diff --git a/office/misc/tesseract/actions.py b/office/misc/tesseract/actions.py new file mode 100644 index 0000000000..c16dcb9a94 --- /dev/null +++ b/office/misc/tesseract/actions.py @@ -0,0 +1,26 @@ +#!/usr/bin/env python +# -*- coding: utf-8 -*- +# +# Licensed under the GNU General Public License, version 3. +# See the file http://www.gnu.org/licenses/gpl.txt + +from pisi.actionsapi import shelltools +from pisi.actionsapi import autotools +from pisi.actionsapi import pisitools +from pisi.actionsapi import get + +def setup(): + shelltools.system("./autogen.sh") + autotools.configure("--disable-static") + # fıx unused dırect dependency analysis + pisitools.dosed("libtool", " -shared ", " -Wl,-O1,--as-needed -shared ") + +def build(): + autotools.make() + autotools.make("training") + +def install(): + autotools.rawInstall("DESTDIR=%s" % get.installDIR()) + autotools.make("training-install DESTDIR=%s" % get.installDIR()) + + pisitools.dodoc("LICENSE", "README*") \ No newline at end of file diff --git a/office/misc/tesseract/pspec.xml b/office/misc/tesseract/pspec.xml new file mode 100644 index 0000000000..5951bb1662 --- /dev/null +++ b/office/misc/tesseract/pspec.xml @@ -0,0 +1,110 @@ + + + + + tesseract + https://github.com/tesseract-ocr/tesseract + + PisiLinux Community + admins@pisilinux.org + + office.misc + APACHE + app:console + Tesseract OCR Engine + This package contains an OCR engine - libtesseract and a command line program - tesseract. Tesseract 4 adds a new neural net (LSTM) based OCR engine which is focused on line recognition, but also still supports the legacy Tesseract OCR engine of Tess + https://github.com/tesseract-ocr/tesseract/archive/4.1.1.tar.gz + + leptonica-devel + cairo-devel + icu4c-devel + pango-devel + tiff-devel + libgomp + + libxslt-devel + asciidoc + + + + + tesseract + Tesseract OCR Engine + + leptonica + libgomp + tiff + + + /usr/lib/libtesseract.so.4.0.1 + /usr/lib/libtesseract.so.4 + /usr/lib/libtesseract.so + /usr/bin/tesseract + /usr/share/tessdata + /usr/share/doc + /usr/share + + + + + tesseract-devel + Development files for tesseract + + tesseract + + + /usr/lib/pkgconfig/tesseract.pc + /usr/include/tesseract + + + + + tesseract-tools + Tools for tesseract + + tesseract + fontconfig + leptonica + cairo + icu4c + pango + + + /usr/bin/set_unicharset_properties + /usr/bin/language-specific.sh + /usr/bin/unicharset_extractor + /usr/bin/tesstrain_utils.sh + /usr/bin/combine_lang_model + /usr/bin/classifier_tester + /usr/bin/merge_unicharsets + /usr/bin/combine_tessdata + /usr/bin/shapeclustering + /usr/bin/ambiguous_words + /usr/bin/dawg2wordlist + /usr/bin/wordlist2dawg + /usr/bin/tesstrain.sh + /usr/bin/lstmtraining + /usr/bin/cntraining + /usr/bin/mftraining + /usr/bin/text2image + /usr/bin/lstmeval + + + + + + 2020-01-19 + 4.1.1 + Version bump. + Blue Devil + bluedevil@sctzine.com + + + 2019-01-25 + 4.0.0 + First release + PisiLinux Community + admins@pisilinux.org + + + \ No newline at end of file diff --git a/office/misc/tesseract/translations.xml b/office/misc/tesseract/translations.xml new file mode 100644 index 0000000000..1990ebf2e2 --- /dev/null +++ b/office/misc/tesseract/translations.xml @@ -0,0 +1,20 @@ + + + + tesseract + Tesseract OCR Motoru + tesseract, en kararlı açık kaynaklı OCR (Optik Karakter Tanıma) motorlarından biridir. Uygulama girdi olarak aldığı çeşitli resim biçimlerini analiz ederek metin biçiminde çıktı üretir. + + + + tesseract-devel + tesseract için geliştirme dosyaları + tesseract-devel, tesseract için geliştirme dosyaları içerir. + + + + tesseract-tools + tesseract için araçlar + tesseract-tools, tesseract için araçlar içerir. + + \ No newline at end of file