summaryrefslogtreecommitdiff
path: root/src/pkg/encoding/csv/reader.go
blob: b328dcc375cd7cccf8f680540fe34cc01c791be8 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
// Copyright 2011 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.

// Package csv reads and writes comma-separated values (CSV) files.
//
// A csv file contains zero or more records of one or more fields per record.
// Each record is separated by the newline character. The final record may
// optionally be followed by a newline character.
//
//	field1,field2,field3
//
// White space is considered part of a field.
//
// Carriage returns before newline characters are silently removed.
//
// Blank lines are ignored.  A line with only whitespace characters (excluding
// the ending newline character) is not considered a blank line.
//
// Fields which start and stop with the quote character " are called
// quoted-fields.  The beginning and ending quote are not part of the
// field.
//
// The source:
//
//	normal string,"quoted-field"
//
// results in the fields
//
//	{`normal string`, `quoted-field`}
//
// Within a quoted-field a quote character followed by a second quote
// character is considered a single quote.
//
//	"the ""word"" is true","a ""quoted-field"""
//
// results in
//
//	{`the "word" is true`, `a "quoted-field"`}
//
// Newlines and commas may be included in a quoted-field
//
//	"Multi-line
//	field","comma is ,"
//
// results in
//
//	{`Multi-line
//	field`, `comma is ,`}
package csv

import (
	"bufio"
	"bytes"
	"errors"
	"fmt"
	"io"
	"unicode"
)

// A ParseError is returned for parsing errors.
// The first line is 1.  The first column is 0.
type ParseError struct {
	Line   int   // Line where the error occurred
	Column int   // Column (rune index) where the error occurred
	Err    error // The actual error
}

func (e *ParseError) Error() string {
	return fmt.Sprintf("line %d, column %d: %s", e.Line, e.Column, e.Err)
}

// These are the errors that can be returned in ParseError.Error
var (
	ErrTrailingComma = errors.New("extra delimiter at end of line") // no longer used
	ErrBareQuote     = errors.New("bare \" in non-quoted-field")
	ErrQuote         = errors.New("extraneous \" in field")
	ErrFieldCount    = errors.New("wrong number of fields in line")
)

// A Reader reads records from a CSV-encoded file.
//
// As returned by NewReader, a Reader expects input conforming to RFC 4180.
// The exported fields can be changed to customize the details before the
// first call to Read or ReadAll.
//
// Comma is the field delimiter.  It defaults to ','.
//
// Comment, if not 0, is the comment character. Lines beginning with the
// Comment character are ignored.
//
// If FieldsPerRecord is positive, Read requires each record to
// have the given number of fields.  If FieldsPerRecord is 0, Read sets it to
// the number of fields in the first record, so that future records must
// have the same field count.  If FieldsPerRecord is negative, no check is
// made and records may have a variable number of fields.
//
// If LazyQuotes is true, a quote may appear in an unquoted field and a
// non-doubled quote may appear in a quoted field.
//
// If TrimLeadingSpace is true, leading white space in a field is ignored.
type Reader struct {
	Comma            rune // field delimiter (set to ',' by NewReader)
	Comment          rune // comment character for start of line
	FieldsPerRecord  int  // number of expected fields per record
	LazyQuotes       bool // allow lazy quotes
	TrailingComma    bool // ignored; here for backwards compatibility
	TrimLeadingSpace bool // trim leading space
	line             int
	column           int
	r                *bufio.Reader
	field            bytes.Buffer
}

// NewReader returns a new Reader that reads from r.
func NewReader(r io.Reader) *Reader {
	return &Reader{
		Comma: ',',
		r:     bufio.NewReader(r),
	}
}

// error creates a new ParseError based on err.
func (r *Reader) error(err error) error {
	return &ParseError{
		Line:   r.line,
		Column: r.column,
		Err:    err,
	}
}

// Read reads one record from r.  The record is a slice of strings with each
// string representing one field.
func (r *Reader) Read() (record []string, err error) {
	for {
		record, err = r.parseRecord()
		if record != nil {
			break
		}
		if err != nil {
			return nil, err
		}
	}

	if r.FieldsPerRecord > 0 {
		if len(record) != r.FieldsPerRecord {
			r.column = 0 // report at start of record
			return record, r.error(ErrFieldCount)
		}
	} else if r.FieldsPerRecord == 0 {
		r.FieldsPerRecord = len(record)
	}
	return record, nil
}

// ReadAll reads all the remaining records from r.
// Each record is a slice of fields.
// A successful call returns err == nil, not err == EOF. Because ReadAll is
// defined to read until EOF, it does not treat end of file as an error to be
// reported.
func (r *Reader) ReadAll() (records [][]string, err error) {
	for {
		record, err := r.Read()
		if err == io.EOF {
			return records, nil
		}
		if err != nil {
			return nil, err
		}
		records = append(records, record)
	}
}

// readRune reads one rune from r, folding \r\n to \n and keeping track
// of how far into the line we have read.  r.column will point to the start
// of this rune, not the end of this rune.
func (r *Reader) readRune() (rune, error) {
	r1, _, err := r.r.ReadRune()

	// Handle \r\n here.  We make the simplifying assumption that
	// anytime \r is followed by \n that it can be folded to \n.
	// We will not detect files which contain both \r\n and bare \n.
	if r1 == '\r' {
		r1, _, err = r.r.ReadRune()
		if err == nil {
			if r1 != '\n' {
				r.r.UnreadRune()
				r1 = '\r'
			}
		}
	}
	r.column++
	return r1, err
}

// unreadRune puts the last rune read from r back.
func (r *Reader) unreadRune() {
	r.r.UnreadRune()
	r.column--
}

// skip reads runes up to and including the rune delim or until error.
func (r *Reader) skip(delim rune) error {
	for {
		r1, err := r.readRune()
		if err != nil {
			return err
		}
		if r1 == delim {
			return nil
		}
	}
}

// parseRecord reads and parses a single csv record from r.
func (r *Reader) parseRecord() (fields []string, err error) {
	// Each record starts on a new line.  We increment our line
	// number (lines start at 1, not 0) and set column to -1
	// so as we increment in readRune it points to the character we read.
	r.line++
	r.column = -1

	// Peek at the first rune.  If it is an error we are done.
	// If we are support comments and it is the comment character
	// then skip to the end of line.

	r1, _, err := r.r.ReadRune()
	if err != nil {
		return nil, err
	}

	if r.Comment != 0 && r1 == r.Comment {
		return nil, r.skip('\n')
	}
	r.r.UnreadRune()

	// At this point we have at least one field.
	for {
		haveField, delim, err := r.parseField()
		if haveField {
			fields = append(fields, r.field.String())
		}
		if delim == '\n' || err == io.EOF {
			return fields, err
		} else if err != nil {
			return nil, err
		}
	}
}

// parseField parses the next field in the record.  The read field is
// located in r.field.  Delim is the first character not part of the field
// (r.Comma or '\n').
func (r *Reader) parseField() (haveField bool, delim rune, err error) {
	r.field.Reset()

	r1, err := r.readRune()
	for err == nil && r.TrimLeadingSpace && r1 != '\n' && unicode.IsSpace(r1) {
		r1, err = r.readRune()
	}

	if err == io.EOF && r.column != 0 {
		return true, 0, err
	}
	if err != nil {
		return false, 0, err
	}

	switch r1 {
	case r.Comma:
		// will check below

	case '\n':
		// We are a trailing empty field or a blank line
		if r.column == 0 {
			return false, r1, nil
		}
		return true, r1, nil

	case '"':
		// quoted field
	Quoted:
		for {
			r1, err = r.readRune()
			if err != nil {
				if err == io.EOF {
					if r.LazyQuotes {
						return true, 0, err
					}
					return false, 0, r.error(ErrQuote)
				}
				return false, 0, err
			}
			switch r1 {
			case '"':
				r1, err = r.readRune()
				if err != nil || r1 == r.Comma {
					break Quoted
				}
				if r1 == '\n' {
					return true, r1, nil
				}
				if r1 != '"' {
					if !r.LazyQuotes {
						r.column--
						return false, 0, r.error(ErrQuote)
					}
					// accept the bare quote
					r.field.WriteRune('"')
				}
			case '\n':
				r.line++
				r.column = -1
			}
			r.field.WriteRune(r1)
		}

	default:
		// unquoted field
		for {
			r.field.WriteRune(r1)
			r1, err = r.readRune()
			if err != nil || r1 == r.Comma {
				break
			}
			if r1 == '\n' {
				return true, r1, nil
			}
			if !r.LazyQuotes && r1 == '"' {
				return false, 0, r.error(ErrBareQuote)
			}
		}
	}

	if err != nil {
		if err == io.EOF {
			return true, 0, err
		}
		return false, 0, err
	}

	return true, r1, nil
}