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
|
package magic
import (
"bytes"
"encoding/csv"
"io"
)
// Csv matches a comma-separated values file.
func Csv(raw []byte, limit uint32) bool {
return sv(raw, ',', limit)
}
// Tsv matches a tab-separated values file.
func Tsv(raw []byte, limit uint32) bool {
return sv(raw, '\t', limit)
}
func sv(in []byte, comma rune, limit uint32) bool {
r := csv.NewReader(dropLastLine(in, limit))
r.Comma = comma
r.TrimLeadingSpace = true
r.LazyQuotes = true
r.Comment = '#'
lines, err := r.ReadAll()
return err == nil && r.FieldsPerRecord > 1 && len(lines) > 1
}
// dropLastLine drops the last incomplete line from b.
//
// mimetype limits itself to ReadLimit bytes when performing a detection.
// This means, for file formats like CSV for NDJSON, the last line of the input
// can be an incomplete line.
func dropLastLine(b []byte, cutAt uint32) io.Reader {
if cutAt == 0 {
return bytes.NewReader(b)
}
if uint32(len(b)) >= cutAt {
for i := cutAt - 1; i > 0; i-- {
if b[i] == '\n' {
return bytes.NewReader(b[:i])
}
}
// No newline was found between the 0 index and cutAt.
return bytes.NewReader(b[:cutAt])
}
return bytes.NewReader(b)
}
|
