summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorJonas Knobloch <jonas.knobloch@t-online.de>2026-04-21 19:49:03 +0200
committerJonas Knobloch <jonas.knobloch@t-online.de>2026-04-22 12:53:44 +0200
commit2f8e06698271c077fdd9a88dfe3ad5db90f55f83 (patch)
tree704bd2131ad1aae7cdbad6056bc77db81709e617
parent753f9c9e5fb1707ffdbac4f3c790c0218343e1f2 (diff)
Skip AppleDouble metadata files
-rw-r--r--dataset/parquet.go10
1 files changed, 8 insertions, 2 deletions
diff --git a/dataset/parquet.go b/dataset/parquet.go
index 81454f8..1ba2a9b 100644
--- a/dataset/parquet.go
+++ b/dataset/parquet.go
@@ -7,6 +7,7 @@ import (
"iter"
"path/filepath"
"slices"
+ "strings"
"github.com/apache/arrow-go/v18/arrow"
"github.com/apache/arrow-go/v18/arrow/array"
@@ -23,12 +24,17 @@ type ParquetReader struct {
}
func NewParquetReader(name string) (*ParquetReader, error) {
- var shards []string
+ shards := make([]string, 0)
if matches, err := filepath.Glob(filepath.Join(name, "*.parquet")); err != nil {
return nil, err
} else {
- shards = matches
+ // skip AppleDouble metadata files
+ for _, m := range matches {
+ if !strings.HasPrefix(filepath.Base(m), "._") {
+ shards = append(shards, m)
+ }
+ }
}
if len(shards) == 0 {