mapleFU commented on code in PR #38367:
URL: https://github.com/apache/arrow/pull/38367#discussion_r1375773113
##########
go/parquet/internal/encoding/boolean_decoder.go:
##########
@@ -109,3 +114,76 @@ func (dec *PlainBooleanDecoder) DecodeSpaced(out []bool,
nullCount int, validBit
}
return dec.Decode(out)
}
+
+type RleBooleanDecoder struct {
+ decoder
+
+ rleDec *utils.RleDecoder
+}
+
+func (RleBooleanDecoder) Type() parquet.Type {
+ return parquet.Types.Boolean
+}
+
+func (dec *RleBooleanDecoder) SetData(nvals int, data []byte) error {
+ dec.nvals = nvals
+
+ if len(data) < 4 {
+ return fmt.Errorf("invalid length - %d (corrupt data page?)",
len(data))
+ }
+
+ // load the first 4 bytes in little-endian which indicates the length
+ nbytes := binary.LittleEndian.Uint32(data[:4])
+ if nbytes > uint32(len(data)-4) {
+ return fmt.Errorf("received invalid number of bytes - %d
(corrupt data page?)", nbytes)
+ }
+
+ dec.data = data[4:]
+ if dec.rleDec == nil {
+ dec.rleDec = utils.NewRleDecoder(bytes.NewReader(dec.data), 1)
+ } else {
+ dec.rleDec.Reset(bytes.NewReader(dec.data), 1)
+ }
+ return nil
+}
+
+func (dec *RleBooleanDecoder) Decode(out []bool) (int, error) {
+ max := shared_utils.MinInt(len(out), dec.nvals)
+
+ var (
+ buf [1024]uint64
+ n = max
+ )
+
+ for n > 0 {
+ batch := shared_utils.MinInt(len(buf), n)
+ decoded := dec.rleDec.GetBatch(buf[:batch])
+ if decoded != batch {
+ return max - n, io.ErrUnexpectedEOF
Review Comment:
> We are writing out the decoded values via the loop at line 165. The input
to this function is an output slice to write to and we populate it after the
call to GetBatch. So returning max - n here informs the caller of how many
values were populated into that slice before the error was hit.
No, `max` is not the real maximum of the underlying data. For rep-def
existing scenerio, it means "rep-def" size, rather than the actual size. So it
may greater than existing value. (At least this is the case in C++)
So here, If user want decode more value than expected
1. when rep-def might not exist, it will return ok, since max is limited by
max
2. when rep-def might exists, it will get a larger than expected max, and
raise an Error here.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
