anotherhadi/eleakxir
1
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

take 3 first and 3 last

Browse Source 
Signed-off-by: Hadi <112569860+anotherhadi@users.noreply.github.com>
This commit is contained in:
Hadi
2025-10-04 16:39:46 +02:00
parent c036649a70
commit b51c149488
1 changed files with 12 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
back/search/dataleak/dataleak.go
View File

@@ -217,11 +217,20 @@ func getFromClause(s *server.Server) string {
} }
return fmt.Sprintf("read_parquet([%s], union_by_name=true, filename=true)", strings.Join(parquets, ", ")) return fmt.Sprintf("read_parquet([%s], union_by_name=true, filename=true)", strings.Join(parquets, ", "))
} }
func GetDataleakSample(s server.Server, path string) ([][]string, error) { func GetDataleakSample(s server.Server, path string) ([][]string, error) {
rowsData := [][]string{} rowsData := [][]string{}
query := fmt.Sprintf("SELECT * FROM read_parquet('%s') LIMIT 5", path) // Use row_number() to get first 3 and last 3 rows
query := fmt.Sprintf(`
WITH numbered AS (
SELECT *, row_number() OVER () AS rn, count(*) OVER () AS total_rows
FROM read_parquet('%s')
)
SELECT * EXCLUDE (rn, total_rows)
FROM numbered
WHERE rn <= 3 OR rn > total_rows - 3
`, path)
rows, err := s.Duckdb.Query(query) rows, err := s.Duckdb.Query(query)
if err != nil { if err != nil {
return rowsData, err return rowsData, err
@@ -233,7 +242,7 @@ func GetDataleakSample(s server.Server, path string) ([][]string, error) {
return rowsData, err return rowsData, err
} }
rowsData = append(rowsData, cols) rowsData = append(rowsData, cols) // header
rawResult := make([][]byte, len(cols)) rawResult := make([][]byte, len(cols))
dest := make([]any, len(cols)) dest := make([]any, len(cols))