DIY Book Scanner

Hello.

I'm running openSUSE and maintain a few packages for book scanning that may be helpful for you.

cuneiform — OCR System
djvusmooth — Graphical Editor for DjVu
exact-image — Fast Image Manipulation Programs
jbig2enc — Encoder for JBIG2
minidjvu — Bitonal DjVu Encoder/Decoder
ocrodjvu — Wrapper for OCR Systems
pdf2djvu — PDF to DjVu Converter
scantailor — Interactive Post-Processing Tool for Scanned Pages
yagf — Graphical Front-end for Cuneiform OCR Tool

You can find packages for oS 11.1, 11.2, 11.3, SLE 11SP1 in repository
http://download.opensuse.org/repositori ... se_version>/

what jbig2enc version?

I built 0.27 (upon leptonica-lib 1.63) with several alternative versions including various patches

http://dokupuppylinux.co.cc/programs:encoders

.pet package can be uncompressed renaming it in .tar.gz

jbig2enc-0.27+git20100608, built against liblept-devel-static-1.66 with akrykukov patch.

Thanks

I seen you also commented line 27 in thessalonica-pdf.py (calling PIL image library) in order to make working also without PIL

anyway, not being a STATICAL COMPILATION (it needs GLIbc 2.8) cannot run in my 2.6.21.7 kernel with a previous version of GLIBC

Lazy_Kent, I just wanted to drop in here and say: thanks for letting us know about your work! If you don't mind my very n00bish questions, what package manager does OpenSUSE use? Will I be able to use these packages on other distros?

daniel_reetz wrote:Lazy_Kent, I just wanted to drop in here and say: thanks for letting us know about your work! If you don't mind my very n00bish questions, what package manager does OpenSUSE use? Will I be able to use these packages on other distros?

It is a matter of GLIBC compatibility essentially

these RPMs are archives compressed with lzma algorithm, if your current distro matches GLIBC 2.8, then you can

- download RPM
- uncompress (with PEAZIP)
- pick the executable, moving to /usr/local/bin and use

you can try and view if app downloaded works

this is why, in oder to allow universal use of an app. this must be compiled STATICALLY (not LINKED)

dingodog wrote: these RPMs are archives compressed with lzma algorithm, if your current distro matches GLIBC 2.8, then you can

- download RPM
- uncompress (with PEAZIP)
- pick the executable, moving to /usr/local/bin and use

Of course, if you want to keep your sanity, always install software through your package manager... especially if you are new to Linux. The difference between "can" and "should" is sometimes significant when your OS gives you so much freedom to tinker. And if you're on a debian-based system, last I knew alien could turn most rpm's in deb's, but it's been a long time since I used anything other than Gentoo.

packages are nothing else than compressed archives with one or several directories inside

knowing the Linux filesystem and its internal structure, smarter users can take a package, un-compress (for testing a package, for instance) and pick binaries in local bin dir /urs/local/bin/

this is the whole process that so-called package manager only automatize

for instance: you can not have a lzma decoding libs, so, in order to test opensuse packages, you extract this with PeaZip and move in directory you want

dingodog wrote:I seen you also commented line 27 in thessalonica-pdf.py (calling PIL image library) in order to make working also without PIL anyway, not being a STATICAL COMPILATION (it needs GLIbc 2.8) cannot run in my 2.6.21.7 kernel with a previous version of GLIBC

Not me. I grabbed it from http://dokupuppylinux.co.cc/_media/prog ... ica-pdf.py.
I keep the script with original name (pdf.py) for compatibility.

jbig2enc depends on system shared libraries
% ldd /usr/bin/jbig2
linux-gate.so.1 => (0xffffe000)
libpng12.so.0 => /usr/lib/libpng12.so.0 (0xb76f2000)
libjpeg.so.62 => /usr/lib/libjpeg.so.62 (0xb76a1000)
libtiff.so.3 => /usr/lib/libtiff.so.3 (0xb7646000)
libstdc++.so.6 => /usr/lib/libstdc++.so.6 (0xb7552000)
libm.so.6 => /lib/libm.so.6 (0xb7528000)
libgcc_s.so.1 => /lib/libgcc_s.so.1 (0xb7519000)
libc.so.6 => /lib/libc.so.6 (0xb73bd000)
libz.so.1 => /lib/libz.so.1 (0xb73a8000)
/lib/ld-linux.so.2 (0xb7737000)

daniel_reetz wrote:what package manager does OpenSUSE use? Will I be able to use these packages on other distros?

I'm sorry, you willn't. openSUSE use RPM. For other rpm-based distros you can grab specs and patches from my openSUSE Build Service project and build packages locally.
https://build.opensuse.org/project/pack ... ALazy_Kent

New package for openSUSE users.
img2djvu — Single-pass DjVu Encoder Based on DjVu Libre and ImageMagick.
http://software.opensuse.org/search?q=img2djvu

Author: Alexey Shipunov
URL: http://github.com/ashipunov/img2djvu

It works with Cuneiform. Waiting for ocrodjvu 0.7.0 for Tesseract 3.00 support.
Any comments and requests are welcome. I will contact the author.