From ca5014f0ad35ad475860ff817a2ebed121bc1f27 Mon Sep 17 00:00:00 2001
From: Nick White <git@njw.name>
Date: Fri, 13 Mar 2020 16:52:05 +0000
Subject: Add simple "getbests" utility, useful for statistics gathering

---
 cmd/getbests/main.go | 70 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 70 insertions(+)
 create mode 100644 cmd/getbests/main.go

(limited to 'cmd')

diff --git a/cmd/getbests/main.go b/cmd/getbests/main.go
new file mode 100644
index 0000000..2b0c40a
--- /dev/null
+++ b/cmd/getbests/main.go
@@ -0,0 +1,70 @@
+// Copyright 2020 Nick White.
+// Use of this source code is governed by the GPLv3
+// license that can be found in the LICENSE file.
+
+package main
+
+import (
+	"flag"
+	"fmt"
+	"log"
+	"strings"
+
+	"rescribe.xyz/bookpipeline"
+)
+
+const usage = `Usage: getbests
+
+Downloads every 'best' file from a set of OCRed books. This is
+useful for statistics.
+`
+
+// null writer to enable non-verbose logging to be discarded
+type NullWriter bool
+
+func (w NullWriter) Write(p []byte) (n int, err error) {
+	return len(p), nil
+}
+
+type Pipeliner interface {
+	Init() error
+	ListObjects(bucket string, prefix string) ([]string, error)
+	Download(bucket string, key string, fn string) error
+	WIPStorageId() string
+}
+
+func main() {
+	flag.Usage = func() {
+		fmt.Fprintf(flag.CommandLine.Output(), usage)
+		flag.PrintDefaults()
+	}
+	flag.Parse()
+
+	var n NullWriter
+	verboselog := log.New(n, "", log.LstdFlags)
+
+	var conn Pipeliner
+	conn = &bookpipeline.AwsConn{Region: "eu-west-2", Logger: verboselog}
+
+	err := conn.Init()
+	if err != nil {
+		log.Fatalln("Error setting up cloud connection:", err)
+	}
+
+	log.Println("Getting list of all available objects to filter through")
+	objs, err := conn.ListObjects(conn.WIPStorageId(), "")
+	if err != nil {
+		log.Fatalln("Failed to get list of files", err)
+	}
+
+	log.Println("Downloading all best files found")
+	for _, i := range objs {
+		parts := strings.Split(i, "/")
+		if parts[len(parts) - 1] == "best" {
+			err = conn.Download(conn.WIPStorageId(), i, parts[0] + "-best")
+			if err != nil {
+				log.Fatalln("Failed to download file", i, err)
+			}
+		}
+	}
+}
-- 
cgit v1.2.1-24-ge1ad