Stowage/go
2
0
Fork 0
mirror of https://github.com/golang/go.git synced 2025-12-08 06:10:04 +00:00
Code Issues Projects Releases Packages Wiki Activity

encoding/csv: add FieldPos method

Browse source 
This enables a consumer of a CSV to find out the position
of a CSV field without implementing an intermediate buffer.
This is useful to produce good higher level error messages when
the CSV syntax is OK but the field values don't match expectations.

This also changes the existing semantics of the `ParseError.Column`
field to bring it in line with precedent elsewhere in the Go
standard library (notably go/token.Position) - the column is
now 1-based and indicates a byte count rather than a rune count,
and the error position reporting at the end of a last line without
a newline is now fixed.

This change has some impact on performance:

```
name                                     old time/op    new time/op    delta
Read-8                                     2.14µs ± 0%    2.15µs ± 0%    ~     (p=0.056 n=5+5)
ReadWithFieldsPerRecord-8                  2.15µs ± 2%    2.15µs ± 1%    ~     (p=0.151 n=5+5)
ReadWithoutFieldsPerRecord-8               2.15µs ± 0%    2.15µs ± 0%  +0.37%  (p=0.024 n=5+5)
ReadLargeFields-8                          3.55µs ± 2%    3.59µs ± 0%    ~     (p=0.206 n=5+5)
ReadReuseRecord-8                          1.18µs ± 1%    1.22µs ± 1%  +2.93%  (p=0.008 n=5+5)
ReadReuseRecordWithFieldsPerRecord-8       1.18µs ± 0%    1.21µs ± 0%  +2.54%  (p=0.008 n=5+5)
ReadReuseRecordWithoutFieldsPerRecord-8    1.18µs ± 0%    1.22µs ± 1%  +3.66%  (p=0.008 n=5+5)
ReadReuseRecordLargeFields-8               2.53µs ± 1%    2.57µs ± 1%  +1.70%  (p=0.008 n=5+5)
Write-8                                    1.02µs ± 1%    1.01µs ± 0%  -1.18%  (p=0.016 n=5+4)
```

Fixes #44221.

Change-Id: Id37c50fc396024eef406c5bad45380ecd414f5ea
Reviewed-on: https://go-review.googlesource.com/c/go/+/291290
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Trust: Paul Jolly <paul@myitcv.org.uk>
This commit is contained in:
Roger Peppe 2021-02-11 13:54:45 +00:00 • committed by roger peppe
parent 2c05ba4ae0
commit 6d95e5a4ff
2 changed files with 587 additions and 397 deletions
Download patch file Download diff file Expand all files Collapse all files

71
src/encoding/csv/reader.go
View file

@ -66,7 +66,7 @@ import (
type ParseError struct { type ParseError struct {
StartLine int // Line where the record starts StartLine int // Line where the record starts
Line int // Line where the error occurred Line int // Line where the error occurred
Column int // Column (rune index) where the error occurred Column int // Column (1-based byte index) where the error occurred
Err error // The actual error Err error // The actual error
} }
@ -162,6 +162,10 @@ type Reader struct {
// The i'th field ends at offset fieldIndexes[i] in recordBuffer. // The i'th field ends at offset fieldIndexes[i] in recordBuffer.
fieldIndexes []int fieldIndexes []int
// fieldPositions is an index of field positions for the
// last record returned by Read.
fieldPositions []position
// lastRecord is a record cache and only used when ReuseRecord == true. // lastRecord is a record cache and only used when ReuseRecord == true.
lastRecord []string lastRecord []string
} }
@ -192,6 +196,25 @@ func (r *Reader) Read() (record []string, err error) {
return record, err return record, err
} }
// FieldPos returns the line and column corresponding to
// the start of the field with the given index in the slice most recently
// returned by Read. Numbering of lines and columns starts at 1;
// columns are counted in bytes, not runes.
//
// If this is called with an out-of-bounds index, it panics.
func (r *Reader) FieldPos(field int) (line, column int) {
if field < 0 || field >= len(r.fieldPositions) {
panic("out of range index passed to FieldPos")
}
p := &r.fieldPositions[field]
return p.line, p.col
}
// pos holds the position of a field in the current line.
type position struct {
line, col int
}
// ReadAll reads all the remaining records from r. // ReadAll reads all the remaining records from r.
// Each record is a slice of fields. // Each record is a slice of fields.
// A successful call returns err == nil, not err == io.EOF. Because ReadAll is // A successful call returns err == nil, not err == io.EOF. Because ReadAll is
@ -260,7 +283,7 @@ func (r *Reader) readRecord(dst []string) ([]string, error) {
} }
// Read line (automatically skipping past empty lines and any comments). // Read line (automatically skipping past empty lines and any comments).
var line, fullLine []byte var line []byte
var errRead error var errRead error
for errRead == nil { for errRead == nil {
line, errRead = r.readLine() line, errRead = r.readLine()
@ -272,7 +295,6 @@ func (r *Reader) readRecord(dst []string) ([]string, error) {
line = nil line = nil
continue // Skip empty lines continue // Skip empty lines
} }
fullLine = line
break break
} }
if errRead == io.EOF { if errRead == io.EOF {
@ -286,10 +308,20 @@ func (r *Reader) readRecord(dst []string) ([]string, error) {
recLine := r.numLine // Starting line for record recLine := r.numLine // Starting line for record
r.recordBuffer = r.recordBuffer[:0] r.recordBuffer = r.recordBuffer[:0]
r.fieldIndexes = r.fieldIndexes[:0] r.fieldIndexes = r.fieldIndexes[:0]
r.fieldPositions = r.fieldPositions[:0]
pos := position{line: r.numLine, col: 1}
parseField: parseField:
for { for {
if r.TrimLeadingSpace { if r.TrimLeadingSpace {
line = bytes.TrimLeftFunc(line, unicode.IsSpace) i := bytes.IndexFunc(line, func(r rune) bool {
return !unicode.IsSpace(r)
})
if i < 0 {
i = len(line)
pos.col -= lengthNL(line)
}
line = line[i:]
pos.col += i
} }
if len(line) == 0 || line[0] != '"' { if len(line) == 0 || line[0] != '"' {
// Non-quoted string field // Non-quoted string field
@ -303,48 +335,56 @@ parseField:
// Check to make sure a quote does not appear in field. // Check to make sure a quote does not appear in field.
if !r.LazyQuotes { if !r.LazyQuotes {
if j := bytes.IndexByte(field, '"'); j >= 0 { if j := bytes.IndexByte(field, '"'); j >= 0 {
col := utf8.RuneCount(fullLine[:len(fullLine)-len(line[j:])]) col := pos.col + j
err = &ParseError{StartLine: recLine, Line: r.numLine, Column: col, Err: ErrBareQuote} err = &ParseError{StartLine: recLine, Line: r.numLine, Column: col, Err: ErrBareQuote}
break parseField break parseField
} }
} }
r.recordBuffer = append(r.recordBuffer, field...) r.recordBuffer = append(r.recordBuffer, field...)
r.fieldIndexes = append(r.fieldIndexes, len(r.recordBuffer)) r.fieldIndexes = append(r.fieldIndexes, len(r.recordBuffer))
r.fieldPositions = append(r.fieldPositions, pos)
if i >= 0 { if i >= 0 {
line = line[i+commaLen:] line = line[i+commaLen:]
pos.col += i + commaLen
continue parseField continue parseField
} }
break parseField break parseField
} else { } else {
// Quoted string field // Quoted string field
fieldPos := pos
line = line[quoteLen:] line = line[quoteLen:]
pos.col += quoteLen
for { for {
i := bytes.IndexByte(line, '"') i := bytes.IndexByte(line, '"')
if i >= 0 { if i >= 0 {
// Hit next quote. // Hit next quote.
r.recordBuffer = append(r.recordBuffer, line[:i]...) r.recordBuffer = append(r.recordBuffer, line[:i]...)
line = line[i+quoteLen:] line = line[i+quoteLen:]
pos.col += i + quoteLen
switch rn := nextRune(line); { switch rn := nextRune(line); {
case rn == '"': case rn == '"':
// `""` sequence (append quote). // `""` sequence (append quote).
r.recordBuffer = append(r.recordBuffer, '"') r.recordBuffer = append(r.recordBuffer, '"')
line = line[quoteLen:] line = line[quoteLen:]
pos.col += quoteLen
case rn == r.Comma: case rn == r.Comma:
// `",` sequence (end of field). // `",` sequence (end of field).
line = line[commaLen:] line = line[commaLen:]
pos.col += commaLen
r.fieldIndexes = append(r.fieldIndexes, len(r.recordBuffer)) r.fieldIndexes = append(r.fieldIndexes, len(r.recordBuffer))
r.fieldPositions = append(r.fieldPositions, fieldPos)
continue parseField continue parseField
case lengthNL(line) == len(line): case lengthNL(line) == len(line):
// `"\n` sequence (end of line). // `"\n` sequence (end of line).
r.fieldIndexes = append(r.fieldIndexes, len(r.recordBuffer)) r.fieldIndexes = append(r.fieldIndexes, len(r.recordBuffer))
r.fieldPositions = append(r.fieldPositions, fieldPos)
break parseField break parseField
case r.LazyQuotes: case r.LazyQuotes:
// `"` sequence (bare quote). // `"` sequence (bare quote).
r.recordBuffer = append(r.recordBuffer, '"') r.recordBuffer = append(r.recordBuffer, '"')
default: default:
// `"*` sequence (invalid non-escaped quote). // `"*` sequence (invalid non-escaped quote).
col := utf8.RuneCount(fullLine[:len(fullLine)-len(line)-quoteLen]) err = &ParseError{StartLine: recLine, Line: r.numLine, Column: pos.col - quoteLen, Err: ErrQuote}
err = &ParseError{StartLine: recLine, Line: r.numLine, Column: col, Err: ErrQuote}
break parseField break parseField
} }
} else if len(line) > 0 { } else if len(line) > 0 {
@ -353,19 +393,23 @@ parseField:
if errRead != nil { if errRead != nil {
break parseField break parseField
} }
pos.col += len(line)
line, errRead = r.readLine() line, errRead = r.readLine()
if len(line) > 0 {
pos.line++
pos.col = 1
}
if errRead == io.EOF { if errRead == io.EOF {
errRead = nil errRead = nil
} }
fullLine = line
} else { } else {
// Abrupt end of file (EOF or error). // Abrupt end of file (EOF or error).
if !r.LazyQuotes && errRead == nil { if !r.LazyQuotes && errRead == nil {
col := utf8.RuneCount(fullLine) err = &ParseError{StartLine: recLine, Line: pos.line, Column: pos.col, Err: ErrQuote}
err = &ParseError{StartLine: recLine, Line: r.numLine, Column: col, Err: ErrQuote}
break parseField break parseField
} }
r.fieldIndexes = append(r.fieldIndexes, len(r.recordBuffer)) r.fieldIndexes = append(r.fieldIndexes, len(r.recordBuffer))
r.fieldPositions = append(r.fieldPositions, fieldPos)
break parseField break parseField
} }
} }
@ -392,7 +436,12 @@ parseField:
// Check or update the expected fields per record. // Check or update the expected fields per record.
if r.FieldsPerRecord > 0 { if r.FieldsPerRecord > 0 {
if len(dst) != r.FieldsPerRecord && err == nil { if len(dst) != r.FieldsPerRecord && err == nil {
err = &ParseError{StartLine: recLine, Line: recLine, Err: ErrFieldCount} err = &ParseError{
StartLine: recLine,
Line: recLine,
Column: 1,
Err: ErrFieldCount,
}
} }
} else if r.FieldsPerRecord == 0 { } else if r.FieldsPerRecord == 0 {
r.FieldsPerRecord = len(dst) r.FieldsPerRecord = len(dst)

479
src/encoding/csv/reader_test.go
View file

@ -5,6 +5,8 @@
package csv package csv
import ( import (
"errors"
"fmt"
"io" "io"
"reflect" "reflect"
"strings" "strings"
@ -12,12 +14,12 @@ import (
"unicode/utf8" "unicode/utf8"
) )
func TestRead(t *testing.T) { type readTest struct {
tests := []struct {
Name string Name string
Input string Input string
Output [][]string Output [][]string
Error error Positions [][][2]int
Errors []error
// These fields are copied into the Reader // These fields are copied into the Reader
Comma rune Comma rune
@ -27,25 +29,32 @@ func TestRead(t *testing.T) {
LazyQuotes bool LazyQuotes bool
TrimLeadingSpace bool TrimLeadingSpace bool
ReuseRecord bool ReuseRecord bool
}{{ }
// In these tests, the §, ¶ and ∑ characters in readTest.Input are used to denote
// the start of a field, a record boundary and the position of an error respectively.
// They are removed before parsing and are used to verify the position
// information reported by FieldPos.
var readTests = []readTest{{
Name: "Simple", Name: "Simple",
Input: "a,b,c\n", Input: "§a,§b,§c\n",
Output: [][]string{{"a", "b", "c"}}, Output: [][]string{{"a", "b", "c"}},
}, { }, {
Name: "CRLF", Name: "CRLF",
Input: "a,b\r\nc,d\r\n", Input: "§a,§b\r\n¶§c,§d\r\n",
Output: [][]string{{"a", "b"}, {"c", "d"}}, Output: [][]string{{"a", "b"}, {"c", "d"}},
}, { }, {
Name: "BareCR", Name: "BareCR",
Input: "a,b\rc,d\r\n", Input: "§a,§b\rc,§d\r\n",
Output: [][]string{{"a", "b\rc", "d"}}, Output: [][]string{{"a", "b\rc", "d"}},
}, { }, {
Name: "RFC4180test", Name: "RFC4180test",
Input: `#field1,field2,field3 Input: `§#field1,§field2,§field3
"aaa","bb §"aaa",§"bb
b","ccc" b",§"ccc"
"a,a","b""bb","ccc" §"a,a",§"b""bb",§"ccc"
zzz,yyy,xxx §zzz,§yyy,§xxx
`, `,
Output: [][]string{ Output: [][]string{
{"#field1", "field2", "field3"}, {"#field1", "field2", "field3"},
@ -55,147 +64,156 @@ zzz,yyy,xxx
}, },
UseFieldsPerRecord: true, UseFieldsPerRecord: true,
FieldsPerRecord: 0, FieldsPerRecord: 0,
}, { }, {
Name: "NoEOLTest", Name: "NoEOLTest",
Input: "a,b,c", Input: "§a,§b,§c",
Output: [][]string{{"a", "b", "c"}}, Output: [][]string{{"a", "b", "c"}},
}, { }, {
Name: "Semicolon", Name: "Semicolon",
Input: "a;b;c\n", Input: "§a;§b;§c\n",
Output: [][]string{{"a", "b", "c"}}, Output: [][]string{{"a", "b", "c"}},
Comma: ';', Comma: ';',
}, { }, {
Name: "MultiLine", Name: "MultiLine",
Input: `"two Input: `§"two
line","one line","three line",§"one line",§"three
line line
field"`, field"`,
Output: [][]string{{"two\nline", "one line", "three\nline\nfield"}}, Output: [][]string{{"two\nline", "one line", "three\nline\nfield"}},
}, { }, {
Name: "BlankLine", Name: "BlankLine",
Input: "a,b,c\n\nd,e,f\n\n", Input: "§a,§b,§c\n\n¶§d,§e,§f\n\n",
Output: [][]string{ Output: [][]string{
{"a", "b", "c"}, {"a", "b", "c"},
{"d", "e", "f"}, {"d", "e", "f"},
}, },
}, { }, {
Name: "BlankLineFieldCount", Name: "BlankLineFieldCount",
Input: "a,b,c\n\nd,e,f\n\n", Input: "§a,§b,§c\n\n¶§d,§e,§f\n\n",
Output: [][]string{ Output: [][]string{
{"a", "b", "c"}, {"a", "b", "c"},
{"d", "e", "f"}, {"d", "e", "f"},
}, },
UseFieldsPerRecord: true, UseFieldsPerRecord: true,
FieldsPerRecord: 0, FieldsPerRecord: 0,
}, { }, {
Name: "TrimSpace", Name: "TrimSpace",
Input: " a, b, c\n", Input: " §a, §b, §c\n",
Output: [][]string{{"a", "b", "c"}}, Output: [][]string{{"a", "b", "c"}},
TrimLeadingSpace: true, TrimLeadingSpace: true,
}, { }, {
Name: "LeadingSpace", Name: "LeadingSpace",
Input: " a, b, c\n", Input: "§ a,§ b,§ c\n",
Output: [][]string{{" a", " b", " c"}}, Output: [][]string{{" a", " b", " c"}},
}, { }, {
Name: "Comment", Name: "Comment",
Input: "#1,2,3\na,b,c\n#comment", Input: "#1,2,3\n§a,§b,§c\n#comment",
Output: [][]string{{"a", "b", "c"}}, Output: [][]string{{"a", "b", "c"}},
Comment: '#', Comment: '#',
}, { }, {
Name: "NoComment", Name: "NoComment",
Input: "#1,2,3\na,b,c", Input: "§#1,§2,§3\n¶§a,§b,§c",
Output: [][]string{{"#1", "2", "3"}, {"a", "b", "c"}}, Output: [][]string{{"#1", "2", "3"}, {"a", "b", "c"}},
}, { }, {
Name: "LazyQuotes", Name: "LazyQuotes",
Input: `a "word","1"2",a","b`, Input: `§a "word",§"1"2",§a",§"b`,
Output: [][]string{{`a "word"`, `1"2`, `a"`, `b`}}, Output: [][]string{{`a "word"`, `1"2`, `a"`, `b`}},
LazyQuotes: true, LazyQuotes: true,
}, { }, {
Name: "BareQuotes", Name: "BareQuotes",
Input: `a "word","1"2",a"`, Input: `§a "word",§"1"2",§a"`,
Output: [][]string{{`a "word"`, `1"2`, `a"`}}, Output: [][]string{{`a "word"`, `1"2`, `a"`}},
LazyQuotes: true, LazyQuotes: true,
}, { }, {
Name: "BareDoubleQuotes", Name: "BareDoubleQuotes",
Input: `a""b,c`, Input: `§a""b,§c`,
Output: [][]string{{`a""b`, `c`}}, Output: [][]string{{`a""b`, `c`}},
LazyQuotes: true, LazyQuotes: true,
}, { }, {
Name: "BadDoubleQuotes", Name: "BadDoubleQuotes",
Input: `a""b,c`, Input: `§a∑""b,c`,
Error: &ParseError{StartLine: 1, Line: 1, Column: 1, Err: ErrBareQuote}, Errors: []error{&ParseError{Err: ErrBareQuote}},
}, { }, {
Name: "TrimQuote", Name: "TrimQuote",
Input: ` "a"," b",c`, Input: ` §"a",§" b",§c`,
Output: [][]string{{"a", " b", "c"}}, Output: [][]string{{"a", " b", "c"}},
TrimLeadingSpace: true, TrimLeadingSpace: true,
}, { }, {
Name: "BadBareQuote", Name: "BadBareQuote",
Input: `a "word","b"`, Input: `§a ∑"word","b"`,
Error: &ParseError{StartLine: 1, Line: 1, Column: 2, Err: ErrBareQuote}, Errors: []error{&ParseError{Err: ErrBareQuote}},
}, { }, {
Name: "BadTrailingQuote", Name: "BadTrailingQuote",
Input: `"a word",b"`, Input: `§"a word",b∑"`,
Error: &ParseError{StartLine: 1, Line: 1, Column: 10, Err: ErrBareQuote}, Errors: []error{&ParseError{Err: ErrBareQuote}},
}, { }, {
Name: "ExtraneousQuote", Name: "ExtraneousQuote",
Input: `"a "word","b"`, Input: `§"a ∑"word","b"`,
Error: &ParseError{StartLine: 1, Line: 1, Column: 3, Err: ErrQuote}, Errors: []error{&ParseError{Err: ErrQuote}},
}, { }, {
Name: "BadFieldCount", Name: "BadFieldCount",
Input: "a,b,c\nd,e", Input: "§a,§b,§c\n¶∑§d,§e",
Error: &ParseError{StartLine: 2, Line: 2, Err: ErrFieldCount}, Errors: []error{nil, &ParseError{Err: ErrFieldCount}},
Output: [][]string{{"a", "b", "c"}, {"d", "e"}},
UseFieldsPerRecord: true, UseFieldsPerRecord: true,
FieldsPerRecord: 0, FieldsPerRecord: 0,
}, { }, {
Name: "BadFieldCountMultiple",
Input: "§a,§b,§c\n¶∑§d,§e\n¶∑§f",
Errors: []error{nil, &ParseError{Err: ErrFieldCount}, &ParseError{Err: ErrFieldCount}},
Output: [][]string{{"a", "b", "c"}, {"d", "e"}, {"f"}},
UseFieldsPerRecord: true,
FieldsPerRecord: 0,
}, {
Name: "BadFieldCount1", Name: "BadFieldCount1",
Input: `a,b,c`, Input: `§∑a,§b,§c`,
Error: &ParseError{StartLine: 1, Line: 1, Err: ErrFieldCount}, Errors: []error{&ParseError{Err: ErrFieldCount}},
Output: [][]string{{"a", "b", "c"}},
UseFieldsPerRecord: true, UseFieldsPerRecord: true,
FieldsPerRecord: 2, FieldsPerRecord: 2,
}, { }, {
Name: "FieldCount", Name: "FieldCount",
Input: "a,b,c\nd,e", Input: "§a,§b,§c\n¶§d,§e",
Output: [][]string{{"a", "b", "c"}, {"d", "e"}}, Output: [][]string{{"a", "b", "c"}, {"d", "e"}},
}, { }, {
Name: "TrailingCommaEOF", Name: "TrailingCommaEOF",
Input: "a,b,c,", Input: "§a,§b,§c,§",
Output: [][]string{{"a", "b", "c", ""}}, Output: [][]string{{"a", "b", "c", ""}},
}, { }, {
Name: "TrailingCommaEOL", Name: "TrailingCommaEOL",
Input: "a,b,c,\n", Input: "§a,§b,§c,§\n",
Output: [][]string{{"a", "b", "c", ""}}, Output: [][]string{{"a", "b", "c", ""}},
}, { }, {
Name: "TrailingCommaSpaceEOF", Name: "TrailingCommaSpaceEOF",
Input: "a,b,c, ", Input: "§a,§b,§c, §",
Output: [][]string{{"a", "b", "c", ""}}, Output: [][]string{{"a", "b", "c", ""}},
TrimLeadingSpace: true, TrimLeadingSpace: true,
}, { }, {
Name: "TrailingCommaSpaceEOL", Name: "TrailingCommaSpaceEOL",
Input: "a,b,c, \n", Input: "§a,§b,§c, §\n",
Output: [][]string{{"a", "b", "c", ""}}, Output: [][]string{{"a", "b", "c", ""}},
TrimLeadingSpace: true, TrimLeadingSpace: true,
}, { }, {
Name: "TrailingCommaLine3", Name: "TrailingCommaLine3",
Input: "a,b,c\nd,e,f\ng,hi,", Input: "§a,§b,§c\n¶§d,§e,§f\n¶§g,§hi,§",
Output: [][]string{{"a", "b", "c"}, {"d", "e", "f"}, {"g", "hi", ""}}, Output: [][]string{{"a", "b", "c"}, {"d", "e", "f"}, {"g", "hi", ""}},
TrimLeadingSpace: true, TrimLeadingSpace: true,
}, { }, {
Name: "NotTrailingComma3", Name: "NotTrailingComma3",
Input: "a,b,c, \n", Input: "§a,§b,§c,§ \n",
Output: [][]string{{"a", "b", "c", " "}}, Output: [][]string{{"a", "b", "c", " "}},
}, { }, {
Name: "CommaFieldTest", Name: "CommaFieldTest",
Input: `x,y,z,w Input: `§x,§y,§z,§w
x,y,z, §x,§y,§z,§
x,y,, §x,§y,§,§
x,,, §x,§,§,§
,,, §,§,§,§
"x","y","z","w" §"x",§"y",§"z",§"w"
"x","y","z","" §"x",§"y",§"z",§""
"x","y","","" §"x",§"y",§"",§""
"x","","","" §"x",§"",§"",§""
"","","","" §"",§"",§"",§""
`, `,
Output: [][]string{ Output: [][]string{
{"x", "y", "z", "w"}, {"x", "y", "z", "w"},
@ -209,184 +227,186 @@ x,,,
{"x", "", "", ""}, {"x", "", "", ""},
{"", "", "", ""}, {"", "", "", ""},
}, },
}, { }, {
Name: "TrailingCommaIneffective1", Name: "TrailingCommaIneffective1",
Input: "a,b,\nc,d,e", Input: "§a,§b,§\n¶§c,§d,§e",
Output: [][]string{ Output: [][]string{
{"a", "b", ""}, {"a", "b", ""},
{"c", "d", "e"}, {"c", "d", "e"},
}, },
TrimLeadingSpace: true, TrimLeadingSpace: true,
}, { }, {
Name: "ReadAllReuseRecord", Name: "ReadAllReuseRecord",
Input: "a,b\nc,d", Input: "§a,§b\n¶§c,§d",
Output: [][]string{ Output: [][]string{
{"a", "b"}, {"a", "b"},
{"c", "d"}, {"c", "d"},
}, },
ReuseRecord: true, ReuseRecord: true,
}, { }, {
Name: "StartLine1", // Issue 19019 Name: "StartLine1", // Issue 19019
Input: "a,\"b\nc\"d,e", Input: "§a,\"b\nc∑\"d,e",
Error: &ParseError{StartLine: 1, Line: 2, Column: 1, Err: ErrQuote}, Errors: []error{&ParseError{Err: ErrQuote}},
}, { }, {
Name: "StartLine2", Name: "StartLine2",
Input: "a,b\n\"d\n\n,e", Input: "§a,§b\n¶§\"d\n\n,e∑",
Error: &ParseError{StartLine: 2, Line: 5, Column: 0, Err: ErrQuote}, Errors: []error{nil, &ParseError{Err: ErrQuote}},
}, { Output: [][]string{{"a", "b"}},
}, {
Name: "CRLFInQuotedField", // Issue 21201 Name: "CRLFInQuotedField", // Issue 21201
Input: "A,\"Hello\r\nHi\",B\r\n", Input: "§A,§\"Hello\r\nHi\",§B\r\n",
Output: [][]string{ Output: [][]string{
{"A", "Hello\nHi", "B"}, {"A", "Hello\nHi", "B"},
}, },
}, { }, {
Name: "BinaryBlobField", // Issue 19410 Name: "BinaryBlobField", // Issue 19410
Input: "x09\x41\xb4\x1c,aktau", Input: "§x09\x41\xb4\x1c,§aktau",
Output: [][]string{{"x09A\xb4\x1c", "aktau"}}, Output: [][]string{{"x09A\xb4\x1c", "aktau"}},
}, { }, {
Name: "TrailingCR", Name: "TrailingCR",
Input: "field1,field2\r", Input: "§field1,§field2\r",
Output: [][]string{{"field1", "field2"}}, Output: [][]string{{"field1", "field2"}},
}, { }, {
Name: "QuotedTrailingCR", Name: "QuotedTrailingCR",
Input: "\"field\"\r", Input: "§\"field\"\r",
Output: [][]string{{"field"}}, Output: [][]string{{"field"}},
}, { }, {
Name: "QuotedTrailingCRCR", Name: "QuotedTrailingCRCR",
Input: "\"field\"\r\r", Input: "§\"field∑\"\r\r",
Error: &ParseError{StartLine: 1, Line: 1, Column: 6, Err: ErrQuote}, Errors: []error{&ParseError{Err: ErrQuote}},
}, { }, {
Name: "FieldCR", Name: "FieldCR",
Input: "field\rfield\r", Input: "§field\rfield\r",
Output: [][]string{{"field\rfield"}}, Output: [][]string{{"field\rfield"}},
}, { }, {
Name: "FieldCRCR", Name: "FieldCRCR",
Input: "field\r\rfield\r\r", Input: "§field\r\rfield\r\r",
Output: [][]string{{"field\r\rfield\r"}}, Output: [][]string{{"field\r\rfield\r"}},
}, { }, {
Name: "FieldCRCRLF", Name: "FieldCRCRLF",
Input: "field\r\r\nfield\r\r\n", Input: "§field\r\r\n¶§field\r\r\n",
Output: [][]string{{"field\r"}, {"field\r"}}, Output: [][]string{{"field\r"}, {"field\r"}},
}, { }, {
Name: "FieldCRCRLFCR", Name: "FieldCRCRLFCR",
Input: "field\r\r\n\rfield\r\r\n\r", Input: "§field\r\r\n¶§\rfield\r\r\n\r",
Output: [][]string{{"field\r"}, {"\rfield\r"}}, Output: [][]string{{"field\r"}, {"\rfield\r"}},
}, { }, {
Name: "FieldCRCRLFCRCR", Name: "FieldCRCRLFCRCR",
Input: "field\r\r\n\r\rfield\r\r\n\r\r", Input: "§field\r\r\n¶§\r\rfield\r\r\n¶§\r\r",
Output: [][]string{{"field\r"}, {"\r\rfield\r"}, {"\r"}}, Output: [][]string{{"field\r"}, {"\r\rfield\r"}, {"\r"}},
}, { }, {
Name: "MultiFieldCRCRLFCRCR", Name: "MultiFieldCRCRLFCRCR",
Input: "field1,field2\r\r\n\r\rfield1,field2\r\r\n\r\r,", Input: "§field1,§field2\r\r\n¶§\r\rfield1,§field2\r\r\n¶§\r\r,§",
Output: [][]string{ Output: [][]string{
{"field1", "field2\r"}, {"field1", "field2\r"},
{"\r\rfield1", "field2\r"}, {"\r\rfield1", "field2\r"},
{"\r\r", ""}, {"\r\r", ""},
}, },
}, { }, {
Name: "NonASCIICommaAndComment", Name: "NonASCIICommaAndComment",
Input: "a£b,c£ \td,e\n€ comment\n", Input: "§a£§b,c£ \t§d,e\n€ comment\n",
Output: [][]string{{"a", "b,c", "d,e"}}, Output: [][]string{{"a", "b,c", "d,e"}},
TrimLeadingSpace: true, TrimLeadingSpace: true,
Comma: '£', Comma: '£',
Comment: '€', Comment: '€',
}, { }, {
Name: "NonASCIICommaAndCommentWithQuotes", Name: "NonASCIICommaAndCommentWithQuotes",
Input: "a€\" b,\"€ c\nλ comment\n", Input: "§a€§\" b,\"€§ c\nλ comment\n",
Output: [][]string{{"a", " b,", " c"}}, Output: [][]string{{"a", " b,", " c"}},
Comma: '€', Comma: '€',
Comment: 'λ', Comment: 'λ',
}, { }, {
// λ and θ start with the same byte. // λ and θ start with the same byte.
// This tests that the parser doesn't confuse such characters. // This tests that the parser doesn't confuse such characters.
Name: "NonASCIICommaConfusion", Name: "NonASCIICommaConfusion",
Input: "\"abθcd\"λefθgh", Input: "§\"abθcd\"λ§efθgh",
Output: [][]string{{"abθcd", "efθgh"}}, Output: [][]string{{"abθcd", "efθgh"}},
Comma: 'λ', Comma: 'λ',
Comment: '€', Comment: '€',
}, { }, {
Name: "NonASCIICommentConfusion", Name: "NonASCIICommentConfusion",
Input: "λ\nλ\nθ\nλ\n", Input: "§λ\n¶§λ\nθ\n¶§λ\n",
Output: [][]string{{"λ"}, {"λ"}, {"λ"}}, Output: [][]string{{"λ"}, {"λ"}, {"λ"}},
Comment: 'θ', Comment: 'θ',
}, { }, {
Name: "QuotedFieldMultipleLF", Name: "QuotedFieldMultipleLF",
Input: "\"\n\n\n\n\"", Input: "§\"\n\n\n\n\"",
Output: [][]string{{"\n\n\n\n"}}, Output: [][]string{{"\n\n\n\n"}},
}, { }, {
Name: "MultipleCRLF", Name: "MultipleCRLF",
Input: "\r\n\r\n\r\n\r\n", Input: "\r\n\r\n\r\n\r\n",
}, { }, {
// The implementation may read each line in several chunks if it doesn't fit entirely // The implementation may read each line in several chunks if it doesn't fit entirely
// in the read buffer, so we should test the code to handle that condition. // in the read buffer, so we should test the code to handle that condition.
Name: "HugeLines", Name: "HugeLines",
Input: strings.Repeat("#ignore\n", 10000) + strings.Repeat("@", 5000) + "," + strings.Repeat("*", 5000), Input: strings.Repeat("#ignore\n", 10000) + "§" + strings.Repeat("@", 5000) + ",§" + strings.Repeat("*", 5000),
Output: [][]string{{strings.Repeat("@", 5000), strings.Repeat("*", 5000)}}, Output: [][]string{{strings.Repeat("@", 5000), strings.Repeat("*", 5000)}},
Comment: '#', Comment: '#',
}, { }, {
Name: "QuoteWithTrailingCRLF", Name: "QuoteWithTrailingCRLF",
Input: "\"foo\"bar\"\r\n", Input: "§\"foo∑\"bar\"\r\n",
Error: &ParseError{StartLine: 1, Line: 1, Column: 4, Err: ErrQuote}, Errors: []error{&ParseError{Err: ErrQuote}},
}, { }, {
Name: "LazyQuoteWithTrailingCRLF", Name: "LazyQuoteWithTrailingCRLF",
Input: "\"foo\"bar\"\r\n", Input: "§\"foo\"bar\"\r\n",
Output: [][]string{{`foo"bar`}}, Output: [][]string{{`foo"bar`}},
LazyQuotes: true, LazyQuotes: true,
}, { }, {
Name: "DoubleQuoteWithTrailingCRLF", Name: "DoubleQuoteWithTrailingCRLF",
Input: "\"foo\"\"bar\"\r\n", Input: "§\"foo\"\"bar\"\r\n",
Output: [][]string{{`foo"bar`}}, Output: [][]string{{`foo"bar`}},
}, { }, {
Name: "EvenQuotes", Name: "EvenQuotes",
Input: `""""""""`, Input: `§""""""""`,
Output: [][]string{{`"""`}}, Output: [][]string{{`"""`}},
}, { }, {
Name: "OddQuotes", Name: "OddQuotes",
Input: `"""""""`, Input: `§"""""""∑`,
Error: &ParseError{StartLine: 1, Line: 1, Column: 7, Err: ErrQuote}, Errors: []error{&ParseError{Err: ErrQuote}},
}, { }, {
Name: "LazyOddQuotes", Name: "LazyOddQuotes",
Input: `"""""""`, Input: `§"""""""`,
Output: [][]string{{`"""`}}, Output: [][]string{{`"""`}},
LazyQuotes: true, LazyQuotes: true,
}, { }, {
Name: "BadComma1", Name: "BadComma1",
Comma: '\n', Comma: '\n',
Error: errInvalidDelim, Errors: []error{errInvalidDelim},
}, { }, {
Name: "BadComma2", Name: "BadComma2",
Comma: '\r', Comma: '\r',
Error: errInvalidDelim, Errors: []error{errInvalidDelim},
}, { }, {
Name: "BadComma3", Name: "BadComma3",
Comma: '"', Comma: '"',
Error: errInvalidDelim, Errors: []error{errInvalidDelim},
}, { }, {
Name: "BadComma4", Name: "BadComma4",
Comma: utf8.RuneError, Comma: utf8.RuneError,
Error: errInvalidDelim, Errors: []error{errInvalidDelim},
}, { }, {
Name: "BadComment1", Name: "BadComment1",
Comment: '\n', Comment: '\n',
Error: errInvalidDelim, Errors: []error{errInvalidDelim},
}, { }, {
Name: "BadComment2", Name: "BadComment2",
Comment: '\r', Comment: '\r',
Error: errInvalidDelim, Errors: []error{errInvalidDelim},
}, { }, {
Name: "BadComment3", Name: "BadComment3",
Comment: utf8.RuneError, Comment: utf8.RuneError,
Error: errInvalidDelim, Errors: []error{errInvalidDelim},
}, { }, {
Name: "BadCommaComment", Name: "BadCommaComment",
Comma: 'X', Comma: 'X',
Comment: 'X', Comment: 'X',
Error: errInvalidDelim, Errors: []error{errInvalidDelim},
}} }}
for _, tt := range tests { func TestRead(t *testing.T) {
t.Run(tt.Name, func(t *testing.T) { newReader := func(tt readTest) (*Reader, [][][2]int, map[int][2]int) {
r := NewReader(strings.NewReader(tt.Input)) positions, errPositions, input := makePositions(tt.Input)
r := NewReader(strings.NewReader(input))
if tt.Comma != 0 { if tt.Comma != 0 {
r.Comma = tt.Comma r.Comma = tt.Comma
@ -400,17 +420,138 @@ x,,,
r.LazyQuotes = tt.LazyQuotes r.LazyQuotes = tt.LazyQuotes
r.TrimLeadingSpace = tt.TrimLeadingSpace r.TrimLeadingSpace = tt.TrimLeadingSpace
r.ReuseRecord = tt.ReuseRecord r.ReuseRecord = tt.ReuseRecord
return r, positions, errPositions
}
for _, tt := range readTests {
t.Run(tt.Name, func(t *testing.T) {
r, positions, errPositions := newReader(tt)
out, err := r.ReadAll() out, err := r.ReadAll()
if !reflect.DeepEqual(err, tt.Error) { if wantErr := firstError(tt.Errors, positions, errPositions); wantErr != nil {
t.Errorf("ReadAll() error:\ngot %v\nwant %v", err, tt.Error) if !reflect.DeepEqual(err, wantErr) {
} else if !reflect.DeepEqual(out, tt.Output) { t.Fatalf("ReadAll() error mismatch:\ngot %v (%#v)\nwant %v (%#v)", err, err, wantErr, wantErr)
t.Errorf("ReadAll() output:\ngot %q\nwant %q", out, tt.Output) }
if out != nil {
t.Fatalf("ReadAll() output:\ngot %q\nwant nil", out)
}
} else {
if err != nil {
t.Fatalf("unexpected Readall() error: %v", err)
}
if !reflect.DeepEqual(out, tt.Output) {
t.Fatalf("ReadAll() output:\ngot %q\nwant %q", out, tt.Output)
}
}
// Check field and error positions.
r, _, _ = newReader(tt)
for recNum := 0; ; recNum++ {
rec, err := r.Read()
var wantErr error
if recNum < len(tt.Errors) && tt.Errors[recNum] != nil {
wantErr = errorWithPosition(tt.Errors[recNum], recNum, positions, errPositions)
} else if recNum >= len(tt.Output) {
wantErr = io.EOF
}
if !reflect.DeepEqual(err, wantErr) {
t.Fatalf("Read() error at record %d:\ngot %v (%#v)\nwant %v (%#v)", recNum, err, err, wantErr, wantErr)
}
// ErrFieldCount is explicitly non-fatal.
if err != nil && !errors.Is(err, ErrFieldCount) {
if recNum < len(tt.Output) {
t.Fatalf("need more records; got %d want %d", recNum, len(tt.Output))
}
break
}
if got, want := rec, tt.Output[recNum]; !reflect.DeepEqual(got, want) {
t.Errorf("Read vs ReadAll mismatch;\ngot %q\nwant %q", got, want)
}
pos := positions[recNum]
if len(pos) != len(rec) {
t.Fatalf("mismatched position length at record %d", recNum)
}
for i := range rec {
line, col := r.FieldPos(i)
if got, want := [2]int{line, col}, pos[i]; got != want {
t.Errorf("position mismatch at record %d, field %d;\ngot %v\nwant %v", recNum, i, got, want)
}
}
} }
}) })
} }
} }
// firstError returns the first non-nil error in errs,
// with the position adjusted according to the error's
// index inside positions.
func firstError(errs []error, positions [][][2]int, errPositions map[int][2]int) error {
for i, err := range errs {
if err != nil {
return errorWithPosition(err, i, positions, errPositions)
}
}
return nil
}
func errorWithPosition(err error, recNum int, positions [][][2]int, errPositions map[int][2]int) error {
parseErr, ok := err.(*ParseError)
if !ok {
return err
}
if recNum >= len(positions) {
panic(fmt.Errorf("no positions found for error at record %d", recNum))
}
errPos, ok := errPositions[recNum]
if !ok {
panic(fmt.Errorf("no error position found for error at record %d", recNum))
}
parseErr1 := *parseErr
parseErr1.StartLine = positions[recNum][0][0]
parseErr1.Line = errPos[0]
parseErr1.Column = errPos[1]
return &parseErr1
}
// makePositions returns the expected field positions of all
// the fields in text, the positions of any errors, and the text with the position markers
// removed.
//
// The start of each field is marked with a § symbol;
// CSV lines are separated by ¶ symbols;
// Error positions are marked with ∑ symbols.
func makePositions(text string) ([][][2]int, map[int][2]int, string) {
buf := make([]byte, 0, len(text))
var positions [][][2]int
errPositions := make(map[int][2]int)
line, col := 1, 1
recNum := 0
for len(text) > 0 {
r, size := utf8.DecodeRuneInString(text)
switch r {
case '\n':
line++
col = 1
buf = append(buf, '\n')
case '§':
if len(positions) == 0 {
positions = append(positions, [][2]int{})
}
positions[len(positions)-1] = append(positions[len(positions)-1], [2]int{line, col})
case '¶':
positions = append(positions, [][2]int{})
recNum++
case '∑':
errPositions[recNum] = [2]int{line, col}
default:
buf = append(buf, text[:size]...)
col += size
}
text = text[size:]
}
return positions, errPositions, string(buf)
}
// nTimes is an io.Reader which yields the string s n times. // nTimes is an io.Reader which yields the string s n times.
type nTimes struct { type nTimes struct {
s string s string