CVS commit: pkgsrc/graphics/tesseract

["Adam Ciarcinski" <adam%netbsd.org@localhost>]

Module Name:    pkgsrc
Committed By:   adam
Date:           Mon Jan 30 07:57:28 UTC 2023

Modified Files:
        pkgsrc/graphics/tesseract: Makefile PLIST buildlink3.mk distinfo
        pkgsrc/graphics/tesseract/patches: patch-CMakeLists.txt

Log Message:
tesseract: updated to 5.3.0

5.3.0

Fix memory issues in ScrollView::MessageReceiver
autotools: Add rule for svpaint executable
Replace call of exit function by return statement in main function
Fix the build on CodeQL/Analyze
CI: Remove Ubuntu 18.04
configure.ac: fix build on aarch64_be
SW CI: Add paths filter
Create .mailmap
Fix tesseract.pc from cmake to match autotools
Update README.md
Fixed 2 errors
fix issue 3940 - remove colormap before thresholding
Update upload-artifact action
Update checkout action to version 3
Fix Markdownlint
Fix broken links in CONTRIBUTING.md
pdfrenderer.cpp: Ignore non-text blocks
lstm.train: allow .box from .raw.png too
Fix a number of performance issues
Fix training tools for legacy engine
Fix function tesseract::WriteFeature
Modernize function ObjectCache::DeleteUnusedObjects
More fixes for issue


To generate a diff of this commit:
cvs rdiff -u -r1.63 -r1.64 pkgsrc/graphics/tesseract/Makefile
cvs rdiff -u -r1.12 -r1.13 pkgsrc/graphics/tesseract/PLIST
cvs rdiff -u -r1.4 -r1.5 pkgsrc/graphics/tesseract/buildlink3.mk
cvs rdiff -u -r1.26 -r1.27 pkgsrc/graphics/tesseract/distinfo
cvs rdiff -u -r1.1 -r1.2 \
    pkgsrc/graphics/tesseract/patches/patch-CMakeLists.txt

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:1.63 pkgsrc/graphics/tesseract/Makefile:1.64
--- pkgsrc/graphics/tesseract/Makefile:1.63     Sun Jan 29 21:16:47 2023
+++ pkgsrc/graphics/tesseract/Makefile  Mon Jan 30 07:57:28 2023
@@ -1,7 +1,6 @@
-# $NetBSD: Makefile,v 1.63 2023/01/29 21:16:47 ryoon Exp $
+# $NetBSD: Makefile,v 1.64 2023/01/30 07:57:28 adam Exp $
 
-DISTNAME=      tesseract-4.1.1
-PKGREVISION=   13
+DISTNAME=      tesseract-5.3.0
 CATEGORIES=    graphics
 MASTER_SITES=  ${MASTER_SITE_GITHUB:=tesseract-ocr/}
 DISTFILES=     ${DEFAULT_DISTFILES}
@@ -11,9 +10,9 @@ HOMEPAGE=     https://github.com/tesseract-o
 COMMENT=       Open Source OCR Engine
 LICENSE=       apache-2.0
 
-LANGVER=       4.0.0
+LANGVER=       4.1.0
 DISTFILES+=    tessdata-${LANGVER}${EXTRACT_SUFX}
-SITES.tessdata-${LANGVER}.tar.gz=      -${MASTER_SITES:Q}tessdata/archive/${LANGVER}.tar.gz
+SITES.tessdata-${LANGVER}.tar.gz=      -${MASTER_SITES:Q}tessdata_best/archive/${LANGVER}.tar.gz
 
 BUILD_DEPENDS+=                asciidoc-[0-9]*:../../textproc/asciidoc
 
@@ -22,12 +21,14 @@ USE_LIBTOOL=                yes
 USE_TOOLS+=            autoconf automake bash gmake pax pkg-config
 GNU_CONFIGURE=         yes
 CONFIGURE_ENV+=                LIBLEPT_HEADERSDIR=${BUILDLINK_PREFIX.leptonica}/include
-
 INSTALL_TARGET=                install training-install
 REPLACE_BASH+=         src/training/*.sh
 
+#CMAKE_ARGS+=  -DUSE_SYSTEM_ICU=ON
+
 post-extract:
-       ${MV} ${WRKDIR}/tessdata-${LANGVER}/* ${WRKSRC}/tessdata
+       ${MV} ${WRKDIR}/tessdata_best-${LANGVER}/*.traineddata ${WRKSRC}/tessdata/
+       ${MV} ${WRKDIR}/tessdata_best-${LANGVER}/script ${WRKSRC}/tessdata/
 
 # For some reason this release tarball lacks these
 pre-configure:
@@ -42,6 +43,8 @@ post-install:
        cd ${WRKSRC}/tessdata && ${SETENV} ${INSTALL_ENV} ${MAKE_ENV} \
                ${MAKE_PROGRAM} ${MAKE_FLAGS} ${INSTALL_MAKE_FLAGS} install-langs
 
+.include "../../archivers/libarchive/buildlink3.mk"
+#.include "../../devel/cmake/build.mk"
 .include "../../devel/pango/buildlink3.mk"
 .include "../../graphics/cairo/buildlink3.mk"
 .include "../../graphics/leptonica/buildlink3.mk"

Index: pkgsrc/graphics/tesseract/PLIST
diff -u pkgsrc/graphics/tesseract/PLIST:1.12 pkgsrc/graphics/tesseract/PLIST:1.13
--- pkgsrc/graphics/tesseract/PLIST:1.12        Mon Jul  8 18:37:03 2019
+++ pkgsrc/graphics/tesseract/PLIST     Mon Jan 30 07:57:28 2023
@@ -1,11 +1,10 @@
-@comment $NetBSD: PLIST,v 1.12 2019/07/08 18:37:03 adam Exp $
+@comment $NetBSD: PLIST,v 1.13 2023/01/30 07:57:28 adam Exp $
 bin/ambiguous_words
 bin/classifier_tester
 bin/cntraining
 bin/combine_lang_model
 bin/combine_tessdata
 bin/dawg2wordlist
-bin/language-specific.sh
 bin/lstmeval
 bin/lstmtraining
 bin/merge_unicharsets
@@ -13,30 +12,21 @@ bin/mftraining
 bin/set_unicharset_properties
 bin/shapeclustering
 bin/tesseract
-bin/tesstrain.sh
-bin/tesstrain_utils.sh
 bin/text2image
 bin/unicharset_extractor
 bin/wordlist2dawg
-include/tesseract/apitypes.h
 include/tesseract/baseapi.h
 include/tesseract/capi.h
-include/tesseract/genericvector.h
-include/tesseract/helpers.h
+include/tesseract/export.h
 include/tesseract/ltrresultiterator.h
 include/tesseract/ocrclass.h
 include/tesseract/osdetect.h
 include/tesseract/pageiterator.h
-include/tesseract/platform.h
 include/tesseract/publictypes.h
 include/tesseract/renderer.h
 include/tesseract/resultiterator.h
-include/tesseract/serialis.h
-include/tesseract/strngs.h
-include/tesseract/tess_version.h
-include/tesseract/tesscallback.h
-include/tesseract/thresholder.h
 include/tesseract/unichar.h
+include/tesseract/version.h
 lib/libtesseract.la
 lib/pkgconfig/tesseract.pc
 man/man1/ambiguous_words.1
@@ -105,9 +95,7 @@ share/tessdata/configs/wordstrbox
 share/tessdata/cos.traineddata
 share/tessdata/cym.traineddata
 share/tessdata/dan.traineddata
-share/tessdata/dan_frak.traineddata
 share/tessdata/deu.traineddata
-share/tessdata/deu_frak.traineddata
 share/tessdata/div.traineddata
 share/tessdata/dzo.traineddata
 share/tessdata/ell.traineddata
@@ -116,7 +104,6 @@ share/tessdata/eng.user-patterns
 share/tessdata/eng.user-words
 share/tessdata/enm.traineddata
 share/tessdata/epo.traineddata
-share/tessdata/equ.traineddata
 share/tessdata/est.traineddata
 share/tessdata/eus.traineddata
 share/tessdata/fao.traineddata
@@ -152,10 +139,9 @@ share/tessdata/kat_old.traineddata
 share/tessdata/kaz.traineddata
 share/tessdata/khm.traineddata
 share/tessdata/kir.traineddata
+share/tessdata/kmr.traineddata
 share/tessdata/kor.traineddata
 share/tessdata/kor_vert.traineddata
-share/tessdata/kur.traineddata
-share/tessdata/kur_ara.traineddata
 share/tessdata/lao.traineddata
 share/tessdata/lat.traineddata
 share/tessdata/lav.traineddata
@@ -186,7 +172,6 @@ share/tessdata/rus.traineddata
 share/tessdata/san.traineddata
 share/tessdata/sin.traineddata
 share/tessdata/slk.traineddata
-share/tessdata/slk_frak.traineddata
 share/tessdata/slv.traineddata
 share/tessdata/snd.traineddata
 share/tessdata/spa.traineddata
@@ -208,7 +193,6 @@ share/tessdata/tessconfigs/msdemo
 share/tessdata/tessconfigs/nobatch
 share/tessdata/tessconfigs/segdemo
 share/tessdata/tgk.traineddata
-share/tessdata/tgl.traineddata
 share/tessdata/tha.traineddata
 share/tessdata/tir.traineddata
 share/tessdata/ton.traineddata

Index: pkgsrc/graphics/tesseract/buildlink3.mk
diff -u pkgsrc/graphics/tesseract/buildlink3.mk:1.4 pkgsrc/graphics/tesseract/buildlink3.mk:1.5
--- pkgsrc/graphics/tesseract/buildlink3.mk:1.4 Tue Jan  3 17:36:27 2023
+++ pkgsrc/graphics/tesseract/buildlink3.mk     Mon Jan 30 07:57:28 2023
@@ -1,4 +1,4 @@
-# $NetBSD: buildlink3.mk,v 1.4 2023/01/03 17:36:27 wiz Exp $
+# $NetBSD: buildlink3.mk,v 1.5 2023/01/30 07:57:28 adam Exp $
 
 BUILDLINK_TREE+=       tesseract
 
@@ -6,7 +6,7 @@ BUILDLINK_TREE+=        tesseract
 TESSERACT_BUILDLINK3_MK:=
 
 BUILDLINK_API_DEPENDS.tesseract+=      tesseract>=3.05.01nb6
-BUILDLINK_ABI_DEPENDS.tesseract?=      tesseract>=4.1.1nb12
+BUILDLINK_ABI_DEPENDS.tesseract+=      tesseract>=4.1.1nb12
 BUILDLINK_PKGSRCDIR.tesseract?=                ../../graphics/tesseract
 
 .include "../../graphics/leptonica/buildlink3.mk"

Index: pkgsrc/graphics/tesseract/distinfo
diff -u pkgsrc/graphics/tesseract/distinfo:1.26 pkgsrc/graphics/tesseract/distinfo:1.27
--- pkgsrc/graphics/tesseract/distinfo:1.26     Tue Oct 26 10:46:56 2021
+++ pkgsrc/graphics/tesseract/distinfo  Mon Jan 30 07:57:28 2023
@@ -1,10 +1,10 @@
-$NetBSD: distinfo,v 1.26 2021/10/26 10:46:56 nia Exp $
+$NetBSD: distinfo,v 1.27 2023/01/30 07:57:28 adam Exp $
 
-BLAKE2s (tessdata-4.0.0.tar.gz) = 1bd390cbb6f59cf52a50d062121883ea1be58b23cd408a78a58644a720038800
-SHA512 (tessdata-4.0.0.tar.gz) = cd71bb99d44eefb53b359ba64b472c509fff773b2737a8d51e10d5d52d9a3a7ff870d470b1c72a7c78be3263b5ecfbb58a6eab13cf7128d8599681676cdcef6b
-Size (tessdata-4.0.0.tar.gz) = 669258747 bytes
-BLAKE2s (tesseract-4.1.1.tar.gz) = f3bdf47c24202445d85d8df94283de127e95e802163cc8f104de1566f302eccd
-SHA512 (tesseract-4.1.1.tar.gz) = 017723a2268be789fe98978eed02fd294968cc8050dde376dee026f56f2b99df42db935049ae5e72c4519a920e263b40af1a6a40d9942e66608145b3131a71a2
-Size (tesseract-4.1.1.tar.gz) = 1974988 bytes
-SHA1 (patch-CMakeLists.txt) = d54f14faadc70cffac45fa6c3b9612845bf6ed27
+BLAKE2s (tessdata-4.1.0.tar.gz) = 639157a6e38486715a14670ae0d93884e57e7e3c1353ed6dc1fe4ec26cf7d159
+SHA512 (tessdata-4.1.0.tar.gz) = fbb8c43cbbbc74789a78d1930d859ba1631058c04a57c7b67350ef6245d705a1ad1df3f899b54b30bbdf253216a6707bbb0e589a5876062e9de04b2b8cb5f872
+Size (tessdata-4.1.0.tar.gz) = 1387839134 bytes
+BLAKE2s (tesseract-5.3.0.tar.gz) = d2412da0202f24c01469ce3f0b3720baf6a303f499b4e2e770ab7e50e5905855
+SHA512 (tesseract-5.3.0.tar.gz) = fc0cfda980059390b7d6d3d8e475b4ecc9dcfee1abc3b16fc9ffc8904db0a0847a5f6f1b32bd2e1f6f5a714a8c4df0fb3dc7c6aa25ece6162bf52fe86b4d0f34
+Size (tesseract-5.3.0.tar.gz) = 1913678 bytes
+SHA1 (patch-CMakeLists.txt) = e29e3b732d45551795046cb5cd09d6d8735269aa
 SHA1 (patch-tessdata_Makefile.am) = 8fe773d1c6318392296ba06996b51692edf32919

Index: pkgsrc/graphics/tesseract/patches/patch-CMakeLists.txt
diff -u pkgsrc/graphics/tesseract/patches/patch-CMakeLists.txt:1.1 pkgsrc/graphics/tesseract/patches/patch-CMakeLists.txt:1.2
--- pkgsrc/graphics/tesseract/patches/patch-CMakeLists.txt:1.1  Fri Jul 16 09:16:27 2021
+++ pkgsrc/graphics/tesseract/patches/patch-CMakeLists.txt      Mon Jan 30 07:57:28 2023
@@ -1,15 +1,15 @@
-$NetBSD: patch-CMakeLists.txt,v 1.1 2021/07/16 09:16:27 jperkin Exp $
+$NetBSD: patch-CMakeLists.txt,v 1.2 2023/01/30 07:57:28 adam Exp $
 
 Avoid C++ <version> issue on case-insensitive file systems.
 
---- CMakeLists.txt.orig        2019-12-26 14:21:51.000000000 +0000
+--- CMakeLists.txt.orig        2022-12-22 13:57:57.000000000 +0000
 +++ CMakeLists.txt
-@@ -41,7 +41,7 @@ set_property(GLOBAL PROPERTY PREDEFINED_
+@@ -48,7 +48,7 @@ endif()
  project(tesseract C CXX)
  
  # Get version with components from VERSION file.
 -file(STRINGS "VERSION" VERSION_PLAIN)
 +file(STRINGS "VERSION.txt" VERSION_PLAIN)
  string(REGEX REPLACE "^([^.]*)\\..*" "\\1" VERSION_MAJOR ${VERSION_PLAIN})
- string(REGEX REPLACE "^[^.]*\\.([^.]*)\\..*" "\\1" VERSION_MINOR ${VERSION_PLAIN})
- string(REGEX REPLACE "^[^.]*\\.[^.]*\\.([0-9]*).*" "\\1" VERSION_PATCH ${VERSION_PLAIN})
+ string(REGEX REPLACE "^[^.]*\\.([^.]*)\\..*" "\\1" VERSION_MINOR
+                      ${VERSION_PLAIN})