Add tool to extract plain text from hocr

1 files changed, 30 insertions, 0 deletions

diff --git a/hocrtotxt/main.go b/hocrtotxt/main.go
new file mode 100644
index 0000000..6821a9e
--- /dev/null
+++ b/hocrtotxt/main.go
@@ -0,0 +1,30 @@
+package main
+
+import (
+	"flag"
+	"fmt"
+	"log"
+	"os"
+
+	"rescribe.xyz/go.git/lib/hocr"
+)
+
+func main() {
+	flag.Usage = func() {
+		fmt.Fprintf(os.Stderr, "Usage: hocrtotxt hocrfile\n")
+		fmt.Fprintf(os.Stderr, "Prints the text from a hocr file.\n")
+		flag.PrintDefaults()
+	}
+	flag.Parse()
+	if flag.NArg() != 1 {
+		flag.Usage()
+		os.Exit(1)
+	}
+
+	text, err := hocr.GetText(flag.Arg(0))
+	if err != nil {
+		log.Fatal(err)
+	}
+
+	fmt.Printf("%s\n", text)
+}