printer-notifications/node_modules/sucrase/dist/parser/tokenizer/index.js
2023-11-13 16:10:04 -05:00

1005 lines
34 KiB
JavaScript

"use strict";Object.defineProperty(exports, "__esModule", {value: true}); function _interopRequireDefault(obj) { return obj && obj.__esModule ? obj : { default: obj }; }/* eslint max-len: 0 */
var _base = require('../traverser/base');
var _util = require('../traverser/util');
var _charcodes = require('../util/charcodes');
var _identifier = require('../util/identifier');
var _whitespace = require('../util/whitespace');
var _keywords = require('./keywords');
var _readWord = require('./readWord'); var _readWord2 = _interopRequireDefault(_readWord);
var _types = require('./types');
var IdentifierRole; (function (IdentifierRole) {
const Access = 0; IdentifierRole[IdentifierRole["Access"] = Access] = "Access";
const ExportAccess = Access + 1; IdentifierRole[IdentifierRole["ExportAccess"] = ExportAccess] = "ExportAccess";
const TopLevelDeclaration = ExportAccess + 1; IdentifierRole[IdentifierRole["TopLevelDeclaration"] = TopLevelDeclaration] = "TopLevelDeclaration";
const FunctionScopedDeclaration = TopLevelDeclaration + 1; IdentifierRole[IdentifierRole["FunctionScopedDeclaration"] = FunctionScopedDeclaration] = "FunctionScopedDeclaration";
const BlockScopedDeclaration = FunctionScopedDeclaration + 1; IdentifierRole[IdentifierRole["BlockScopedDeclaration"] = BlockScopedDeclaration] = "BlockScopedDeclaration";
const ObjectShorthandTopLevelDeclaration = BlockScopedDeclaration + 1; IdentifierRole[IdentifierRole["ObjectShorthandTopLevelDeclaration"] = ObjectShorthandTopLevelDeclaration] = "ObjectShorthandTopLevelDeclaration";
const ObjectShorthandFunctionScopedDeclaration = ObjectShorthandTopLevelDeclaration + 1; IdentifierRole[IdentifierRole["ObjectShorthandFunctionScopedDeclaration"] = ObjectShorthandFunctionScopedDeclaration] = "ObjectShorthandFunctionScopedDeclaration";
const ObjectShorthandBlockScopedDeclaration = ObjectShorthandFunctionScopedDeclaration + 1; IdentifierRole[IdentifierRole["ObjectShorthandBlockScopedDeclaration"] = ObjectShorthandBlockScopedDeclaration] = "ObjectShorthandBlockScopedDeclaration";
const ObjectShorthand = ObjectShorthandBlockScopedDeclaration + 1; IdentifierRole[IdentifierRole["ObjectShorthand"] = ObjectShorthand] = "ObjectShorthand";
// Any identifier bound in an import statement, e.g. both A and b from
// `import A, * as b from 'A';`
const ImportDeclaration = ObjectShorthand + 1; IdentifierRole[IdentifierRole["ImportDeclaration"] = ImportDeclaration] = "ImportDeclaration";
const ObjectKey = ImportDeclaration + 1; IdentifierRole[IdentifierRole["ObjectKey"] = ObjectKey] = "ObjectKey";
// The `foo` in `import {foo as bar} from "./abc";`.
const ImportAccess = ObjectKey + 1; IdentifierRole[IdentifierRole["ImportAccess"] = ImportAccess] = "ImportAccess";
})(IdentifierRole || (exports.IdentifierRole = IdentifierRole = {}));
/**
* Extra information on jsxTagStart tokens, used to determine which of the three
* jsx functions are called in the automatic transform.
*/
var JSXRole; (function (JSXRole) {
// The element is self-closing or has a body that resolves to empty. We
// shouldn't emit children at all in this case.
const NoChildren = 0; JSXRole[JSXRole["NoChildren"] = NoChildren] = "NoChildren";
// The element has a single explicit child, which might still be an arbitrary
// expression like an array. We should emit that expression as the children.
const OneChild = NoChildren + 1; JSXRole[JSXRole["OneChild"] = OneChild] = "OneChild";
// The element has at least two explicitly-specified children or has spread
// children, so child positions are assumed to be "static". We should wrap
// these children in an array.
const StaticChildren = OneChild + 1; JSXRole[JSXRole["StaticChildren"] = StaticChildren] = "StaticChildren";
// The element has a prop named "key" after a prop spread, so we should fall
// back to the createElement function.
const KeyAfterPropSpread = StaticChildren + 1; JSXRole[JSXRole["KeyAfterPropSpread"] = KeyAfterPropSpread] = "KeyAfterPropSpread";
})(JSXRole || (exports.JSXRole = JSXRole = {}));
function isDeclaration(token) {
const role = token.identifierRole;
return (
role === IdentifierRole.TopLevelDeclaration ||
role === IdentifierRole.FunctionScopedDeclaration ||
role === IdentifierRole.BlockScopedDeclaration ||
role === IdentifierRole.ObjectShorthandTopLevelDeclaration ||
role === IdentifierRole.ObjectShorthandFunctionScopedDeclaration ||
role === IdentifierRole.ObjectShorthandBlockScopedDeclaration
);
} exports.isDeclaration = isDeclaration;
function isNonTopLevelDeclaration(token) {
const role = token.identifierRole;
return (
role === IdentifierRole.FunctionScopedDeclaration ||
role === IdentifierRole.BlockScopedDeclaration ||
role === IdentifierRole.ObjectShorthandFunctionScopedDeclaration ||
role === IdentifierRole.ObjectShorthandBlockScopedDeclaration
);
} exports.isNonTopLevelDeclaration = isNonTopLevelDeclaration;
function isTopLevelDeclaration(token) {
const role = token.identifierRole;
return (
role === IdentifierRole.TopLevelDeclaration ||
role === IdentifierRole.ObjectShorthandTopLevelDeclaration ||
role === IdentifierRole.ImportDeclaration
);
} exports.isTopLevelDeclaration = isTopLevelDeclaration;
function isBlockScopedDeclaration(token) {
const role = token.identifierRole;
// Treat top-level declarations as block scope since the distinction doesn't matter here.
return (
role === IdentifierRole.TopLevelDeclaration ||
role === IdentifierRole.BlockScopedDeclaration ||
role === IdentifierRole.ObjectShorthandTopLevelDeclaration ||
role === IdentifierRole.ObjectShorthandBlockScopedDeclaration
);
} exports.isBlockScopedDeclaration = isBlockScopedDeclaration;
function isFunctionScopedDeclaration(token) {
const role = token.identifierRole;
return (
role === IdentifierRole.FunctionScopedDeclaration ||
role === IdentifierRole.ObjectShorthandFunctionScopedDeclaration
);
} exports.isFunctionScopedDeclaration = isFunctionScopedDeclaration;
function isObjectShorthandDeclaration(token) {
return (
token.identifierRole === IdentifierRole.ObjectShorthandTopLevelDeclaration ||
token.identifierRole === IdentifierRole.ObjectShorthandBlockScopedDeclaration ||
token.identifierRole === IdentifierRole.ObjectShorthandFunctionScopedDeclaration
);
} exports.isObjectShorthandDeclaration = isObjectShorthandDeclaration;
// Object type used to represent tokens. Note that normally, tokens
// simply exist as properties on the parser object. This is only
// used for the onToken callback and the external tokenizer.
class Token {
constructor() {
this.type = _base.state.type;
this.contextualKeyword = _base.state.contextualKeyword;
this.start = _base.state.start;
this.end = _base.state.end;
this.scopeDepth = _base.state.scopeDepth;
this.isType = _base.state.isType;
this.identifierRole = null;
this.jsxRole = null;
this.shadowsGlobal = false;
this.isAsyncOperation = false;
this.contextId = null;
this.rhsEndIndex = null;
this.isExpression = false;
this.numNullishCoalesceStarts = 0;
this.numNullishCoalesceEnds = 0;
this.isOptionalChainStart = false;
this.isOptionalChainEnd = false;
this.subscriptStartIndex = null;
this.nullishStartIndex = null;
}
// Initially false for all tokens, then may be computed in a follow-up step that does scope
// analysis.
// Initially false for all tokens, but may be set during transform to mark it as containing an
// await operation.
// For assignments, the index of the RHS. For export tokens, the end of the export.
// For class tokens, records if the class is a class expression or a class statement.
// Number of times to insert a `nullishCoalesce(` snippet before this token.
// Number of times to insert a `)` snippet after this token.
// If true, insert an `optionalChain([` snippet before this token.
// If true, insert a `])` snippet after this token.
// Tag for `.`, `?.`, `[`, `?.[`, `(`, and `?.(` to denote the "root" token for this
// subscript chain. This can be used to determine if this chain is an optional chain.
// Tag for `??` operators to denote the root token for this nullish coalescing call.
} exports.Token = Token;
// ## Tokenizer
// Move to the next token
function next() {
_base.state.tokens.push(new Token());
nextToken();
} exports.next = next;
// Call instead of next when inside a template, since that needs to be handled differently.
function nextTemplateToken() {
_base.state.tokens.push(new Token());
_base.state.start = _base.state.pos;
readTmplToken();
} exports.nextTemplateToken = nextTemplateToken;
// The tokenizer never parses regexes by default. Instead, the parser is responsible for
// instructing it to parse a regex when we see a slash at the start of an expression.
function retokenizeSlashAsRegex() {
if (_base.state.type === _types.TokenType.assign) {
--_base.state.pos;
}
readRegexp();
} exports.retokenizeSlashAsRegex = retokenizeSlashAsRegex;
function pushTypeContext(existingTokensInType) {
for (let i = _base.state.tokens.length - existingTokensInType; i < _base.state.tokens.length; i++) {
_base.state.tokens[i].isType = true;
}
const oldIsType = _base.state.isType;
_base.state.isType = true;
return oldIsType;
} exports.pushTypeContext = pushTypeContext;
function popTypeContext(oldIsType) {
_base.state.isType = oldIsType;
} exports.popTypeContext = popTypeContext;
function eat(type) {
if (match(type)) {
next();
return true;
} else {
return false;
}
} exports.eat = eat;
function eatTypeToken(tokenType) {
const oldIsType = _base.state.isType;
_base.state.isType = true;
eat(tokenType);
_base.state.isType = oldIsType;
} exports.eatTypeToken = eatTypeToken;
function match(type) {
return _base.state.type === type;
} exports.match = match;
function lookaheadType() {
const snapshot = _base.state.snapshot();
next();
const type = _base.state.type;
_base.state.restoreFromSnapshot(snapshot);
return type;
} exports.lookaheadType = lookaheadType;
class TypeAndKeyword {
constructor(type, contextualKeyword) {
this.type = type;
this.contextualKeyword = contextualKeyword;
}
} exports.TypeAndKeyword = TypeAndKeyword;
function lookaheadTypeAndKeyword() {
const snapshot = _base.state.snapshot();
next();
const type = _base.state.type;
const contextualKeyword = _base.state.contextualKeyword;
_base.state.restoreFromSnapshot(snapshot);
return new TypeAndKeyword(type, contextualKeyword);
} exports.lookaheadTypeAndKeyword = lookaheadTypeAndKeyword;
function nextTokenStart() {
return nextTokenStartSince(_base.state.pos);
} exports.nextTokenStart = nextTokenStart;
function nextTokenStartSince(pos) {
_whitespace.skipWhiteSpace.lastIndex = pos;
const skip = _whitespace.skipWhiteSpace.exec(_base.input);
return pos + skip[0].length;
} exports.nextTokenStartSince = nextTokenStartSince;
function lookaheadCharCode() {
return _base.input.charCodeAt(nextTokenStart());
} exports.lookaheadCharCode = lookaheadCharCode;
// Read a single token, updating the parser object's token-related
// properties.
function nextToken() {
skipSpace();
_base.state.start = _base.state.pos;
if (_base.state.pos >= _base.input.length) {
const tokens = _base.state.tokens;
// We normally run past the end a bit, but if we're way past the end, avoid an infinite loop.
// Also check the token positions rather than the types since sometimes we rewrite the token
// type to something else.
if (
tokens.length >= 2 &&
tokens[tokens.length - 1].start >= _base.input.length &&
tokens[tokens.length - 2].start >= _base.input.length
) {
_util.unexpected.call(void 0, "Unexpectedly reached the end of input.");
}
finishToken(_types.TokenType.eof);
return;
}
readToken(_base.input.charCodeAt(_base.state.pos));
} exports.nextToken = nextToken;
function readToken(code) {
// Identifier or keyword. '\uXXXX' sequences are allowed in
// identifiers, so '\' also dispatches to that.
if (
_identifier.IS_IDENTIFIER_START[code] ||
code === _charcodes.charCodes.backslash ||
(code === _charcodes.charCodes.atSign && _base.input.charCodeAt(_base.state.pos + 1) === _charcodes.charCodes.atSign)
) {
_readWord2.default.call(void 0, );
} else {
getTokenFromCode(code);
}
}
function skipBlockComment() {
while (
_base.input.charCodeAt(_base.state.pos) !== _charcodes.charCodes.asterisk ||
_base.input.charCodeAt(_base.state.pos + 1) !== _charcodes.charCodes.slash
) {
_base.state.pos++;
if (_base.state.pos > _base.input.length) {
_util.unexpected.call(void 0, "Unterminated comment", _base.state.pos - 2);
return;
}
}
_base.state.pos += 2;
}
function skipLineComment(startSkip) {
let ch = _base.input.charCodeAt((_base.state.pos += startSkip));
if (_base.state.pos < _base.input.length) {
while (
ch !== _charcodes.charCodes.lineFeed &&
ch !== _charcodes.charCodes.carriageReturn &&
ch !== _charcodes.charCodes.lineSeparator &&
ch !== _charcodes.charCodes.paragraphSeparator &&
++_base.state.pos < _base.input.length
) {
ch = _base.input.charCodeAt(_base.state.pos);
}
}
} exports.skipLineComment = skipLineComment;
// Called at the start of the parse and after every token. Skips
// whitespace and comments.
function skipSpace() {
while (_base.state.pos < _base.input.length) {
const ch = _base.input.charCodeAt(_base.state.pos);
switch (ch) {
case _charcodes.charCodes.carriageReturn:
if (_base.input.charCodeAt(_base.state.pos + 1) === _charcodes.charCodes.lineFeed) {
++_base.state.pos;
}
case _charcodes.charCodes.lineFeed:
case _charcodes.charCodes.lineSeparator:
case _charcodes.charCodes.paragraphSeparator:
++_base.state.pos;
break;
case _charcodes.charCodes.slash:
switch (_base.input.charCodeAt(_base.state.pos + 1)) {
case _charcodes.charCodes.asterisk:
_base.state.pos += 2;
skipBlockComment();
break;
case _charcodes.charCodes.slash:
skipLineComment(2);
break;
default:
return;
}
break;
default:
if (_whitespace.IS_WHITESPACE[ch]) {
++_base.state.pos;
} else {
return;
}
}
}
} exports.skipSpace = skipSpace;
// Called at the end of every token. Sets various fields, and skips the space after the token, so
// that the next one's `start` will point at the right position.
function finishToken(
type,
contextualKeyword = _keywords.ContextualKeyword.NONE,
) {
_base.state.end = _base.state.pos;
_base.state.type = type;
_base.state.contextualKeyword = contextualKeyword;
} exports.finishToken = finishToken;
// ### Token reading
// This is the function that is called to fetch the next token. It
// is somewhat obscure, because it works in character codes rather
// than characters, and because operator parsing has been inlined
// into it.
//
// All in the name of speed.
function readToken_dot() {
const nextChar = _base.input.charCodeAt(_base.state.pos + 1);
if (nextChar >= _charcodes.charCodes.digit0 && nextChar <= _charcodes.charCodes.digit9) {
readNumber(true);
return;
}
if (nextChar === _charcodes.charCodes.dot && _base.input.charCodeAt(_base.state.pos + 2) === _charcodes.charCodes.dot) {
_base.state.pos += 3;
finishToken(_types.TokenType.ellipsis);
} else {
++_base.state.pos;
finishToken(_types.TokenType.dot);
}
}
function readToken_slash() {
const nextChar = _base.input.charCodeAt(_base.state.pos + 1);
if (nextChar === _charcodes.charCodes.equalsTo) {
finishOp(_types.TokenType.assign, 2);
} else {
finishOp(_types.TokenType.slash, 1);
}
}
function readToken_mult_modulo(code) {
// '%*'
let tokenType = code === _charcodes.charCodes.asterisk ? _types.TokenType.star : _types.TokenType.modulo;
let width = 1;
let nextChar = _base.input.charCodeAt(_base.state.pos + 1);
// Exponentiation operator **
if (code === _charcodes.charCodes.asterisk && nextChar === _charcodes.charCodes.asterisk) {
width++;
nextChar = _base.input.charCodeAt(_base.state.pos + 2);
tokenType = _types.TokenType.exponent;
}
// Match *= or %=, disallowing *=> which can be valid in flow.
if (
nextChar === _charcodes.charCodes.equalsTo &&
_base.input.charCodeAt(_base.state.pos + 2) !== _charcodes.charCodes.greaterThan
) {
width++;
tokenType = _types.TokenType.assign;
}
finishOp(tokenType, width);
}
function readToken_pipe_amp(code) {
// '|&'
const nextChar = _base.input.charCodeAt(_base.state.pos + 1);
if (nextChar === code) {
if (_base.input.charCodeAt(_base.state.pos + 2) === _charcodes.charCodes.equalsTo) {
// ||= or &&=
finishOp(_types.TokenType.assign, 3);
} else {
// || or &&
finishOp(code === _charcodes.charCodes.verticalBar ? _types.TokenType.logicalOR : _types.TokenType.logicalAND, 2);
}
return;
}
if (code === _charcodes.charCodes.verticalBar) {
// '|>'
if (nextChar === _charcodes.charCodes.greaterThan) {
finishOp(_types.TokenType.pipeline, 2);
return;
} else if (nextChar === _charcodes.charCodes.rightCurlyBrace && _base.isFlowEnabled) {
// '|}'
finishOp(_types.TokenType.braceBarR, 2);
return;
}
}
if (nextChar === _charcodes.charCodes.equalsTo) {
finishOp(_types.TokenType.assign, 2);
return;
}
finishOp(code === _charcodes.charCodes.verticalBar ? _types.TokenType.bitwiseOR : _types.TokenType.bitwiseAND, 1);
}
function readToken_caret() {
// '^'
const nextChar = _base.input.charCodeAt(_base.state.pos + 1);
if (nextChar === _charcodes.charCodes.equalsTo) {
finishOp(_types.TokenType.assign, 2);
} else {
finishOp(_types.TokenType.bitwiseXOR, 1);
}
}
function readToken_plus_min(code) {
// '+-'
const nextChar = _base.input.charCodeAt(_base.state.pos + 1);
if (nextChar === code) {
// Tentatively call this a prefix operator, but it might be changed to postfix later.
finishOp(_types.TokenType.preIncDec, 2);
return;
}
if (nextChar === _charcodes.charCodes.equalsTo) {
finishOp(_types.TokenType.assign, 2);
} else if (code === _charcodes.charCodes.plusSign) {
finishOp(_types.TokenType.plus, 1);
} else {
finishOp(_types.TokenType.minus, 1);
}
}
function readToken_lt() {
const nextChar = _base.input.charCodeAt(_base.state.pos + 1);
if (nextChar === _charcodes.charCodes.lessThan) {
if (_base.input.charCodeAt(_base.state.pos + 2) === _charcodes.charCodes.equalsTo) {
finishOp(_types.TokenType.assign, 3);
return;
}
// We see <<, but need to be really careful about whether to treat it as a
// true left-shift or as two < tokens.
if (_base.state.isType) {
// Within a type, << might come up in a snippet like `Array<<T>() => void>`,
// so treat it as two < tokens. Importantly, this should only override <<
// rather than other tokens like <= . If we treated <= as < in a type
// context, then the snippet `a as T <= 1` would incorrectly start parsing
// a type argument on T. We don't need to worry about `a as T << 1`
// because TypeScript disallows that syntax.
finishOp(_types.TokenType.lessThan, 1);
} else {
// Outside a type, this might be a true left-shift operator, or it might
// still be two open-type-arg tokens, such as in `f<<T>() => void>()`. We
// look at the token while considering the `f`, so we don't yet know that
// we're in a type context. In this case, we initially tokenize as a
// left-shift and correct after-the-fact as necessary in
// tsParseTypeArgumentsWithPossibleBitshift .
finishOp(_types.TokenType.bitShiftL, 2);
}
return;
}
if (nextChar === _charcodes.charCodes.equalsTo) {
// <=
finishOp(_types.TokenType.relationalOrEqual, 2);
} else {
finishOp(_types.TokenType.lessThan, 1);
}
}
function readToken_gt() {
if (_base.state.isType) {
// Avoid right-shift for things like `Array<Array<string>>` and
// greater-than-or-equal for things like `const a: Array<number>=[];`.
finishOp(_types.TokenType.greaterThan, 1);
return;
}
const nextChar = _base.input.charCodeAt(_base.state.pos + 1);
if (nextChar === _charcodes.charCodes.greaterThan) {
const size = _base.input.charCodeAt(_base.state.pos + 2) === _charcodes.charCodes.greaterThan ? 3 : 2;
if (_base.input.charCodeAt(_base.state.pos + size) === _charcodes.charCodes.equalsTo) {
finishOp(_types.TokenType.assign, size + 1);
return;
}
finishOp(_types.TokenType.bitShiftR, size);
return;
}
if (nextChar === _charcodes.charCodes.equalsTo) {
// >=
finishOp(_types.TokenType.relationalOrEqual, 2);
} else {
finishOp(_types.TokenType.greaterThan, 1);
}
}
/**
* Reinterpret a possible > token when transitioning from a type to a non-type
* context.
*
* This comes up in two situations where >= needs to be treated as one token:
* - After an `as` expression, like in the code `a as T >= 1`.
* - In a type argument in an expression context, e.g. `f(a < b, c >= d)`, we
* need to see the token as >= so that we get an error and backtrack to
* normal expression parsing.
*
* Other situations require >= to be seen as two tokens, e.g.
* `const x: Array<T>=[];`, so it's important to treat > as its own token in
* typical type parsing situations.
*/
function rescan_gt() {
if (_base.state.type === _types.TokenType.greaterThan) {
_base.state.pos -= 1;
readToken_gt();
}
} exports.rescan_gt = rescan_gt;
function readToken_eq_excl(code) {
// '=!'
const nextChar = _base.input.charCodeAt(_base.state.pos + 1);
if (nextChar === _charcodes.charCodes.equalsTo) {
finishOp(_types.TokenType.equality, _base.input.charCodeAt(_base.state.pos + 2) === _charcodes.charCodes.equalsTo ? 3 : 2);
return;
}
if (code === _charcodes.charCodes.equalsTo && nextChar === _charcodes.charCodes.greaterThan) {
// '=>'
_base.state.pos += 2;
finishToken(_types.TokenType.arrow);
return;
}
finishOp(code === _charcodes.charCodes.equalsTo ? _types.TokenType.eq : _types.TokenType.bang, 1);
}
function readToken_question() {
// '?'
const nextChar = _base.input.charCodeAt(_base.state.pos + 1);
const nextChar2 = _base.input.charCodeAt(_base.state.pos + 2);
if (
nextChar === _charcodes.charCodes.questionMark &&
// In Flow (but not TypeScript), ??string is a valid type that should be
// tokenized as two individual ? tokens.
!(_base.isFlowEnabled && _base.state.isType)
) {
if (nextChar2 === _charcodes.charCodes.equalsTo) {
// '??='
finishOp(_types.TokenType.assign, 3);
} else {
// '??'
finishOp(_types.TokenType.nullishCoalescing, 2);
}
} else if (
nextChar === _charcodes.charCodes.dot &&
!(nextChar2 >= _charcodes.charCodes.digit0 && nextChar2 <= _charcodes.charCodes.digit9)
) {
// '.' not followed by a number
_base.state.pos += 2;
finishToken(_types.TokenType.questionDot);
} else {
++_base.state.pos;
finishToken(_types.TokenType.question);
}
}
function getTokenFromCode(code) {
switch (code) {
case _charcodes.charCodes.numberSign:
++_base.state.pos;
finishToken(_types.TokenType.hash);
return;
// The interpretation of a dot depends on whether it is followed
// by a digit or another two dots.
case _charcodes.charCodes.dot:
readToken_dot();
return;
// Punctuation tokens.
case _charcodes.charCodes.leftParenthesis:
++_base.state.pos;
finishToken(_types.TokenType.parenL);
return;
case _charcodes.charCodes.rightParenthesis:
++_base.state.pos;
finishToken(_types.TokenType.parenR);
return;
case _charcodes.charCodes.semicolon:
++_base.state.pos;
finishToken(_types.TokenType.semi);
return;
case _charcodes.charCodes.comma:
++_base.state.pos;
finishToken(_types.TokenType.comma);
return;
case _charcodes.charCodes.leftSquareBracket:
++_base.state.pos;
finishToken(_types.TokenType.bracketL);
return;
case _charcodes.charCodes.rightSquareBracket:
++_base.state.pos;
finishToken(_types.TokenType.bracketR);
return;
case _charcodes.charCodes.leftCurlyBrace:
if (_base.isFlowEnabled && _base.input.charCodeAt(_base.state.pos + 1) === _charcodes.charCodes.verticalBar) {
finishOp(_types.TokenType.braceBarL, 2);
} else {
++_base.state.pos;
finishToken(_types.TokenType.braceL);
}
return;
case _charcodes.charCodes.rightCurlyBrace:
++_base.state.pos;
finishToken(_types.TokenType.braceR);
return;
case _charcodes.charCodes.colon:
if (_base.input.charCodeAt(_base.state.pos + 1) === _charcodes.charCodes.colon) {
finishOp(_types.TokenType.doubleColon, 2);
} else {
++_base.state.pos;
finishToken(_types.TokenType.colon);
}
return;
case _charcodes.charCodes.questionMark:
readToken_question();
return;
case _charcodes.charCodes.atSign:
++_base.state.pos;
finishToken(_types.TokenType.at);
return;
case _charcodes.charCodes.graveAccent:
++_base.state.pos;
finishToken(_types.TokenType.backQuote);
return;
case _charcodes.charCodes.digit0: {
const nextChar = _base.input.charCodeAt(_base.state.pos + 1);
// '0x', '0X', '0o', '0O', '0b', '0B'
if (
nextChar === _charcodes.charCodes.lowercaseX ||
nextChar === _charcodes.charCodes.uppercaseX ||
nextChar === _charcodes.charCodes.lowercaseO ||
nextChar === _charcodes.charCodes.uppercaseO ||
nextChar === _charcodes.charCodes.lowercaseB ||
nextChar === _charcodes.charCodes.uppercaseB
) {
readRadixNumber();
return;
}
}
// Anything else beginning with a digit is an integer, octal
// number, or float.
case _charcodes.charCodes.digit1:
case _charcodes.charCodes.digit2:
case _charcodes.charCodes.digit3:
case _charcodes.charCodes.digit4:
case _charcodes.charCodes.digit5:
case _charcodes.charCodes.digit6:
case _charcodes.charCodes.digit7:
case _charcodes.charCodes.digit8:
case _charcodes.charCodes.digit9:
readNumber(false);
return;
// Quotes produce strings.
case _charcodes.charCodes.quotationMark:
case _charcodes.charCodes.apostrophe:
readString(code);
return;
// Operators are parsed inline in tiny state machines. '=' (charCodes.equalsTo) is
// often referred to. `finishOp` simply skips the amount of
// characters it is given as second argument, and returns a token
// of the type given by its first argument.
case _charcodes.charCodes.slash:
readToken_slash();
return;
case _charcodes.charCodes.percentSign:
case _charcodes.charCodes.asterisk:
readToken_mult_modulo(code);
return;
case _charcodes.charCodes.verticalBar:
case _charcodes.charCodes.ampersand:
readToken_pipe_amp(code);
return;
case _charcodes.charCodes.caret:
readToken_caret();
return;
case _charcodes.charCodes.plusSign:
case _charcodes.charCodes.dash:
readToken_plus_min(code);
return;
case _charcodes.charCodes.lessThan:
readToken_lt();
return;
case _charcodes.charCodes.greaterThan:
readToken_gt();
return;
case _charcodes.charCodes.equalsTo:
case _charcodes.charCodes.exclamationMark:
readToken_eq_excl(code);
return;
case _charcodes.charCodes.tilde:
finishOp(_types.TokenType.tilde, 1);
return;
default:
break;
}
_util.unexpected.call(void 0, `Unexpected character '${String.fromCharCode(code)}'`, _base.state.pos);
} exports.getTokenFromCode = getTokenFromCode;
function finishOp(type, size) {
_base.state.pos += size;
finishToken(type);
}
function readRegexp() {
const start = _base.state.pos;
let escaped = false;
let inClass = false;
for (;;) {
if (_base.state.pos >= _base.input.length) {
_util.unexpected.call(void 0, "Unterminated regular expression", start);
return;
}
const code = _base.input.charCodeAt(_base.state.pos);
if (escaped) {
escaped = false;
} else {
if (code === _charcodes.charCodes.leftSquareBracket) {
inClass = true;
} else if (code === _charcodes.charCodes.rightSquareBracket && inClass) {
inClass = false;
} else if (code === _charcodes.charCodes.slash && !inClass) {
break;
}
escaped = code === _charcodes.charCodes.backslash;
}
++_base.state.pos;
}
++_base.state.pos;
// Need to use `skipWord` because '\uXXXX' sequences are allowed here (don't ask).
skipWord();
finishToken(_types.TokenType.regexp);
}
/**
* Read a decimal integer. Note that this can't be unified with the similar code
* in readRadixNumber (which also handles hex digits) because "e" needs to be
* the end of the integer so that we can properly handle scientific notation.
*/
function readInt() {
while (true) {
const code = _base.input.charCodeAt(_base.state.pos);
if ((code >= _charcodes.charCodes.digit0 && code <= _charcodes.charCodes.digit9) || code === _charcodes.charCodes.underscore) {
_base.state.pos++;
} else {
break;
}
}
}
function readRadixNumber() {
_base.state.pos += 2; // 0x
// Walk to the end of the number, allowing hex digits.
while (true) {
const code = _base.input.charCodeAt(_base.state.pos);
if (
(code >= _charcodes.charCodes.digit0 && code <= _charcodes.charCodes.digit9) ||
(code >= _charcodes.charCodes.lowercaseA && code <= _charcodes.charCodes.lowercaseF) ||
(code >= _charcodes.charCodes.uppercaseA && code <= _charcodes.charCodes.uppercaseF) ||
code === _charcodes.charCodes.underscore
) {
_base.state.pos++;
} else {
break;
}
}
const nextChar = _base.input.charCodeAt(_base.state.pos);
if (nextChar === _charcodes.charCodes.lowercaseN) {
++_base.state.pos;
finishToken(_types.TokenType.bigint);
} else {
finishToken(_types.TokenType.num);
}
}
// Read an integer, octal integer, or floating-point number.
function readNumber(startsWithDot) {
let isBigInt = false;
let isDecimal = false;
if (!startsWithDot) {
readInt();
}
let nextChar = _base.input.charCodeAt(_base.state.pos);
if (nextChar === _charcodes.charCodes.dot) {
++_base.state.pos;
readInt();
nextChar = _base.input.charCodeAt(_base.state.pos);
}
if (nextChar === _charcodes.charCodes.uppercaseE || nextChar === _charcodes.charCodes.lowercaseE) {
nextChar = _base.input.charCodeAt(++_base.state.pos);
if (nextChar === _charcodes.charCodes.plusSign || nextChar === _charcodes.charCodes.dash) {
++_base.state.pos;
}
readInt();
nextChar = _base.input.charCodeAt(_base.state.pos);
}
if (nextChar === _charcodes.charCodes.lowercaseN) {
++_base.state.pos;
isBigInt = true;
} else if (nextChar === _charcodes.charCodes.lowercaseM) {
++_base.state.pos;
isDecimal = true;
}
if (isBigInt) {
finishToken(_types.TokenType.bigint);
return;
}
if (isDecimal) {
finishToken(_types.TokenType.decimal);
return;
}
finishToken(_types.TokenType.num);
}
function readString(quote) {
_base.state.pos++;
for (;;) {
if (_base.state.pos >= _base.input.length) {
_util.unexpected.call(void 0, "Unterminated string constant");
return;
}
const ch = _base.input.charCodeAt(_base.state.pos);
if (ch === _charcodes.charCodes.backslash) {
_base.state.pos++;
} else if (ch === quote) {
break;
}
_base.state.pos++;
}
_base.state.pos++;
finishToken(_types.TokenType.string);
}
// Reads template string tokens.
function readTmplToken() {
for (;;) {
if (_base.state.pos >= _base.input.length) {
_util.unexpected.call(void 0, "Unterminated template");
return;
}
const ch = _base.input.charCodeAt(_base.state.pos);
if (
ch === _charcodes.charCodes.graveAccent ||
(ch === _charcodes.charCodes.dollarSign && _base.input.charCodeAt(_base.state.pos + 1) === _charcodes.charCodes.leftCurlyBrace)
) {
if (_base.state.pos === _base.state.start && match(_types.TokenType.template)) {
if (ch === _charcodes.charCodes.dollarSign) {
_base.state.pos += 2;
finishToken(_types.TokenType.dollarBraceL);
return;
} else {
++_base.state.pos;
finishToken(_types.TokenType.backQuote);
return;
}
}
finishToken(_types.TokenType.template);
return;
}
if (ch === _charcodes.charCodes.backslash) {
_base.state.pos++;
}
_base.state.pos++;
}
}
// Skip to the end of the current word. Note that this is the same as the snippet at the end of
// readWord, but calling skipWord from readWord seems to slightly hurt performance from some rough
// measurements.
function skipWord() {
while (_base.state.pos < _base.input.length) {
const ch = _base.input.charCodeAt(_base.state.pos);
if (_identifier.IS_IDENTIFIER_CHAR[ch]) {
_base.state.pos++;
} else if (ch === _charcodes.charCodes.backslash) {
// \u
_base.state.pos += 2;
if (_base.input.charCodeAt(_base.state.pos) === _charcodes.charCodes.leftCurlyBrace) {
while (
_base.state.pos < _base.input.length &&
_base.input.charCodeAt(_base.state.pos) !== _charcodes.charCodes.rightCurlyBrace
) {
_base.state.pos++;
}
_base.state.pos++;
}
} else {
break;
}
}
} exports.skipWord = skipWord;