Add pgconf tool that prints the overall confidence for a whole page of hocr

2 files changed, 64 insertions, 0 deletions

diff --git a/lib/hocr/lines.go b/lib/hocr/lines.go
index 20c6550..74e8f9a 100644
--- a/lib/hocr/lines.go
+++ b/lib/hocr/lines.go
@@ -112,3 +112,20 @@ func GetLineDetails(hocrfn string) (line.Details, error) {
 	n := strings.Replace(filepath.Base(hocrfn), ".hocr", "", 1)
 	return parseLineDetails(h, img, n)
 }
+
+func GetLineBasics(hocrfn string) (line.Details, error) {
+	var newlines line.Details
+
+	file, err := ioutil.ReadFile(hocrfn)
+	if err != nil {
+		return newlines, err
+	}
+
+	h, err := Parse(file)
+	if err != nil {
+		return newlines, err
+	}
+
+	n := strings.Replace(filepath.Base(hocrfn), ".hocr", "", 1)
+	return parseLineDetails(h, nil, n)
+}
diff --git a/pgconf/main.go b/pgconf/main.go
new file mode 100644
index 0000000..1b70ecc
--- /dev/null
+++ b/pgconf/main.go
@@ -0,0 +1,47 @@
+package main
+
+import (
+	"flag"
+	"fmt"
+	"log"
+	"os"
+
+	"rescribe.xyz/go.git/lib/hocr"
+	"rescribe.xyz/go.git/lib/line"
+)
+
+func main() {
+	flag.Usage = func() {
+		fmt.Fprintf(os.Stderr, "Usage: pgconf hocr\n")
+		fmt.Fprintf(os.Stderr, "Prints the total confidence for a page, as an average of the confidence of each word.\n")
+		flag.PrintDefaults()
+	}
+	flag.Parse()
+	if flag.NArg() != 1 {
+		flag.Usage()
+		os.Exit(1)
+	}
+
+	var err error
+	lines := make(line.Details, 0)
+
+	for _, f := range flag.Args() {
+		var newlines line.Details
+		newlines, err = hocr.GetLineBasics(f)
+		if err != nil {
+			log.Fatal(err)
+		}
+
+		for _, l := range newlines {
+			lines = append(lines, l)
+		}
+	}
+
+	var total float64
+	for _, l := range lines {
+		total += l.Avgconf
+	}
+	avg := total / float64(len(lines))
+
+	fmt.Printf("%0.2f\n", avg)
+}