csv-parse
Version:
CSV parsing implementing the Node.js `stream.Transform` API
728 lines (720 loc) • 28.6 kB
JavaScript
import {normalize_columns_array} from './normalize_columns_array.js';
import {init_state} from './init_state.js';
import {normalize_options} from './normalize_options.js';
import {CsvError} from './CsvError.js';
const isRecordEmpty = function(record){
return record.every((field) => field == null || field.toString && field.toString().trim() === '');
};
const cr = 13; // `\r`, carriage return, 0x0D in hexadécimal, 13 in decimal
const nl = 10; // `\n`, newline, 0x0A in hexadecimal, 10 in decimal
const boms = {
// Note, the following are equals:
// Buffer.from("\ufeff")
// Buffer.from([239, 187, 191])
// Buffer.from('EFBBBF', 'hex')
'utf8': Buffer.from([239, 187, 191]),
// Note, the following are equals:
// Buffer.from "\ufeff", 'utf16le
// Buffer.from([255, 254])
'utf16le': Buffer.from([255, 254])
};
const transform = function(original_options = {}) {
const info = {
bytes: 0,
comment_lines: 0,
empty_lines: 0,
invalid_field_length: 0,
lines: 1,
records: 0
};
const options = normalize_options(original_options);
return {
info: info,
original_options: original_options,
options: options,
state: init_state(options),
__needMoreData: function(i, bufLen, end){
if(end) return false;
const {encoding, escape, quote} = this.options;
const {quoting, needMoreDataSize, recordDelimiterMaxLength} = this.state;
const numOfCharLeft = bufLen - i - 1;
const requiredLength = Math.max(
needMoreDataSize,
// Skip if the remaining buffer smaller than record delimiter
// If "record_delimiter" is yet to be discovered:
// 1. It is equals to `[]` and "recordDelimiterMaxLength" equals `0`
// 2. We set the length to windows line ending in the current encoding
// Note, that encoding is known from user or bom discovery at that point
// recordDelimiterMaxLength,
recordDelimiterMaxLength === 0 ? Buffer.from('\r\n', encoding).length : recordDelimiterMaxLength,
// Skip if remaining buffer can be an escaped quote
quoting ? ((escape === null ? 0 : escape.length) + quote.length) : 0,
// Skip if remaining buffer can be record delimiter following the closing quote
quoting ? (quote.length + recordDelimiterMaxLength) : 0,
);
return numOfCharLeft < requiredLength;
},
// Central parser implementation
parse: function(nextBuf, end, push, close){
const {bom, comment_no_infix, encoding, from_line, ltrim, max_record_size,raw, relax_quotes, rtrim, skip_empty_lines, to, to_line} = this.options;
let {comment, escape, quote, record_delimiter} = this.options;
const {bomSkipped, previousBuf, rawBuffer, escapeIsQuote} = this.state;
let buf;
if(previousBuf === undefined){
if(nextBuf === undefined){
// Handle empty string
close();
return;
}else{
buf = nextBuf;
}
}else if(previousBuf !== undefined && nextBuf === undefined){
buf = previousBuf;
}else{
buf = Buffer.concat([previousBuf, nextBuf]);
}
// Handle UTF BOM
if(bomSkipped === false){
if(bom === false){
this.state.bomSkipped = true;
}else if(buf.length < 3){
// No enough data
if(end === false){
// Wait for more data
this.state.previousBuf = buf;
return;
}
}else{
for(const encoding in boms){
if(boms[encoding].compare(buf, 0, boms[encoding].length) === 0){
// Skip BOM
const bomLength = boms[encoding].length;
this.state.bufBytesStart += bomLength;
buf = buf.slice(bomLength);
// Renormalize original options with the new encoding
this.options = normalize_options({...this.original_options, encoding: encoding});
// Options will re-evaluate the Buffer with the new encoding
({comment, escape, quote } = this.options);
break;
}
}
this.state.bomSkipped = true;
}
}
const bufLen = buf.length;
let pos;
for(pos = 0; pos < bufLen; pos++){
// Ensure we get enough space to look ahead
// There should be a way to move this out of the loop
if(this.__needMoreData(pos, bufLen, end)){
break;
}
if(this.state.wasRowDelimiter === true){
this.info.lines++;
this.state.wasRowDelimiter = false;
}
if(to_line !== -1 && this.info.lines > to_line){
this.state.stop = true;
close();
return;
}
// Auto discovery of record_delimiter, unix, mac and windows supported
if(this.state.quoting === false && record_delimiter.length === 0){
const record_delimiterCount = this.__autoDiscoverRecordDelimiter(buf, pos);
if(record_delimiterCount){
record_delimiter = this.options.record_delimiter;
}
}
const chr = buf[pos];
if(raw === true){
rawBuffer.append(chr);
}
if((chr === cr || chr === nl) && this.state.wasRowDelimiter === false){
this.state.wasRowDelimiter = true;
}
// Previous char was a valid escape char
// treat the current char as a regular char
if(this.state.escaping === true){
this.state.escaping = false;
}else{
// Escape is only active inside quoted fields
// We are quoting, the char is an escape chr and there is a chr to escape
// if(escape !== null && this.state.quoting === true && chr === escape && pos + 1 < bufLen){
if(escape !== null && this.state.quoting === true && this.__isEscape(buf, pos, chr) && pos + escape.length < bufLen){
if(escapeIsQuote){
if(this.__isQuote(buf, pos+escape.length)){
this.state.escaping = true;
pos += escape.length - 1;
continue;
}
}else{
this.state.escaping = true;
pos += escape.length - 1;
continue;
}
}
// Not currently escaping and chr is a quote
// TODO: need to compare bytes instead of single char
if(this.state.commenting === false && this.__isQuote(buf, pos)){
if(this.state.quoting === true){
const nextChr = buf[pos+quote.length];
const isNextChrTrimable = rtrim && this.__isCharTrimable(buf, pos+quote.length);
const isNextChrComment = comment !== null && this.__compareBytes(comment, buf, pos+quote.length, nextChr);
const isNextChrDelimiter = this.__isDelimiter(buf, pos+quote.length, nextChr);
const isNextChrRecordDelimiter = record_delimiter.length === 0 ? this.__autoDiscoverRecordDelimiter(buf, pos+quote.length) : this.__isRecordDelimiter(nextChr, buf, pos+quote.length);
// Escape a quote
// Treat next char as a regular character
if(escape !== null && this.__isEscape(buf, pos, chr) && this.__isQuote(buf, pos + escape.length)){
pos += escape.length - 1;
}else if(!nextChr || isNextChrDelimiter || isNextChrRecordDelimiter || isNextChrComment || isNextChrTrimable){
this.state.quoting = false;
this.state.wasQuoting = true;
pos += quote.length - 1;
continue;
}else if(relax_quotes === false){
const err = this.__error(
new CsvError('CSV_INVALID_CLOSING_QUOTE', [
'Invalid Closing Quote:',
`got "${String.fromCharCode(nextChr)}"`,
`at line ${this.info.lines}`,
'instead of delimiter, record delimiter, trimable character',
'(if activated) or comment',
], this.options, this.__infoField())
);
if(err !== undefined) return err;
}else{
this.state.quoting = false;
this.state.wasQuoting = true;
this.state.field.prepend(quote);
pos += quote.length - 1;
}
}else{
if(this.state.field.length !== 0){
// In relax_quotes mode, treat opening quote preceded by chrs as regular
if(relax_quotes === false){
const info = this.__infoField();
const bom = Object.keys(boms).map(b => boms[b].equals(this.state.field.toString()) ? b : false).filter(Boolean)[0];
const err = this.__error(
new CsvError('INVALID_OPENING_QUOTE', [
'Invalid Opening Quote:',
`a quote is found on field ${JSON.stringify(info.column)} at line ${info.lines}, value is ${JSON.stringify(this.state.field.toString(encoding))}`,
bom ? `(${bom} bom)` : undefined
], this.options, info, {
field: this.state.field,
})
);
if(err !== undefined) return err;
}
}else{
this.state.quoting = true;
pos += quote.length - 1;
continue;
}
}
}
if(this.state.quoting === false){
const recordDelimiterLength = this.__isRecordDelimiter(chr, buf, pos);
if(recordDelimiterLength !== 0){
// Do not emit comments which take a full line
const skipCommentLine = this.state.commenting && (this.state.wasQuoting === false && this.state.record.length === 0 && this.state.field.length === 0);
if(skipCommentLine){
this.info.comment_lines++;
// Skip full comment line
}else{
// Activate records emition if above from_line
if(this.state.enabled === false && this.info.lines + (this.state.wasRowDelimiter === true ? 1: 0) >= from_line){
this.state.enabled = true;
this.__resetField();
this.__resetRecord();
pos += recordDelimiterLength - 1;
continue;
}
// Skip if line is empty and skip_empty_lines activated
if(skip_empty_lines === true && this.state.wasQuoting === false && this.state.record.length === 0 && this.state.field.length === 0){
this.info.empty_lines++;
pos += recordDelimiterLength - 1;
continue;
}
this.info.bytes = this.state.bufBytesStart + pos;
const errField = this.__onField();
if(errField !== undefined) return errField;
this.info.bytes = this.state.bufBytesStart + pos + recordDelimiterLength;
const errRecord = this.__onRecord(push);
if(errRecord !== undefined) return errRecord;
if(to !== -1 && this.info.records >= to){
this.state.stop = true;
close();
return;
}
}
this.state.commenting = false;
pos += recordDelimiterLength - 1;
continue;
}
if(this.state.commenting){
continue;
}
const commentCount = comment === null ? 0 : this.__compareBytes(comment, buf, pos, chr);
if(commentCount !== 0 && (comment_no_infix === false || this.state.field.length === 0)){
this.state.commenting = true;
continue;
}
const delimiterLength = this.__isDelimiter(buf, pos, chr);
if(delimiterLength !== 0){
this.info.bytes = this.state.bufBytesStart + pos;
const errField = this.__onField();
if(errField !== undefined) return errField;
pos += delimiterLength - 1;
continue;
}
}
}
if(this.state.commenting === false){
if(max_record_size !== 0 && this.state.record_length + this.state.field.length > max_record_size){
return this.__error(
new CsvError('CSV_MAX_RECORD_SIZE', [
'Max Record Size:',
'record exceed the maximum number of tolerated bytes',
`of ${max_record_size}`,
`at line ${this.info.lines}`,
], this.options, this.__infoField())
);
}
}
const lappend = ltrim === false || this.state.quoting === true || this.state.field.length !== 0 || !this.__isCharTrimable(buf, pos);
// rtrim in non quoting is handle in __onField
const rappend = rtrim === false || this.state.wasQuoting === false;
if(lappend === true && rappend === true){
this.state.field.append(chr);
}else if(rtrim === true && !this.__isCharTrimable(buf, pos)){
return this.__error(
new CsvError('CSV_NON_TRIMABLE_CHAR_AFTER_CLOSING_QUOTE', [
'Invalid Closing Quote:',
'found non trimable byte after quote',
`at line ${this.info.lines}`,
], this.options, this.__infoField())
);
}else{
if(lappend === false){
pos += this.__isCharTrimable(buf, pos) - 1;
}
continue;
}
}
if(end === true){
// Ensure we are not ending in a quoting state
if(this.state.quoting === true){
const err = this.__error(
new CsvError('CSV_QUOTE_NOT_CLOSED', [
'Quote Not Closed:',
`the parsing is finished with an opening quote at line ${this.info.lines}`,
], this.options, this.__infoField())
);
if(err !== undefined) return err;
}else{
// Skip last line if it has no characters
if(this.state.wasQuoting === true || this.state.record.length !== 0 || this.state.field.length !== 0){
this.info.bytes = this.state.bufBytesStart + pos;
const errField = this.__onField();
if(errField !== undefined) return errField;
const errRecord = this.__onRecord(push);
if(errRecord !== undefined) return errRecord;
}else if(this.state.wasRowDelimiter === true){
this.info.empty_lines++;
}else if(this.state.commenting === true){
this.info.comment_lines++;
}
}
}else{
this.state.bufBytesStart += pos;
this.state.previousBuf = buf.slice(pos);
}
if(this.state.wasRowDelimiter === true){
this.info.lines++;
this.state.wasRowDelimiter = false;
}
},
__onRecord: function(push){
const {columns, group_columns_by_name, encoding, info, from, relax_column_count, relax_column_count_less, relax_column_count_more, raw, skip_records_with_empty_values} = this.options;
const {enabled, record} = this.state;
if(enabled === false){
return this.__resetRecord();
}
// Convert the first line into column names
const recordLength = record.length;
if(columns === true){
if(skip_records_with_empty_values === true && isRecordEmpty(record)){
this.__resetRecord();
return;
}
return this.__firstLineToColumns(record);
}
if(columns === false && this.info.records === 0){
this.state.expectedRecordLength = recordLength;
}
if(recordLength !== this.state.expectedRecordLength){
const err = columns === false ?
new CsvError('CSV_RECORD_INCONSISTENT_FIELDS_LENGTH', [
'Invalid Record Length:',
`expect ${this.state.expectedRecordLength},`,
`got ${recordLength} on line ${this.info.lines}`,
], this.options, this.__infoField(), {
record: record,
})
:
new CsvError('CSV_RECORD_INCONSISTENT_COLUMNS', [
'Invalid Record Length:',
`columns length is ${columns.length},`, // rename columns
`got ${recordLength} on line ${this.info.lines}`,
], this.options, this.__infoField(), {
record: record,
});
if(relax_column_count === true ||
(relax_column_count_less === true && recordLength < this.state.expectedRecordLength) ||
(relax_column_count_more === true && recordLength > this.state.expectedRecordLength)){
this.info.invalid_field_length++;
this.state.error = err;
// Error is undefined with skip_records_with_error
}else{
const finalErr = this.__error(err);
if(finalErr) return finalErr;
}
}
if(skip_records_with_empty_values === true && isRecordEmpty(record)){
this.__resetRecord();
return;
}
if(this.state.recordHasError === true){
this.__resetRecord();
this.state.recordHasError = false;
return;
}
this.info.records++;
if(from === 1 || this.info.records >= from){
const {objname} = this.options;
// With columns, records are object
if(columns !== false){
const obj = {};
// Transform record array to an object
for(let i = 0, l = record.length; i < l; i++){
if(columns[i] === undefined || columns[i].disabled) continue;
// Turn duplicate columns into an array
if (group_columns_by_name === true && obj[columns[i].name] !== undefined) {
if (Array.isArray(obj[columns[i].name])) {
obj[columns[i].name] = obj[columns[i].name].concat(record[i]);
} else {
obj[columns[i].name] = [obj[columns[i].name], record[i]];
}
} else {
obj[columns[i].name] = record[i];
}
}
// Without objname (default)
if(raw === true || info === true){
const extRecord = Object.assign(
{record: obj},
(raw === true ? {raw: this.state.rawBuffer.toString(encoding)}: {}),
(info === true ? {info: this.__infoRecord()}: {})
);
const err = this.__push(
objname === undefined ? extRecord : [obj[objname], extRecord]
, push);
if(err){
return err;
}
}else{
const err = this.__push(
objname === undefined ? obj : [obj[objname], obj]
, push);
if(err){
return err;
}
}
// Without columns, records are array
}else{
if(raw === true || info === true){
const extRecord = Object.assign(
{record: record},
raw === true ? {raw: this.state.rawBuffer.toString(encoding)}: {},
info === true ? {info: this.__infoRecord()}: {}
);
const err = this.__push(
objname === undefined ? extRecord : [record[objname], extRecord]
, push);
if(err){
return err;
}
}else{
const err = this.__push(
objname === undefined ? record : [record[objname], record]
, push);
if(err){
return err;
}
}
}
}
this.__resetRecord();
},
__firstLineToColumns: function(record){
const {firstLineToHeaders} = this.state;
try{
const headers = firstLineToHeaders === undefined ? record : firstLineToHeaders.call(null, record);
if(!Array.isArray(headers)){
return this.__error(
new CsvError('CSV_INVALID_COLUMN_MAPPING', [
'Invalid Column Mapping:',
'expect an array from column function,',
`got ${JSON.stringify(headers)}`
], this.options, this.__infoField(), {
headers: headers,
})
);
}
const normalizedHeaders = normalize_columns_array(headers);
this.state.expectedRecordLength = normalizedHeaders.length;
this.options.columns = normalizedHeaders;
this.__resetRecord();
return;
}catch(err){
return err;
}
},
__resetRecord: function(){
if(this.options.raw === true){
this.state.rawBuffer.reset();
}
this.state.error = undefined;
this.state.record = [];
this.state.record_length = 0;
},
__onField: function(){
const {cast, encoding, rtrim, max_record_size} = this.options;
const {enabled, wasQuoting} = this.state;
// Short circuit for the from_line options
if(enabled === false){
return this.__resetField();
}
let field = this.state.field.toString(encoding);
if(rtrim === true && wasQuoting === false){
field = field.trimRight();
}
if(cast === true){
const [err, f] = this.__cast(field);
if(err !== undefined) return err;
field = f;
}
this.state.record.push(field);
// Increment record length if record size must not exceed a limit
if(max_record_size !== 0 && typeof field === 'string'){
this.state.record_length += field.length;
}
this.__resetField();
},
__resetField: function(){
this.state.field.reset();
this.state.wasQuoting = false;
},
__push: function(record, push){
const {on_record} = this.options;
if(on_record !== undefined){
const info = this.__infoRecord();
try{
record = on_record.call(null, record, info);
}catch(err){
return err;
}
if(record === undefined || record === null){ return; }
}
push(record);
},
// Return a tuple with the error and the casted value
__cast: function(field){
const {columns, relax_column_count} = this.options;
const isColumns = Array.isArray(columns);
// Dont loose time calling cast
// because the final record is an object
// and this field can't be associated to a key present in columns
if(isColumns === true && relax_column_count && this.options.columns.length <= this.state.record.length){
return [undefined, undefined];
}
if(this.state.castField !== null){
try{
const info = this.__infoField();
return [undefined, this.state.castField.call(null, field, info)];
}catch(err){
return [err];
}
}
if(this.__isFloat(field)){
return [undefined, parseFloat(field)];
}else if(this.options.cast_date !== false){
const info = this.__infoField();
return [undefined, this.options.cast_date.call(null, field, info)];
}
return [undefined, field];
},
// Helper to test if a character is a space or a line delimiter
__isCharTrimable: function(buf, pos){
const isTrim = (buf, pos) => {
const {timchars} = this.state;
loop1: for(let i = 0; i < timchars.length; i++){
const timchar = timchars[i];
for(let j = 0; j < timchar.length; j++){
if(timchar[j] !== buf[pos+j]) continue loop1;
}
return timchar.length;
}
return 0;
};
return isTrim(buf, pos);
},
// Keep it in case we implement the `cast_int` option
// __isInt(value){
// // return Number.isInteger(parseInt(value))
// // return !isNaN( parseInt( obj ) );
// return /^(\-|\+)?[1-9][0-9]*$/.test(value)
// }
__isFloat: function(value){
return (value - parseFloat(value) + 1) >= 0; // Borrowed from jquery
},
__compareBytes: function(sourceBuf, targetBuf, targetPos, firstByte){
if(sourceBuf[0] !== firstByte) return 0;
const sourceLength = sourceBuf.length;
for(let i = 1; i < sourceLength; i++){
if(sourceBuf[i] !== targetBuf[targetPos+i]) return 0;
}
return sourceLength;
},
__isDelimiter: function(buf, pos, chr){
const {delimiter, ignore_last_delimiters} = this.options;
if(ignore_last_delimiters === true && this.state.record.length === this.options.columns.length - 1){
return 0;
}else if(ignore_last_delimiters !== false && typeof ignore_last_delimiters === 'number' && this.state.record.length === ignore_last_delimiters - 1){
return 0;
}
loop1: for(let i = 0; i < delimiter.length; i++){
const del = delimiter[i];
if(del[0] === chr){
for(let j = 1; j < del.length; j++){
if(del[j] !== buf[pos+j]) continue loop1;
}
return del.length;
}
}
return 0;
},
__isRecordDelimiter: function(chr, buf, pos){
const {record_delimiter} = this.options;
const recordDelimiterLength = record_delimiter.length;
loop1: for(let i = 0; i < recordDelimiterLength; i++){
const rd = record_delimiter[i];
const rdLength = rd.length;
if(rd[0] !== chr){
continue;
}
for(let j = 1; j < rdLength; j++){
if(rd[j] !== buf[pos+j]){
continue loop1;
}
}
return rd.length;
}
return 0;
},
__isEscape: function(buf, pos, chr){
const {escape} = this.options;
if(escape === null) return false;
const l = escape.length;
if(escape[0] === chr){
for(let i = 0; i < l; i++){
if(escape[i] !== buf[pos+i]){
return false;
}
}
return true;
}
return false;
},
__isQuote: function(buf, pos){
const {quote} = this.options;
if(quote === null) return false;
const l = quote.length;
for(let i = 0; i < l; i++){
if(quote[i] !== buf[pos+i]){
return false;
}
}
return true;
},
__autoDiscoverRecordDelimiter: function(buf, pos){
const { encoding } = this.options;
// Note, we don't need to cache this information in state,
// It is only called on the first line until we find out a suitable
// record delimiter.
const rds = [
// Important, the windows line ending must be before mac os 9
Buffer.from('\r\n', encoding),
Buffer.from('\n', encoding),
Buffer.from('\r', encoding),
];
loop: for(let i = 0; i < rds.length; i++){
const l = rds[i].length;
for(let j = 0; j < l; j++){
if(rds[i][j] !== buf[pos + j]){
continue loop;
}
}
this.options.record_delimiter.push(rds[i]);
this.state.recordDelimiterMaxLength = rds[i].length;
return rds[i].length;
}
return 0;
},
__error: function(msg){
const {encoding, raw, skip_records_with_error} = this.options;
const err = typeof msg === 'string' ? new Error(msg) : msg;
if(skip_records_with_error){
this.state.recordHasError = true;
if(this.options.on_skip !== undefined){
this.options.on_skip(err, raw ? this.state.rawBuffer.toString(encoding) : undefined);
}
// this.emit('skip', err, raw ? this.state.rawBuffer.toString(encoding) : undefined);
return undefined;
}else{
return err;
}
},
__infoDataSet: function(){
return {
...this.info,
columns: this.options.columns
};
},
__infoRecord: function(){
const {columns, raw, encoding} = this.options;
return {
...this.__infoDataSet(),
error: this.state.error,
header: columns === true,
index: this.state.record.length,
raw: raw ? this.state.rawBuffer.toString(encoding) : undefined
};
},
__infoField: function(){
const {columns} = this.options;
const isColumns = Array.isArray(columns);
return {
...this.__infoRecord(),
column: isColumns === true ?
(columns.length > this.state.record.length ?
columns[this.state.record.length].name :
null
) :
this.state.record.length,
quoting: this.state.wasQuoting,
};
}
};
};
export {transform, CsvError};