encoding/csv: always allow trailing commas

Original CL by rsc (11916045):

The motivation for disallowing them was RFC 4180 saying
"The last field in the record must not be followed by a comma."
I believe this is an admonition to CSV generators, not readers.
When reading, anything followed by a comma is not the last field.

Fixes #5892.

R=golang-dev, rsc, r
CC=golang-dev
https://golang.org/cl/12294043

diff --git a/src/pkg/encoding/csv/reader.go b/src/pkg/encoding/csv/reader.go
index b099caf60a805cb157d277db5ba636c72fff6c16..b328dcc375cd7cccf8f680540fe34cc01c791be8 100644 (file)
--- a/src/pkg/encoding/csv/reader.go
+++ b/src/pkg/encoding/csv/reader.go
@@ -72,7 +72,7 @@ func (e *ParseError) Error() string {
 
 // These are the errors that can be returned in ParseError.Error
 var (
-       ErrTrailingComma = errors.New("extra delimiter at end of line")
+       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")
@@ -98,16 +98,14 @@ var (
 // If LazyQuotes is true, a quote may appear in an unquoted field and a
 // non-doubled quote may appear in a quoted field.
 //
-// If TrailingComma is true, the last field may be an unquoted empty 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 // Allow trailing comma
-       TrimLeadingSpace bool // Trim leading space
+       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
@@ -257,23 +255,15 @@ func (r *Reader) parseField() (haveField bool, delim rune, err error) {
        r.field.Reset()
 
        r1, err := r.readRune()
-       if err != nil {
-               // If we have EOF and are not at the start of a line
-               // then we return the empty field.  We have already
-               // checked for trailing commas if needed.
-               if err == io.EOF && r.column != 0 {
-                       return true, 0, err
-               }
-               return false, 0, err
+       for err == nil && r.TrimLeadingSpace && r1 != '\n' && unicode.IsSpace(r1) {
+               r1, err = r.readRune()
        }
 
-       if r.TrimLeadingSpace {
-               for r1 != '\n' && unicode.IsSpace(r1) {
-                       r1, err = r.readRune()
-                       if err != nil {
-                               return false, 0, err
-                       }
-               }
+       if err == io.EOF && r.column != 0 {
+               return true, 0, err
+       }
+       if err != nil {
+               return false, 0, err
        }
 
        switch r1 {
@@ -349,25 +339,5 @@ func (r *Reader) parseField() (haveField bool, delim rune, err error) {
                return false, 0, err
        }
 
-       if !r.TrailingComma {
-               // We don't allow trailing commas.  See if we
-               // are at the end of the line (being mindful
-               // of trimming spaces).
-               c := r.column
-               r1, err = r.readRune()
-               if r.TrimLeadingSpace {
-                       for r1 != '\n' && unicode.IsSpace(r1) {
-                               r1, err = r.readRune()
-                               if err != nil {
-                                       break
-                               }
-                       }
-               }
-               if err == io.EOF || r1 == '\n' {
-                       r.column = c // report the comma
-                       return false, 0, r.error(ErrTrailingComma)
-               }
-               r.unreadRune()
-       }
        return true, r1, nil
 }

diff --git a/src/pkg/encoding/csv/reader_test.go b/src/pkg/encoding/csv/reader_test.go
index 5fd84a76bdfb830db453e7d2d2674870a5846956..123df06bc85e7e781b7ca80bb1561ed4061047f4 100644 (file)
--- a/src/pkg/encoding/csv/reader_test.go
+++ b/src/pkg/encoding/csv/reader_test.go
@@ -171,32 +171,32 @@ field"`,
                Output: [][]string{{"a", "b", "c"}, {"d", "e"}},
        },
        {
-               Name:  "BadTrailingCommaEOF",
-               Input: "a,b,c,",
-               Error: "extra delimiter at end of line", Line: 1, Column: 5,
+               Name:   "TrailingCommaEOF",
+               Input:  "a,b,c,",
+               Output: [][]string{{"a", "b", "c", ""}},
        },
        {
-               Name:  "BadTrailingCommaEOL",
-               Input: "a,b,c,\n",
-               Error: "extra delimiter at end of line", Line: 1, Column: 5,
+               Name:   "TrailingCommaEOL",
+               Input:  "a,b,c,\n",
+               Output: [][]string{{"a", "b", "c", ""}},
        },
        {
-               Name:             "BadTrailingCommaSpaceEOF",
+               Name:             "TrailingCommaSpaceEOF",
                TrimLeadingSpace: true,
                Input:            "a,b,c, ",
-               Error:            "extra delimiter at end of line", Line: 1, Column: 5,
+               Output:           [][]string{{"a", "b", "c", ""}},
        },
        {
-               Name:             "BadTrailingCommaSpaceEOL",
+               Name:             "TrailingCommaSpaceEOL",
                TrimLeadingSpace: true,
                Input:            "a,b,c, \n",
-               Error:            "extra delimiter at end of line", Line: 1, Column: 5,
+               Output:           [][]string{{"a", "b", "c", ""}},
        },
        {
-               Name:             "BadTrailingCommaLine3",
+               Name:             "TrailingCommaLine3",
                TrimLeadingSpace: true,
                Input:            "a,b,c\nd,e,f\ng,hi,",
-               Error:            "extra delimiter at end of line", Line: 3, Column: 4,
+               Output:           [][]string{{"a", "b", "c"}, {"d", "e", "f"}, {"g", "hi", ""}},
        },
        {
                Name:   "NotTrailingComma3",
@@ -231,7 +231,7 @@ x,,,
                },
        },
        {
-               Name:             "Issue 2366",
+               Name:             "TrailingCommaIneffective1",
                TrailingComma:    true,
                TrimLeadingSpace: true,
                Input:            "a,b,\nc,d,e",
@@ -241,11 +241,14 @@ x,,,
                },
        },
        {
-               Name:             "Issue 2366a",
+               Name:             "TrailingCommaIneffective2",
                TrailingComma:    false,
                TrimLeadingSpace: true,
                Input:            "a,b,\nc,d,e",
-               Error:            "extra delimiter at end of line",
+               Output: [][]string{
+                       {"a", "b", ""},
+                       {"c", "d", "e"},
+               },
        },
 }