mirror of
https://github.com/featurebasedb/featurebase.git
synced 2026-09-07 17:15:56 +00:00
6d4c1d9db1
* first cut at pre-sort command that works on ndjson * finish pre_sort command for CSV and JSON and add test * try fixing golangci-lint * remove some dumb lint checks * more linter disabling * take .golangci.yml from previous repo * go fmt (facepalm) * remove ioutil to fix lint
29 lines
1.3 KiB
Go
29 lines
1.3 KiB
Go
// Copyright 2021 Molecula Corp. All rights reserved.
|
|
package cmd
|
|
|
|
import (
|
|
"github.com/featurebasedb/featurebase/v3/ctl"
|
|
"github.com/featurebasedb/featurebase/v3/logger"
|
|
"github.com/spf13/cobra"
|
|
)
|
|
|
|
func newPreSortCommand(logdest logger.Logger) *cobra.Command {
|
|
cmd := ctl.NewPreSortCommand(logdest)
|
|
ccmd := &cobra.Command{
|
|
Use: "pre_sort",
|
|
Short: "Sort records within files into files by FB partition for more efficient ingest",
|
|
Long: `
|
|
Takes all input files and writes PartitionN numbered files to a directory, where each file contains only records that will go into the partition it is named for.
|
|
`,
|
|
RunE: usageErrorWrapper(cmd),
|
|
}
|
|
|
|
flags := ccmd.Flags()
|
|
flags.StringVarP(&cmd.File, "file", "", "", "Input file or directory.")
|
|
flags.StringVarP(&cmd.Table, "table", "", "", "Name of table (used to hash keys to determine partition).")
|
|
flags.StringVarP(&cmd.Type, "type", "", cmd.Type, "Input file type (csv or ndjson).")
|
|
flags.StringSliceVar(&cmd.PrimaryKeyFields, "primary-key-fields", []string{}, "Names of primary key fields. For CSV there must be a header row and these pulled from there.")
|
|
flags.IntVar(&cmd.PartitionN, "partition-n", cmd.PartitionN, "Number of partitions.")
|
|
flags.StringVarP(&cmd.OutputDir, "output-dir", "", cmd.OutputDir, "Directory name to write output to.")
|
|
return ccmd
|
|
}
|