[pkgsrc/trunk]: pkgsrc/graphics/tesseract tesseract: updated to 4.1.0

[adam <adam%pkgsrc.org@localhost>]

details:   https://anonhg.NetBSD.org/pkgsrc/rev/acbefd433dfc
branches:  trunk
changeset: 398003:acbefd433dfc
user:      adam <adam%pkgsrc.org@localhost>
date:      Mon Jul 08 18:37:03 2019 +0000

description:
tesseract: updated to 4.1.0

4.1.0 Release
Added new renders Alto, LSTMBox, WordStrBox.
Added character boxes in hOCR output.
Added python training scripts (experimental) as alternative shell scripts.
Better support AVX / AVX2 / SSE.
Disable OpenMP support by default.
Fix for bounding box problem.
Implemented support for whitelist/blacklist in LSTM engine.
Improved cmake configuration.
Code modernization and improvements.
A lot of bug fixes...

diffstat:

 graphics/tesseract/Makefile                           |   5 +--
 graphics/tesseract/PLIST                              |   7 +++-
 graphics/tesseract/distinfo                           |  13 ++++-----
 graphics/tesseract/patches/patch-doc_Makefile.am      |  24 -------------------
 graphics/tesseract/patches/patch-tessdata_Makefile.am |  10 +++---
 5 files changed, 18 insertions(+), 41 deletions(-)

diffs (131 lines):

diff -r 60e6f9874905 -r acbefd433dfc graphics/tesseract/Makefile
--- a/graphics/tesseract/Makefile       Mon Jul 08 17:57:00 2019 +0000
+++ b/graphics/tesseract/Makefile       Mon Jul 08 18:37:03 2019 +0000
@@ -1,7 +1,6 @@
-# $NetBSD: Makefile,v 1.47 2019/05/04 16:05:33 leot Exp $
+# $NetBSD: Makefile,v 1.48 2019/07/08 18:37:03 adam Exp $
 
-DISTNAME=      tesseract-4.0.0
-PKGREVISION=   6
+DISTNAME=      tesseract-4.1.0
 CATEGORIES=    graphics
 MASTER_SITES=  ${MASTER_SITE_GITHUB:=tesseract-ocr/}
 DISTFILES=     ${DEFAULT_DISTFILES}
diff -r 60e6f9874905 -r acbefd433dfc graphics/tesseract/PLIST
--- a/graphics/tesseract/PLIST  Mon Jul 08 17:57:00 2019 +0000
+++ b/graphics/tesseract/PLIST  Mon Jul 08 18:37:03 2019 +0000
@@ -1,4 +1,4 @@
-@comment $NetBSD: PLIST,v 1.11 2018/11/18 18:07:20 adam Exp $
+@comment $NetBSD: PLIST,v 1.12 2019/07/08 18:37:03 adam Exp $
 bin/ambiguous_words
 bin/classifier_tester
 bin/cntraining
@@ -23,7 +23,6 @@
 include/tesseract/capi.h
 include/tesseract/genericvector.h
 include/tesseract/helpers.h
-include/tesseract/host.h
 include/tesseract/ltrresultiterator.h
 include/tesseract/ocrclass.h
 include/tesseract/osdetect.h
@@ -78,18 +77,21 @@
 share/tessdata/chi_tra.traineddata
 share/tessdata/chi_tra_vert.traineddata
 share/tessdata/chr.traineddata
+share/tessdata/configs/alto
 share/tessdata/configs/ambigs.train
 share/tessdata/configs/api_config
 share/tessdata/configs/bigram
 share/tessdata/configs/box.train
 share/tessdata/configs/box.train.stderr
 share/tessdata/configs/digits
+share/tessdata/configs/get.images
 share/tessdata/configs/hocr
 share/tessdata/configs/inter
 share/tessdata/configs/kannada
 share/tessdata/configs/linebox
 share/tessdata/configs/logfile
 share/tessdata/configs/lstm.train
+share/tessdata/configs/lstmbox
 share/tessdata/configs/lstmdebug
 share/tessdata/configs/makebox
 share/tessdata/configs/pdf
@@ -99,6 +101,7 @@
 share/tessdata/configs/tsv
 share/tessdata/configs/txt
 share/tessdata/configs/unlv
+share/tessdata/configs/wordstrbox
 share/tessdata/cos.traineddata
 share/tessdata/cym.traineddata
 share/tessdata/dan.traineddata
diff -r 60e6f9874905 -r acbefd433dfc graphics/tesseract/distinfo
--- a/graphics/tesseract/distinfo       Mon Jul 08 17:57:00 2019 +0000
+++ b/graphics/tesseract/distinfo       Mon Jul 08 18:37:03 2019 +0000
@@ -1,12 +1,11 @@
-$NetBSD: distinfo,v 1.21 2019/05/04 16:05:33 leot Exp $
+$NetBSD: distinfo,v 1.22 2019/07/08 18:37:03 adam Exp $
 
 SHA1 (tessdata-4.0.0.tar.gz) = 94557a6ecdf8ff8bec131598759e7d3b0bca1911
 RMD160 (tessdata-4.0.0.tar.gz) = 2e826e866b56ff8b9cb2c6613f04d8c4a4ff98d7
 SHA512 (tessdata-4.0.0.tar.gz) = cd71bb99d44eefb53b359ba64b472c509fff773b2737a8d51e10d5d52d9a3a7ff870d470b1c72a7c78be3263b5ecfbb58a6eab13cf7128d8599681676cdcef6b
 Size (tessdata-4.0.0.tar.gz) = 669258747 bytes
-SHA1 (tesseract-4.0.0.tar.gz) = 243a4919d44bc64d1e7e4cac660c716c845a8d03
-RMD160 (tesseract-4.0.0.tar.gz) = 0e95d343639ab98c6d3fbc528053b627b6e12282
-SHA512 (tesseract-4.0.0.tar.gz) = 69e57d4ba1fc43d212fd0fff69a2b5d48a3b37cfee7054fdc083cbb7e04d92317609a32e457229661d70ce8d9b16c9d25e81bfc3861db660dd2c8f292202d447
-Size (tesseract-4.0.0.tar.gz) = 1961372 bytes
-SHA1 (patch-doc_Makefile.am) = fd9cc782e766428de5709b77d7a2476be55ec7d8
-SHA1 (patch-tessdata_Makefile.am) = 1fdbed9dafc1527eb52f354c8b78ba82f854b350
+SHA1 (tesseract-4.1.0.tar.gz) = 6e88cc4fd9f1681142bf74dc2df0559202cff3c2
+RMD160 (tesseract-4.1.0.tar.gz) = 034ffd9690478e28945c09001ce51f7fdceb2ff5
+SHA512 (tesseract-4.1.0.tar.gz) = d617f5c5b826640b2871dbe3d7973bcc5e66fafd837921a20e009d683806ed50f0f258aa455019d99fc54f5cb65c2fa0380e3a3c92b39ab0684b8799c730b09d
+Size (tesseract-4.1.0.tar.gz) = 1965053 bytes
+SHA1 (patch-tessdata_Makefile.am) = 8fe773d1c6318392296ba06996b51692edf32919
diff -r 60e6f9874905 -r acbefd433dfc graphics/tesseract/patches/patch-doc_Makefile.am
--- a/graphics/tesseract/patches/patch-doc_Makefile.am  Mon Jul 08 17:57:00 2019 +0000
+++ /dev/null   Thu Jan 01 00:00:00 1970 +0000
@@ -1,24 +0,0 @@
-$NetBSD: patch-doc_Makefile.am,v 1.1 2018/11/29 09:15:23 adam Exp $
-
-Properly format manpages.
-
---- doc/Makefile.am.orig       2018-11-29 08:57:43.000000000 +0000
-+++ doc/Makefile.am
-@@ -2,7 +2,7 @@
- 
- if ASCIIDOC
- 
--asciidoc=asciidoc -d manpage
-+asciidoc=a2x -f manpage
- 
- 
- man_MANS = \
-@@ -36,7 +36,7 @@ EXTRA_DIST = $(man_MANS) Doxyfile
- html: $(patsubst %,%.html,$(man_MANS))
- 
- %: %.asc
--      $(asciidoc) -o $@ $<
-+      $(asciidoc) $<
- 
- %.html: %.asc
-       asciidoc -b html5 -o $@ $<
diff -r 60e6f9874905 -r acbefd433dfc graphics/tesseract/patches/patch-tessdata_Makefile.am
--- a/graphics/tesseract/patches/patch-tessdata_Makefile.am     Mon Jul 08 17:57:00 2019 +0000
+++ b/graphics/tesseract/patches/patch-tessdata_Makefile.am     Mon Jul 08 18:37:03 2019 +0000
@@ -1,12 +1,12 @@
-$NetBSD: patch-tessdata_Makefile.am,v 1.3 2019/05/04 16:05:33 leot Exp $
+$NetBSD: patch-tessdata_Makefile.am,v 1.4 2019/07/08 18:37:03 adam Exp $
 
 Revert a trunk commit that broke install-lang for tesseract<4.
 
---- tessdata/Makefile.am.orig  2018-10-29 08:53:12.000000000 +0000
+--- tessdata/Makefile.am.orig  2019-07-07 12:34:08.000000000 +0000
 +++ tessdata/Makefile.am
-@@ -29,6 +29,27 @@ langdata = bul.traineddata mlt.trainedda
-       chi_tra.traineddata ita.traineddata spa_old.traineddata \
-       deu-frak.traineddata aze.traineddata
+@@ -7,6 +7,27 @@ SUBDIRS = configs tessconfigs
+ 
+ langdata =
  
 +.PHONY: install-langs
 +install-langs: