In the meantime, I worked with Florian Hackenberger's HocrToPdf.java
to convert multi-page hocr to pdf for viewing and searching.
http://hero.com/public/HocrToPdf.java (needs to be compiled with
javac -deprecation)
I also had to patch /usr/local/share/ocropus/scripts/deskew.lua to
make it work
--- /usr/local/share/ocropus/scripts/deskew.lua 2008-11-14
15:08:47.000000000 -0800
+++ /usr/local/share/ocropus/scripts/deskew.lua~ 2008-11-15
08:31:31.000000000 -0800
@@ -6,10 +6,10 @@
os.exit(1)
end
-proc = ocr.make_DeskewPageByRAST()
+proc = make_DeskewPageByRAST()
input = bytearray:new()
output = bytearray:new()
-iulib.read_image_gray(input,arg[1])
+read_image_gray(input,arg[1])
proc:cleanup(output,input)
-iulib.write_image_gray(arg[2],output)
+write_png(arg[2],output)
-KB
This is Issue 119: no hocr browser rendering code (.js or .css) in
project.
-KB
On Nov 14, 8:13 am, kenberland <ken....@gmail.com> wrote:
It would be great to work with your code, no matter what state,
instead of starting from scratch. I have some resources right now
that I can bring to bare on the task.
On Nov 14, 6:19 am, "Thomas Breuel" <tmb...@gmail.com> wrote:
Yes, we have such JavaScript code; it actually lets you switch between
physical and logical views on the fly. It's not checked into the
Ocropus or hOCR projects yet, though.
Can you submit a bug report/feature request in the issue tracker so
that we don't lose track of this, please? I hope Christian can then
get the code out quickly.
On Fri, Nov 14, 2008 at 07:47, kenberland <ken....@gmail.com> wrote:
Is there .css or javascript available within this project to render
the hocr output from "ocroscript recognize" in a browser?
Is this something I need to write?