aboutsummaryrefslogtreecommitdiffstats
path: root/skate/cmd/skate-map/main.go
diff options
context:
space:
mode:
authorMartin Czygan <martin.czygan@gmail.com>2021-04-24 02:29:10 +0200
committerMartin Czygan <martin.czygan@gmail.com>2021-04-24 02:29:10 +0200
commite7f8c117c3586c6387a744043ad68c1850471719 (patch)
tree84f01a142e89635548790a0b73c137c7a888a87d /skate/cmd/skate-map/main.go
parent19aa629f7f4c6345769784147493e2598d746dd3 (diff)
downloadrefcat-e7f8c117c3586c6387a744043ad68c1850471719.tar.gz
refcat-e7f8c117c3586c6387a744043ad68c1850471719.zip
wip: move to prefixed, w/ skate-map
Diffstat (limited to 'skate/cmd/skate-map/main.go')
-rw-r--r--skate/cmd/skate-map/main.go58
1 files changed, 58 insertions, 0 deletions
diff --git a/skate/cmd/skate-map/main.go b/skate/cmd/skate-map/main.go
new file mode 100644
index 0000000..c5fb798
--- /dev/null
+++ b/skate/cmd/skate-map/main.go
@@ -0,0 +1,58 @@
+// skate-map runs a given map function over input data. We mostly want to
+// extract a key from a json document.
+package main
+
+import (
+ "flag"
+ "fmt"
+ "log"
+ "os"
+ "runtime"
+ "text/tabwriter"
+
+ "git.archive.org/martin/cgraph/skate"
+ "git.archive.org/martin/cgraph/skate/parallel"
+)
+
+var (
+ mapperName = flag.String("m", "", "mapper to run")
+ numWorkers = flag.Int("w", runtime.NumCPU(), "number of workers")
+ batchSize = flag.Int("b", 50000, "batch size")
+ verbose = flag.Bool("verbose", false, "show progress")
+ extraValue = flag.String("x", "", "extra value to pass to configurable mappers")
+)
+
+func main() {
+ flag.Parse()
+ // XXX: introduce prefixes
+ availableMappers := map[string]skate.Mapper{
+ "id": skate.Identity,
+ "ff": skate.CreateFixedFieldFunc(*extraValue),
+ "title": skate.MapperTitle,
+ "tnorm": skate.MapperTitleNormalized,
+ "tnysi": skate.MapperTitleNysiis,
+ "tsand": skate.MapperTitleSandcrawler,
+ }
+ switch {
+ case *mapperName != "":
+ if f, ok := availableMappers[*mapperName]; !ok {
+ log.Fatal("unknown mapper name: %v", *mapperName)
+ } else {
+ pp := parallel.NewProcessor(os.Stdin, os.Stdout, f)
+ pp.NumWorkers = *numWorkers
+ pp.BatchSize = *batchSize
+ pp.Verbose = *verbose
+ if err := pp.Run(); err != nil {
+ log.Fatal(err)
+ }
+ }
+ default:
+ fmt.Println("skate-map available mappers")
+ fmt.Println()
+ w := tabwriter.NewWriter(os.Stdout, 0, 0, 4, ' ', 0)
+ for k, v := range availableMappers {
+ fmt.Fprintf(w, "%s\t%s\n", k, skate.NameOf(v))
+ }
+ w.Flush()
+ }
+}