MyFavMirrors/minio
1
0
Fork 0
mirror of https://github.com/minio/minio.git synced 2025-01-24 05:03:16 -05:00
Code Issues Packages Projects Releases Wiki Activity

fix: missing data on multiple columns reading parquet (#11499)

Browse Source 
fixes #11413
This commit is contained in:
Harshavardhana 2021-02-10 08:49:48 -08:00 committed by GitHub
parent 5a18d437ce
commit f53d1de87f
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 19 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
pkg/s3select/internal/parquet-go/page.go
View File

@ -21,6 +21,7 @@ import (
"context"
"errors"
"fmt"
"io"
"math"
"strings"
@ -108,24 +109,39 @@ func readPage(
}
repLevelsLen = pageHeader.DataPageHeaderV2.GetRepetitionLevelsByteLength()
repLevelsBuf = make([]byte, repLevelsLen)
if _, err = thriftReader.Read(repLevelsBuf); err != nil {
n, err := io.ReadFull(thriftReader, repLevelsBuf)
if err != nil {
return nil, err
}
if n != int(repLevelsLen) {
return nil, fmt.Errorf("expected parquet header repetition levels %d, got %d", repLevelsLen, n)
}
defLevelsLen = pageHeader.DataPageHeaderV2.GetDefinitionLevelsByteLength()
defLevelsBuf = make([]byte, defLevelsLen)
if _, err = thriftReader.Read(defLevelsBuf); err != nil {
n, err = io.ReadFull(thriftReader, defLevelsBuf)
if err != nil {
return nil, err
}
if n != int(defLevelsLen) {
return nil, fmt.Errorf("expected parquet header definition levels %d, got %d", defLevelsLen, n)
}
}
dbLen := pageHeader.GetCompressedPageSize() - repLevelsLen - defLevelsLen
if dbLen < 0 {
return nil, errors.New("parquet: negative data length")
}
dataBuf := make([]byte, dbLen)
if _, err = thriftReader.Read(dataBuf); err != nil {
n, err := io.ReadFull(thriftReader, dataBuf)
if err != nil {
return nil, err
}
if n != int(dbLen) {
return nil, fmt.Errorf("expected parquet data buffer %d, got %d", dbLen, n)
}
if dataBuf, err = compressionCodec(metadata.GetCodec()).uncompress(dataBuf); err != nil {
return nil, err