From 3c4c5f7c202b7c54ca8d23e7bd7bff4a4bb696cc Mon Sep 17 00:00:00 2001
From: Nick White <git@njw.name>
Date: Mon, 25 Feb 2019 12:09:06 +0000
Subject: Add tool to extract plain text from hocr
---
hocrtotxt/main.go | 30 ++++++++++++++++++++++++++++++
1 file changed, 30 insertions(+)
create mode 100644 hocrtotxt/main.go
(limited to 'hocrtotxt')
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)
+}
--
cgit v1.2.1-24-ge1ad