CVS commit: pkgsrc/graphics/tesseract

["Filip Hajny" <fhajny%netbsd.org@localhost>]

Module Name:    pkgsrc
Committed By:   fhajny
Date:           Tue Feb 21 17:51:18 UTC 2017

Modified Files:
        pkgsrc/graphics/tesseract: Makefile PLIST distinfo
Added Files:
        pkgsrc/graphics/tesseract/patches: patch-tessdata_Makefile.am
Removed Files:
        pkgsrc/graphics/tesseract/patches: patch-viewer_svutil_cpp

Log Message:
Update graphics/tesseract to 3.05.00

- Made some fine tuning to the hOCR output.
- Added TSV as another optional output format.
- Fixed ABI break introduced in 3.04.00 with the AnalyseLayout()
  method.
- text2image tool - Enable all OpenType ligatures available in a font.
  This feature requires Pango 1.38 or newer.
- Training tools - Replaced asserts with tprintf() and exit(1).
- Fixed Cygwin compatibility.
- Improved multipage tiff processing.
- Improved the embedded pdf font (pdf.ttf).
- Enable selection of OCR engine mode from command line.
- Changed tesseract command line parameter '-psm' to '--psm'.
- Added new C API for orientation and script detection, removed the
  old one.
- Increased minimum autoconf version to 2.59.
- Removed dead code.
- Fixed many compiler warning.
- Fixed memory and resource leaks.
- Fixed some issues with the 'Cube' OCR engine.
- Fixed some openCL issues.
- Added option to build Tesseract with CMake build system.
- Implemented CPPAN support for easy Windows building.


To generate a diff of this commit:
cvs rdiff -u -r1.24 -r1.25 pkgsrc/graphics/tesseract/Makefile
cvs rdiff -u -r1.8 -r1.9 pkgsrc/graphics/tesseract/PLIST
cvs rdiff -u -r1.13 -r1.14 pkgsrc/graphics/tesseract/distinfo
cvs rdiff -u -r0 -r1.1 \
    pkgsrc/graphics/tesseract/patches/patch-tessdata_Makefile.am
cvs rdiff -u -r1.2 -r0 \
    pkgsrc/graphics/tesseract/patches/patch-viewer_svutil_cpp

Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.

Modified files:

Index: pkgsrc/graphics/tesseract/Makefile
diff -u pkgsrc/graphics/tesseract/Makefile:1.24 pkgsrc/graphics/tesseract/Makefile:1.25
--- pkgsrc/graphics/tesseract/Makefile:1.24     Sun Feb 12 06:25:31 2017
+++ pkgsrc/graphics/tesseract/Makefile  Tue Feb 21 17:51:18 2017
@@ -1,7 +1,6 @@
-# $NetBSD: Makefile,v 1.24 2017/02/12 06:25:31 ryoon Exp $
+# $NetBSD: Makefile,v 1.25 2017/02/21 17:51:18 fhajny Exp $
 
-DISTNAME=      tesseract-3.04.01
-PKGREVISION=   4
+DISTNAME=      tesseract-3.05.00
 CATEGORIES=    graphics
 MASTER_SITES=  ${MASTER_SITE_GITHUB:=tesseract-ocr/}
 DISTFILES=     ${DEFAULT_DISTFILES}
@@ -17,7 +16,7 @@ SITES.tessdata-${LANGVER}.tar.gz=     -${MAS
 
 USE_LANGUAGES=         c c++
 USE_LIBTOOL=           yes
-USE_TOOLS+=            gmake pax pkg-config
+USE_TOOLS+=            autoconf automake make pax pkg-config
 GNU_CONFIGURE=         yes
 
 CONFIGURE_ENV+=                LIBLEPT_HEADERSDIR=${BUILDLINK_PREFIX.leptonica}/include
@@ -29,6 +28,10 @@ INSTALLATION_DIRS=   libexec share/doc/tes
 post-extract:
        ${MV} ${WRKDIR}/tessdata-${LANGVER}/* ${WRKSRC}/tessdata
 
+# For some reason this release tarball lacks these
+pre-configure:
+       cd ${WRKSRC} && ${SETENV} ${CONFIGURE_ENV} ${SH} ./autogen.sh
+
 # Building the training tools requires that the base build is done
 post-build:
        cd ${WRKSRC} && ${BUILD_MAKE_CMD} training

Index: pkgsrc/graphics/tesseract/PLIST
diff -u pkgsrc/graphics/tesseract/PLIST:1.8 pkgsrc/graphics/tesseract/PLIST:1.9
--- pkgsrc/graphics/tesseract/PLIST:1.8 Thu Mar 17 12:51:14 2016
+++ pkgsrc/graphics/tesseract/PLIST     Tue Feb 21 17:51:18 2017
@@ -1,4 +1,4 @@
-@comment $NetBSD: PLIST,v 1.8 2016/03/17 12:51:14 fhajny Exp $
+@comment $NetBSD: PLIST,v 1.9 2017/02/21 17:51:18 fhajny Exp $
 bin/ambiguous_words
 bin/classifier_tester
 bin/cntraining
@@ -91,6 +91,7 @@ share/tessdata/configs/pdf
 share/tessdata/configs/quiet
 share/tessdata/configs/rebox
 share/tessdata/configs/strokewidth
+share/tessdata/configs/tsv
 share/tessdata/configs/txt
 share/tessdata/configs/unlv
 share/tessdata/cym.traineddata

Index: pkgsrc/graphics/tesseract/distinfo
diff -u pkgsrc/graphics/tesseract/distinfo:1.13 pkgsrc/graphics/tesseract/distinfo:1.14
--- pkgsrc/graphics/tesseract/distinfo:1.13     Thu Mar 17 12:51:14 2016
+++ pkgsrc/graphics/tesseract/distinfo  Tue Feb 21 17:51:18 2017
@@ -1,12 +1,12 @@
-$NetBSD: distinfo,v 1.13 2016/03/17 12:51:14 fhajny Exp $
+$NetBSD: distinfo,v 1.14 2017/02/21 17:51:18 fhajny Exp $
 
 SHA1 (tessdata-3.04.00.tar.gz) = 6ea24cccf0e823da98589ccc75d51f0950618236
 RMD160 (tessdata-3.04.00.tar.gz) = 0a3c3b3c127b6031e2e037d78e3a6f159fb9e869
 SHA512 (tessdata-3.04.00.tar.gz) = 4fbb66137c729e16c7a9e35b09916a45c1bb5ec5a7002a22647e0b10975362cb44c6d6c0c997baf25866f78749ec2d4a86317ec3fb664bd963243e230516d162
 Size (tessdata-3.04.00.tar.gz) = 499088801 bytes
-SHA1 (tesseract-3.04.01.tar.gz) = 359ffc1925f0270ca100a2b4c1d3b41f4b23701d
-RMD160 (tesseract-3.04.01.tar.gz) = 5e754411afa74cfc4e6b601fe2c770ba93a25f23
-SHA512 (tesseract-3.04.01.tar.gz) = 816f6c29f8efeaf157891ddb022bd0daf8eb27c0cb8b6df5542e745d209f0ce617245e4583c59a561ed3216044d3afb38f8620ee94cdea656364923dcb140b5b
-Size (tesseract-3.04.01.tar.gz) = 2269105 bytes
+SHA1 (tesseract-3.05.00.tar.gz) = 93c036ce510a6ae40db4d2f30844d3dba1e68faa
+RMD160 (tesseract-3.05.00.tar.gz) = 39705f462aaa160582d9f2a7886e692712523f88
+SHA512 (tesseract-3.05.00.tar.gz) = 54ba96bd1b0d19883e8a0f717ecfddd459945f9dda0fd340e29e2dab3af94d0eceb7f3d207f34685ee97d1ec763efac5d94dec7e0498a7ca85fec35da4ed76b1
+Size (tesseract-3.05.00.tar.gz) = 3581853 bytes
+SHA1 (patch-tessdata_Makefile.am) = 013c9b4bbf64a0948a362d334e6b86a240aa944f
 SHA1 (patch-viewer_scrollview.cpp) = 05a9ff5d2a9e302b3a682144db54c612fd4eccc2
-SHA1 (patch-viewer_svutil_cpp) = ec6f8b79c66993e5272a4f2a13aa92f2432cbea0

Added files:

Index: pkgsrc/graphics/tesseract/patches/patch-tessdata_Makefile.am
diff -u /dev/null pkgsrc/graphics/tesseract/patches/patch-tessdata_Makefile.am:1.1
--- /dev/null   Tue Feb 21 17:51:18 2017
+++ pkgsrc/graphics/tesseract/patches/patch-tessdata_Makefile.am        Tue Feb 21 17:51:18 2017
@@ -0,0 +1,34 @@
+$NetBSD: patch-tessdata_Makefile.am,v 1.1 2017/02/21 17:51:18 fhajny Exp $
+
+Revert a trunk commit that broke install-lang for tesseract<4.
+
+--- tessdata/Makefile.am.orig  2017-02-16 17:59:48.000000000 +0000
++++ tessdata/Makefile.am
+@@ -44,6 +44,27 @@ langdata = bul.traineddata mlt.trainedda
+       ita.cube.nn fra.cube.size eng.cube.bigrams ara.cube.lm \
+       rus.cube.nn spa.cube.nn hin.cube.bigrams
+ 
++.PHONY: install-langs
++install-langs:
++      @if [ ! -d  $(DESTDIR)$(datadir) ]; then mkdir -p $(DESTDIR)$(datadir); fi;
++      @if test "${LANGS}" != ""; then \
++              for lang_code in ${LANGS}; do \
++                      echo "installing data for $$lang_code"; \
++                      $(INSTALL) -m 644 $(srcdir)/$$lang_code.* $(DESTDIR)$(datadir); \
++              done; \
++      else \
++              for l in ./*.traineddata; do \
++                      filename=`basename $$l`; \
++                      lang_code=$${filename%.*}; \
++                      if test "$$lang_code" == "*"; then \
++                              echo "No lang present."; \
++                              break; \
++                      fi; \
++                      echo "installing data for $$lang_code"; \
++                      $(INSTALL) -m 644 $(srcdir)/$$lang_code.* $(DESTDIR)$(datadir); \
++              done; \
++      fi;
++
+ uninstall-local:
+       cd $(DESTDIR)$(datadir); \
+       rm -f $(langdata)