tokenCount++] = (type << TYPE_SHIFT) | offset;
}
// finalize buffers
offsetAndType[tokenCount] = (TYPE.EOF << TYPE_SHIFT) | offset; // <EOF-token>
balance[tokenCount] = sourceLength;
balance[sourceLength] = sourceLength; // prevents false positive balance match with any token
while (balanceStart !== 0) {
balancePrev = balanceStart & OFFSET_MASK;
balanceStart = balance[balancePrev];
balance[balancePrev] = sourceLength;
}
// update stream
stream.source = source;
stream.firstCharOffset = start;
stream.offsetAndType = offsetAndType;
stream.tokenCount = tokenCount;
stream.balance = balance;
stream.reset();
stream.next();
return stream;
}
// extend tokenizer with constants
Object.keys(constants).forEach(function(key) {
tokenize[key] = constants[key];
});
// extend tokenizer with static methods from utils
Object.keys(charCodeDefinitions).forEach(function(key) {
tokenize[key] = charCodeDefinitions[key];
});
Object.keys(utils).forEach(function(key) {
tokenize[key] = utils[key];
});
module.exports = tokenize;