@fast-csv/parse
Version:
fast-csv parsing package
99 lines • 4.83 kB
JavaScript
;
Object.defineProperty(exports, "__esModule", { value: true });
exports.QuotedColumnParser = void 0;
const ColumnFormatter_1 = require("./ColumnFormatter");
const Token_1 = require("../Token");
class QuotedColumnParser {
parserOptions;
columnFormatter;
constructor(parserOptions) {
this.parserOptions = parserOptions;
this.columnFormatter = new ColumnFormatter_1.ColumnFormatter(parserOptions);
}
parse(scanner) {
if (!scanner.hasMoreCharacters) {
return null;
}
const originalCursor = scanner.cursor;
const { foundClosingQuote, col } = this.gatherDataBetweenQuotes(scanner);
if (!foundClosingQuote) {
// reset the cursor to the original
scanner.advanceTo(originalCursor);
// if we didnt find a closing quote but we potentially have more data then skip the parsing
// and return the original scanner.
if (!scanner.hasMoreData) {
throw new Error(`Parse Error: missing closing: '${this.parserOptions.quote || ''}' in line: at '${scanner.lineFromCursor.replace(/[\r\n]/g, "\\n'")}'`);
}
return null;
}
this.checkForMalformedColumn(scanner);
return col;
}
gatherDataBetweenQuotes(scanner) {
const { parserOptions } = this;
let foundStartingQuote = false;
let foundClosingQuote = false;
const characters = [];
let nextToken = scanner.nextCharacterToken;
for (; !foundClosingQuote && nextToken !== null; nextToken = scanner.nextCharacterToken) {
const isQuote = Token_1.Token.isTokenQuote(nextToken, parserOptions);
// ignore first quote
if (!foundStartingQuote && isQuote) {
foundStartingQuote = true;
}
else if (foundStartingQuote) {
if (Token_1.Token.isTokenEscapeCharacter(nextToken, parserOptions)) {
// advance past the escape character so we can get the next one in line
scanner.advancePastToken(nextToken);
const tokenFollowingEscape = scanner.nextCharacterToken;
// if the character following the escape is a quote character then just add
// the quote and advance to that character
if (tokenFollowingEscape !== null &&
(Token_1.Token.isTokenQuote(tokenFollowingEscape, parserOptions) ||
Token_1.Token.isTokenEscapeCharacter(tokenFollowingEscape, parserOptions))) {
characters.push(tokenFollowingEscape.token);
nextToken = tokenFollowingEscape;
}
else if (isQuote) {
// if the escape is also a quote then we found our closing quote and finish early
foundClosingQuote = true;
}
else {
// other wise add the escape token to the characters since it wast escaping anything
characters.push(nextToken.token);
}
}
else if (isQuote) {
// we found our closing quote!
foundClosingQuote = true;
}
else {
// add the token to the characters
characters.push(nextToken.token);
}
}
scanner.advancePastToken(nextToken);
}
return { col: this.columnFormatter.format(characters.join('')), foundClosingQuote };
}
checkForMalformedColumn(scanner) {
const { parserOptions } = this;
const { nextNonSpaceToken } = scanner;
if (nextNonSpaceToken) {
const isNextTokenADelimiter = Token_1.Token.isTokenDelimiter(nextNonSpaceToken, parserOptions);
const isNextTokenARowDelimiter = Token_1.Token.isTokenRowDelimiter(nextNonSpaceToken);
if (!(isNextTokenADelimiter || isNextTokenARowDelimiter)) {
// if the final quote was NOT followed by a column (,) or row(\n) delimiter then its a bad column
// tldr: only part of the column was quoted
const linePreview = scanner.lineFromCursor.substr(0, 10).replace(/[\r\n]/g, "\\n'");
throw new Error(`Parse Error: expected: '${parserOptions.escapedDelimiter}' OR new line got: '${nextNonSpaceToken.token}'. at '${linePreview}'`);
}
scanner.advanceToToken(nextNonSpaceToken);
}
else if (!scanner.hasMoreData) {
scanner.advancePastLine();
}
}
}
exports.QuotedColumnParser = QuotedColumnParser;
//# sourceMappingURL=QuotedColumnParser.js.map