encoding/csv: Preallocate records slice

Currently parseRecord will always start with a nil
slice and then resize the slice on append. For input
with a fixed number of fields per record we can preallocate
the slice to avoid having to resize the slice.

This change implements this optimization by using
FieldsPerRecord as capacity if it's > 0 and also adds a
benchmark to better show the differences.

benchmark         old ns/op     new ns/op     delta
BenchmarkRead     19741         17909         -9.28%

benchmark         old allocs     new allocs     delta
BenchmarkRead     59             41             -30.51%

benchmark         old bytes     new bytes     delta
BenchmarkRead     6276          5844          -6.88%

Change-Id: I7c2abc9c80a23571369bcfcc99a8ffc474eae7ab
Reviewed-on: https://go-review.googlesource.com/8880
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

diff --git a/src/encoding/csv/reader.go b/src/encoding/csv/reader.go
index d9432954ac9d6ec7e92c547971d2f5e5e1593dd0..d0a09044fb487797d31dcadf16bd336f6d6b3670 100644 (file)
--- a/src/encoding/csv/reader.go
+++ b/src/encoding/csv/reader.go
@@ -228,6 +228,12 @@ func (r *Reader) parseRecord() (fields []string, err error) {
        }
        r.r.UnreadRune()
 
+       // If FieldsPerRecord is greater then 0 we can assume the final
+       // length of fields to be equal to FieldsPerRecord.
+       if r.FieldsPerRecord > 0 {
+               fields = make([]string, 0, r.FieldsPerRecord)
+       }
+
        // At this point we have at least one field.
        for {
                haveField, delim, err := r.parseField()

diff --git a/src/encoding/csv/reader_test.go b/src/encoding/csv/reader_test.go
index 123df06bc85e7e781b7ca80bb1561ed4061047f4..b3c4f3bf18fe05141f2979a0ec8c095a412e5a3d 100644 (file)
--- a/src/encoding/csv/reader_test.go
+++ b/src/encoding/csv/reader_test.go
@@ -282,3 +282,25 @@ func TestRead(t *testing.T) {
                }
        }
 }
+
+func BenchmarkRead(b *testing.B) {
+       data := `x,y,z,w
+x,y,z,
+x,y,,
+x,,,
+,,,
+"x","y","z","w"
+"x","y","z",""
+"x","y","",""
+"x","","",""
+"","","",""
+`
+
+       for i := 0; i < b.N; i++ {
+               _, err := NewReader(strings.NewReader(data)).ReadAll()
+
+               if err != nil {
+                       b.Fatalf("could not read data: %s", err)
+               }
+       }
+}