From 31884c60db9d3c95855e250e895e79dba05bdc53 Mon Sep 17 00:00:00 2001
From: joschrew <jonas.schrewe@gwdg.de>
Date: Wed, 31 Jan 2024 15:22:10 +0100
Subject: [PATCH 01/13] Update dockerfile

---
 Dockerfile | 77 ++++++++++++++++++++++++++++++++++++++++++++++++------
 1 file changed, 69 insertions(+), 8 deletions(-)

diff --git a/Dockerfile b/Dockerfile
index d818bc3..9190dfa 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,4 +1,65 @@
-FROM ocrd/core
+FROM ocrd/core:v2.62.0 AS base
+# set proper locales
+ENV LANG C.UTF-8
+ENV LC_ALL C.UTF-8
+
+# set TESSDATA_PREFIX
+ENV TESSDATA_PREFIX /usr/local/share/tessdata
+
+ARG TESSERACT_REPOSITORY=https://github.com/tesseract-ocr/tesseract.git
+ARG TESSERACT_REF=5.3.3
+
+# set frontend non-interactive to silence interactive tzdata config
+ARG DEBIAN_FRONTEND=noninteractive
+
+
+# install common tools and tesseract build dependencies
+# use provided leptonica
+# tzdata required for proper timezone settings
+RUN apt-get update && apt-get install -y \
+	apt-utils \
+	build-essential \
+	g++ \
+	git \
+	libjpeg-dev \
+	libgif-dev \
+	libwebp-dev \
+	libopenjp2-7-dev \
+	libpng-dev \
+	libtiff-dev \
+	libtool \
+	pkg-config \
+	tzdata \
+	xzgv \
+	zlib1g-dev \
+	libleptonica-dev \
+	libpango1.0-dev \
+	libicu-dev \
+	autotools-dev \
+	automake \
+	libcurl4-nss-dev \
+	libarchive-dev
+
+# set proper date and timezone in container
+RUN echo "Europe/Berlin" > /etc/timezone
+RUN ln -sf /usr/share/zoneinfo/Europe/Berlin /etc/localtime
+RUN dpkg-reconfigure -f noninteractive tzdata
+
+# diagnostic output - check timezone settings
+RUN cat /etc/timezone
+
+# clone and checkout desired tesseract version tag
+RUN git clone ${TESSERACT_REPOSITORY} /build_tesseract
+WORKDIR /build_tesseract
+RUN git checkout ${TESSERACT_REF}
+
+# compile
+RUN ./autogen.sh && ./configure --disable-openmp --disable-shared 'CXXFLAGS=-g -O2 -fno-math-errno -Wall -Wextra -Wpedantic'
+RUN LDFLAGS="-L/usr/local/lib" CFLAGS="-I/usr/local/include" make
+RUN make
+RUN make install && ldconfig
+
+# install ocrd-tesserocr (until here commands for installing tesseract-ocr)
 ARG VCS_REF
 ARG BUILD_DATE
 LABEL \
@@ -7,7 +68,6 @@ LABEL \
     org.label-schema.vcs-url="https://github.com/OCR-D/ocrd_tesserocr" \
     org.label-schema.build-date=$BUILD_DATE
 
-ENV DEBIAN_FRONTEND noninteractive
 ENV PYTHONIOENCODING utf8
 
 # avoid HOME/.local/share (hard to predict USER here)
@@ -15,7 +75,7 @@ ENV PYTHONIOENCODING utf8
 # (can still be overridden by derived stages)
 ENV XDG_DATA_HOME /usr/local/share
 
-WORKDIR /build-ocrd
+WORKDIR /build-ocrd_tesserocr
 COPY setup.py .
 COPY ocrd_tesserocr/ocrd-tool.json .
 COPY README.md .
@@ -23,14 +83,15 @@ COPY requirements.txt .
 COPY requirements_test.txt .
 COPY ocrd_tesserocr ./ocrd_tesserocr
 COPY Makefile .
-RUN make deps-ubuntu && \
-    apt-get install -y --no-install-recommends \
-    g++ \
+RUN apt-get install -y --no-install-recommends \
+    python3 \
+    python3-pip \
     && make deps install \
-    && rm -rf /build-ocrd \
+    && rm -rf /build-ocrd_tesserocr \
     && apt-get -y remove --auto-remove g++ libtesseract-dev make
 # PPA tessdata prefix (= ocrd_tesserocr moduledir) is owned by root
-RUN sudo chmod go+w `dpkg-query -L tesseract-ocr-eng | sed -n s,/eng.traineddata,,p`
+# next line causes failure because tesseract-ocr-eng not existing. Not sure if needed, so skipping
+# RUN sudo chmod go+w `dpkg-query -L tesseract-ocr-eng | sed -n s,/eng.traineddata,,p`
 RUN ocrd resmgr download ocrd-tesserocr-recognize Fraktur.traineddata
 RUN ocrd resmgr download ocrd-tesserocr-recognize deu.traineddata
 

From 043b5c4711717564c7319b310d22fe0ee854caf0 Mon Sep 17 00:00:00 2001
From: Konstantin Baierer <unixprog@gmail.com>
Date: Mon, 5 Feb 2024 19:17:44 +0100
Subject: [PATCH 02/13] build tesseract and tesserocr

---
 .dockerignore  |  2 ++
 .gitignore     |  1 +
 .gitmodules    |  6 ++++++
 Dockerfile     | 42 +++++++++++++++---------------------------
 Makefile       | 34 +++++++++++++++++++++++++++++++++-
 repo/tesseract |  1 +
 repo/tesserocr |  1 +
 7 files changed, 59 insertions(+), 28 deletions(-)
 create mode 100644 .gitmodules
 create mode 160000 repo/tesseract
 create mode 160000 repo/tesserocr

diff --git a/.dockerignore b/.dockerignore
index 27c70c6..b4a4830 100644
--- a/.dockerignore
+++ b/.dockerignore
@@ -5,6 +5,8 @@
 !requirements_test.txt
 !LICENSE
 !README.md
+!repo/tesserocr
+!repo/tesseract
 
 # avoid .git and __pycache__ etc:
 !ocrd_tesserocr/**/*.py
diff --git a/.gitignore b/.gitignore
index 38b47c8..c84bd90 100644
--- a/.gitignore
+++ b/.gitignore
@@ -13,3 +13,4 @@ test-workspace
 /.coverage
 /htmlcov
 /.cache
+build_tesseract
diff --git a/.gitmodules b/.gitmodules
new file mode 100644
index 0000000..cfc2d01
--- /dev/null
+++ b/.gitmodules
@@ -0,0 +1,6 @@
+[submodule "repo/tesserocr"]
+	path = repo/tesserocr
+	url = https://github.com/sirfz/tesserocr/
+[submodule "repo/tesseract"]
+	path = repo/tesseract
+	url = https://github.com/tesseract-ocr/tesseract
diff --git a/Dockerfile b/Dockerfile
index 9190dfa..1204553 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -2,13 +2,20 @@ FROM ocrd/core:v2.62.0 AS base
 # set proper locales
 ENV LANG C.UTF-8
 ENV LC_ALL C.UTF-8
+# install ocrd-tesserocr (until here commands for installing tesseract-ocr)
+ARG VCS_REF
+ARG BUILD_DATE
+LABEL \
+    maintainer="https://ocr-d.de/kontakt" \
+    org.label-schema.vcs-ref=$VCS_REF \
+    org.label-schema.vcs-url="https://github.com/OCR-D/ocrd_tesserocr" \
+    org.label-schema.build-date=$BUILD_DATE
+
+ENV PYTHONIOENCODING utf8
 
 # set TESSDATA_PREFIX
 ENV TESSDATA_PREFIX /usr/local/share/tessdata
 
-ARG TESSERACT_REPOSITORY=https://github.com/tesseract-ocr/tesseract.git
-ARG TESSERACT_REF=5.3.3
-
 # set frontend non-interactive to silence interactive tzdata config
 ARG DEBIAN_FRONTEND=noninteractive
 
@@ -46,29 +53,7 @@ RUN ln -sf /usr/share/zoneinfo/Europe/Berlin /etc/localtime
 RUN dpkg-reconfigure -f noninteractive tzdata
 
 # diagnostic output - check timezone settings
-RUN cat /etc/timezone
-
-# clone and checkout desired tesseract version tag
-RUN git clone ${TESSERACT_REPOSITORY} /build_tesseract
-WORKDIR /build_tesseract
-RUN git checkout ${TESSERACT_REF}
-
-# compile
-RUN ./autogen.sh && ./configure --disable-openmp --disable-shared 'CXXFLAGS=-g -O2 -fno-math-errno -Wall -Wextra -Wpedantic'
-RUN LDFLAGS="-L/usr/local/lib" CFLAGS="-I/usr/local/include" make
-RUN make
-RUN make install && ldconfig
-
-# install ocrd-tesserocr (until here commands for installing tesseract-ocr)
-ARG VCS_REF
-ARG BUILD_DATE
-LABEL \
-    maintainer="https://ocr-d.de/kontakt" \
-    org.label-schema.vcs-ref=$VCS_REF \
-    org.label-schema.vcs-url="https://github.com/OCR-D/ocrd_tesserocr" \
-    org.label-schema.build-date=$BUILD_DATE
-
-ENV PYTHONIOENCODING utf8
+# RUN cat /etc/timezone
 
 # avoid HOME/.local/share (hard to predict USER here)
 # so let XDG_DATA_HOME coincide with fixed system location
@@ -82,13 +67,16 @@ COPY README.md .
 COPY requirements.txt .
 COPY requirements_test.txt .
 COPY ocrd_tesserocr ./ocrd_tesserocr
+COPY repo/tesserocr ./repo/tesserocr
+COPY repo/tesseract ./repo/tesseract
 COPY Makefile .
 RUN apt-get install -y --no-install-recommends \
     python3 \
     python3-pip \
-    && make deps install \
+    && make deps install-tesseract install-tesserocr install \
     && rm -rf /build-ocrd_tesserocr \
     && apt-get -y remove --auto-remove g++ libtesseract-dev make
+
 # PPA tessdata prefix (= ocrd_tesserocr moduledir) is owned by root
 # next line causes failure because tesseract-ocr-eng not existing. Not sure if needed, so skipping
 # RUN sudo chmod go+w `dpkg-query -L tesseract-ocr-eng | sed -n s,/eng.traineddata,,p`
diff --git a/Makefile b/Makefile
index e1648ae..40f3295 100644
--- a/Makefile
+++ b/Makefile
@@ -7,8 +7,20 @@ LOG_LEVEL = INFO
 PYTHONIOENCODING=utf8
 LC_ALL = C.UTF-8
 LANG = C.UTF-8
-export
+ifdef VIRTUAL_ENV
+	TESSERACT_PREFIX = $(VIRTUAL_ENV)
+else
+	TESSERACT_PREFIX = /usr/local
+endif
+
+ifeq ($(PKG_CONFIG_PATH),)
+PKG_CONFIG_PATH := $(TESSERACT_PREFIX)/lib/pkgconfig
+else
+PKG_CONFIG_PATH := $(TESSERACT_PREFIX)/lib/pkgconfig:$(PKG_CONFIG_PATH)
+endif
+export PKG_CONFIG_PATH
 
+export
 
 # pytest args. Set to '-s' to see log output during test execution, '--verbose' to see individual tests. Default: '$(PYTEST_ARGS)'
 PYTEST_ARGS =
@@ -85,6 +97,20 @@ docker:
 	--build-arg BUILD_DATE=$$(date -u +"%Y-%m-%dT%H:%M:%SZ") \
 	-t $(DOCKER_TAG) .
 
+install-tesserocr:
+	cd repo/tesserocr; $(PIP) install .
+
+install-tesseract:
+	cd repo/tesseract; ./autogen.sh
+	mkdir -p $(CURDIR)/build_tesseract
+	cd $(CURDIR)/build_tesseract && $(CURDIR)/repo/tesseract/configure \
+				--prefix=$(TESSERACT_PREFIX) \
+				--disable-openmp \
+				--disable-shared \
+				'CXXFLAGS=-g -O2 -fno-math-errno -Wall -Wextra -Wpedantic -fPIC' ;\
+	cd $(CURDIR)/build_tesseract && $(MAKE) install
+	if [[ "$(TESSERACT_PREFIX)" = "/usr"* ]];then ldconfig ;fi
+
 # Install this package
 install: deps
 	$(PIP) install .
@@ -135,6 +161,12 @@ repo/assets:
 	mkdir -p $(dir $@)
 	git clone https://github.com/OCR-D/assets "$@"
 
+.PHONY: clean
+clean: assets-clean tesseract-clean
+
+tesseract-clean:
+	rm -rf $(CURDIR)/build_tesseract
+	cd repo/tesseract; make distclean
 
 .PHONY: assets-clean
 # Remove symlinks in test/assets
diff --git a/repo/tesseract b/repo/tesseract
new file mode 160000
index 0000000..8ee020e
--- /dev/null
+++ b/repo/tesseract
@@ -0,0 +1 @@
+Subproject commit 8ee020e14cf5be4e3f0e9beb09b6b050a1871854
diff --git a/repo/tesserocr b/repo/tesserocr
new file mode 160000
index 0000000..1f960e9
--- /dev/null
+++ b/repo/tesserocr
@@ -0,0 +1 @@
+Subproject commit 1f960e9e0714dcd5ebdcf86248269efb70ccca5b

From 2c05dbcb407dcdb73f0b51083a4c38eaf4cdac13 Mon Sep 17 00:00:00 2001
From: joschrew <jonas.schrewe@gwdg.de>
Date: Thu, 8 Feb 2024 12:08:16 +0100
Subject: [PATCH 03/13] Dockerfile: add eng traineddata

---
 Dockerfile | 1 +
 1 file changed, 1 insertion(+)

diff --git a/Dockerfile b/Dockerfile
index 1204553..33b0e0e 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -82,6 +82,7 @@ RUN apt-get install -y --no-install-recommends \
 # RUN sudo chmod go+w `dpkg-query -L tesseract-ocr-eng | sed -n s,/eng.traineddata,,p`
 RUN ocrd resmgr download ocrd-tesserocr-recognize Fraktur.traineddata
 RUN ocrd resmgr download ocrd-tesserocr-recognize deu.traineddata
+RUN ocrd resmgr download ocrd-tesserocr-recognize eng.traineddata
 
 WORKDIR /data
 VOLUME /data

From 42e2ba57ee4a17bc34f6355aa54e79b85dd97803 Mon Sep 17 00:00:00 2001
From: joschrew <jonas.schrewe@gwdg.de>
Date: Mon, 12 Feb 2024 10:09:07 +0100
Subject: [PATCH 04/13] Move deps install from Dockerfile to Makefile

Co-authored-by: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
---
 Dockerfile | 39 +++++----------------------------------
 Makefile   | 32 ++++++++++++++++++++------------
 2 files changed, 25 insertions(+), 46 deletions(-)

diff --git a/Dockerfile b/Dockerfile
index 33b0e0e..971c556 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -19,34 +19,6 @@ ENV TESSDATA_PREFIX /usr/local/share/tessdata
 # set frontend non-interactive to silence interactive tzdata config
 ARG DEBIAN_FRONTEND=noninteractive
 
-
-# install common tools and tesseract build dependencies
-# use provided leptonica
-# tzdata required for proper timezone settings
-RUN apt-get update && apt-get install -y \
-	apt-utils \
-	build-essential \
-	g++ \
-	git \
-	libjpeg-dev \
-	libgif-dev \
-	libwebp-dev \
-	libopenjp2-7-dev \
-	libpng-dev \
-	libtiff-dev \
-	libtool \
-	pkg-config \
-	tzdata \
-	xzgv \
-	zlib1g-dev \
-	libleptonica-dev \
-	libpango1.0-dev \
-	libicu-dev \
-	autotools-dev \
-	automake \
-	libcurl4-nss-dev \
-	libarchive-dev
-
 # set proper date and timezone in container
 RUN echo "Europe/Berlin" > /etc/timezone
 RUN ln -sf /usr/share/zoneinfo/Europe/Berlin /etc/localtime
@@ -60,7 +32,7 @@ RUN dpkg-reconfigure -f noninteractive tzdata
 # (can still be overridden by derived stages)
 ENV XDG_DATA_HOME /usr/local/share
 
-WORKDIR /build-ocrd_tesserocr
+WORKDIR /build
 COPY setup.py .
 COPY ocrd_tesserocr/ocrd-tool.json .
 COPY README.md .
@@ -70,11 +42,8 @@ COPY ocrd_tesserocr ./ocrd_tesserocr
 COPY repo/tesserocr ./repo/tesserocr
 COPY repo/tesseract ./repo/tesseract
 COPY Makefile .
-RUN apt-get install -y --no-install-recommends \
-    python3 \
-    python3-pip \
-    && make deps install-tesseract install-tesserocr install \
-    && rm -rf /build-ocrd_tesserocr \
+RUN make deps-ubuntu deps install-tesseract install-tesserocr install \
+    && rm -rf /build \
     && apt-get -y remove --auto-remove g++ libtesseract-dev make
 
 # PPA tessdata prefix (= ocrd_tesserocr moduledir) is owned by root
@@ -83,6 +52,8 @@ RUN apt-get install -y --no-install-recommends \
 RUN ocrd resmgr download ocrd-tesserocr-recognize Fraktur.traineddata
 RUN ocrd resmgr download ocrd-tesserocr-recognize deu.traineddata
 RUN ocrd resmgr download ocrd-tesserocr-recognize eng.traineddata
+RUN ocrd resmgr download ocrd-tesserocr-recognize equ.traineddata
+RUN ocrd resmgr download ocrd-tesserocr-recognize osd.traineddata
 
 WORKDIR /data
 VOLUME /data
diff --git a/Makefile b/Makefile
index 40f3295..8a22a0c 100644
--- a/Makefile
+++ b/Makefile
@@ -62,24 +62,32 @@ help:
 
 # Dependencies for deployment in an ubuntu/debian linux
 # (lib*-dev merely for building tesserocr with pip)
-# (tesseract-ocr: Ubuntu 18.04 now ships 4.0.0,
-#  which is unsupported. Add the tesseract-ocr PPA
-#  from Alexander Pozdnyakov which provides 4.1.0.
-#  See https://launchpad.net/~alex-p/+archive/ubuntu/tesseract-ocr
-#  for details.)
 deps-ubuntu:
-	apt-get install -y --no-install-recommends software-properties-common
-	-add-apt-repository -u -y ppa:alex-p/tesseract-ocr
-	apt-get install -y \
+	apt-get update && apt-get install -y --no-install-recommends \
+		apt-utils \
+		build-essential \
 		g++ \
 		git \
 		python3 \
 		python3-pip \
-		libtesseract-dev \
+		libjpeg-dev \
+		libgif-dev \
+		libwebp-dev \
+		libopenjp2-7-dev \
+		libpng-dev \
+		libtiff-dev \
+		libtool \
+		pkg-config \
+		tzdata \
+		xzgv \
+		zlib1g-dev \
 		libleptonica-dev \
-		tesseract-ocr-eng \
-		tesseract-ocr-script-frak \
-		tesseract-ocr
+		libpango1.0-dev \
+		libicu-dev \
+		autotools-dev \
+		automake \
+		libcurl4-nss-dev \
+		libarchive-dev
 
 # Install Python deps for install via pip
 deps:

From d194816f51f97d7498439e6bf344b64afc60545b Mon Sep 17 00:00:00 2001
From: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
Date: Wed, 14 Feb 2024 10:32:46 +0100
Subject: [PATCH 05/13] CI: update (make install-tesseract)

make deps-ubuntu no longer fetches Tesseract via PPA, so we need to make install-tesseract

also, drop unsupported Python 3.6
---
 .circleci/config.yml | 13 ++++---------
 1 file changed, 4 insertions(+), 9 deletions(-)

diff --git a/.circleci/config.yml b/.circleci/config.yml
index 5afab93..89bf3fa 100644
--- a/.circleci/config.yml
+++ b/.circleci/config.yml
@@ -15,15 +15,10 @@ jobs:
     steps:
       - checkout
       - run: sudo make deps-ubuntu
-      - when:
-          condition:
-            equal: [ '3.6', << parameters.python-version >> ]
-          steps:
-            # speed-up build time for end-of-life Python by holding at latest binary:
-            - run: pip install --prefer-binary -U opencv-python-headless numpy
+      - run: make install-tesseract install-tesserocr
       - run: make install
-      # PPA tessdata prefix (= ocrd_tesserocr moduledir) is owned by root
-      - run: sudo chmod go+w `dpkg-query -L tesseract-ocr-eng | sed -n s,/eng.traineddata,,p`
+      - run: ocrd resmgr download ocrd-tesserocr-recognize eng.traineddata
+      - run: ocrd resmgr download ocrd-tesserocr-recognize osd.traineddata
       - run: ocrd resmgr download ocrd-tesserocr-recognize deu.traineddata
       - run: ocrd resmgr download ocrd-tesserocr-recognize Fraktur.traineddata
       - run: make test-cli
@@ -36,4 +31,4 @@ workflows:
       - build-python:
           matrix:
             parameters:
-              python-version: ['3.6', '3.7', '3.8', '3.9', '3.10']
+              python-version: ['3.7', '3.8', '3.9', '3.10']

From c9f2f748eecbfd7a0f26f0a54d3f3cc9f5b50082 Mon Sep 17 00:00:00 2001
From: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
Date: Wed, 14 Feb 2024 10:41:33 +0100
Subject: [PATCH 06/13] makefile: fix subrepo dependencies

---
 Makefile | 20 ++++++++++++--------
 1 file changed, 12 insertions(+), 8 deletions(-)

diff --git a/Makefile b/Makefile
index 8a22a0c..75745c1 100644
--- a/Makefile
+++ b/Makefile
@@ -105,20 +105,24 @@ docker:
 	--build-arg BUILD_DATE=$$(date -u +"%Y-%m-%dT%H:%M:%SZ") \
 	-t $(DOCKER_TAG) .
 
-install-tesserocr:
-	cd repo/tesserocr; $(PIP) install .
+install-tesserocr: repo/tesserocr
+	$(PIP) install ./$<
 
-install-tesseract:
-	cd repo/tesseract; ./autogen.sh
-	mkdir -p $(CURDIR)/build_tesseract
-	cd $(CURDIR)/build_tesseract && $(CURDIR)/repo/tesseract/configure \
+install-tesseract: repo/tesseract
+	cd $<; ./autogen.sh
+	mkdir -p build_tesseract
+	cd build_tesseract && $(CURDIR)/repo/tesseract/configure \
 				--prefix=$(TESSERACT_PREFIX) \
 				--disable-openmp \
 				--disable-shared \
-				'CXXFLAGS=-g -O2 -fno-math-errno -Wall -Wextra -Wpedantic -fPIC' ;\
-	cd $(CURDIR)/build_tesseract && $(MAKE) install
+				'CXXFLAGS=-g -O2 -fno-math-errno -Wall -Wextra -Wpedantic -fPIC' && \
+				$(MAKE) install
 	if [[ "$(TESSERACT_PREFIX)" = "/usr"* ]];then ldconfig ;fi
 
+repo/tesserocr repo/tesseract:
+	git submodule sync $@
+	git submodule update --init $@
+
 # Install this package
 install: deps
 	$(PIP) install .

From 3bdfeb73f51afa51f298b9659c859c8394a051a1 Mon Sep 17 00:00:00 2001
From: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
Date: Wed, 14 Feb 2024 10:55:14 +0100
Subject: [PATCH 07/13] CI: checkout with submodules

(since normal Circleci `checkout` creates empty submodule directories)
---
 .circleci/config.yml | 1 +
 1 file changed, 1 insertion(+)

diff --git a/.circleci/config.yml b/.circleci/config.yml
index 89bf3fa..71bed66 100644
--- a/.circleci/config.yml
+++ b/.circleci/config.yml
@@ -14,6 +14,7 @@ jobs:
       - image: cimg/python:<< parameters.python-version >>
     steps:
       - checkout
+      - run: git submodule sync && git submodule update --init
       - run: sudo make deps-ubuntu
       - run: make install-tesseract install-tesserocr
       - run: make install

From 36486bb8f0fd18ca992066531aa3c717905b78a1 Mon Sep 17 00:00:00 2001
From: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
Date: Wed, 14 Feb 2024 11:10:10 +0100
Subject: [PATCH 08/13] CI: install-tesseract needs sudo

---
 .circleci/config.yml | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/.circleci/config.yml b/.circleci/config.yml
index 71bed66..7822c9f 100644
--- a/.circleci/config.yml
+++ b/.circleci/config.yml
@@ -16,7 +16,8 @@ jobs:
       - checkout
       - run: git submodule sync && git submodule update --init
       - run: sudo make deps-ubuntu
-      - run: make install-tesseract install-tesserocr
+      - run: sudo make install-tesseract
+      - run: make install-tesserocr
       - run: make install
       - run: ocrd resmgr download ocrd-tesserocr-recognize eng.traineddata
       - run: ocrd resmgr download ocrd-tesserocr-recognize osd.traineddata

From 55ade98e80304cae55a66914ee0a9475c4f8a316 Mon Sep 17 00:00:00 2001
From: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
Date: Wed, 14 Feb 2024 11:33:43 +0100
Subject: [PATCH 09/13] CI: try unprivileged tessdata

using VIRTUAL_ENV from PYENV_ROOT
---
 .circleci/config.yml | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/.circleci/config.yml b/.circleci/config.yml
index 7822c9f..825e5d3 100644
--- a/.circleci/config.yml
+++ b/.circleci/config.yml
@@ -12,11 +12,14 @@ jobs:
         type: string
     docker:
       - image: cimg/python:<< parameters.python-version >>
+    environment:
+      # cimg/python uses pyenv instead of venv
+      VIRTUAL_ENV: $PYENV_ROOT
     steps:
       - checkout
       - run: git submodule sync && git submodule update --init
       - run: sudo make deps-ubuntu
-      - run: sudo make install-tesseract
+      - run: make install-tesseract
       - run: make install-tesserocr
       - run: make install
       - run: ocrd resmgr download ocrd-tesserocr-recognize eng.traineddata

From 09686cd94683733ab6054ebbbacc129631b9f8e3 Mon Sep 17 00:00:00 2001
From: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
Date: Wed, 14 Feb 2024 11:38:52 +0100
Subject: [PATCH 10/13] CI: fix envvar syntax

---
 .circleci/config.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.circleci/config.yml b/.circleci/config.yml
index 825e5d3..be6f770 100644
--- a/.circleci/config.yml
+++ b/.circleci/config.yml
@@ -14,7 +14,7 @@ jobs:
       - image: cimg/python:<< parameters.python-version >>
     environment:
       # cimg/python uses pyenv instead of venv
-      VIRTUAL_ENV: $PYENV_ROOT
+      VIRTUAL_ENV: ${PYENV_ROOT}
     steps:
       - checkout
       - run: git submodule sync && git submodule update --init

From dbfbea706f803b95327a0c266272ae075f07bd14 Mon Sep 17 00:00:00 2001
From: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
Date: Wed, 14 Feb 2024 12:43:27 +0100
Subject: [PATCH 11/13] install-tesseract: make rules reentrant, add training
 utils

---
 Makefile | 72 ++++++++++++++++++++++++++++++--------------------------
 1 file changed, 39 insertions(+), 33 deletions(-)

diff --git a/Makefile b/Makefile
index 75745c1..3589e85 100644
--- a/Makefile
+++ b/Makefile
@@ -28,40 +28,35 @@ PYTEST_ARGS =
 # Docker container tag
 DOCKER_TAG = 'ocrd/tesserocr'
 
-# BEGIN-EVAL makefile-parser --make-help Makefile
-
 help:
 	@echo ""
 	@echo "  Targets"
 	@echo ""
-	@echo "    deps-ubuntu   Dependencies for deployment in an ubuntu/debian linux"
-	@echo "                  (lib*-dev merely for building tesserocr with pip)"
-	@echo "                  (tesseract-ocr: Ubuntu 18.04 now ships 4.0.0,"
-	@echo "                   which is unsupported. Add the tesseract-ocr PPA"
-	@echo "                   from Alexander Pozdnyakov which provides 4.1.0."
-	@echo "                   See https://launchpad.net/~alex-p/+archive/ubuntu/tesseract-ocr"
-	@echo "                   for details.)"
-	@echo "    deps          Install Python deps for install via pip"
-	@echo "    deps-test     Install Python deps for test via pip"
-	@echo "    docker        Build docker image"
-	@echo "    install       Install this package"
-	@echo "    test          Run unit tests"
-	@echo "    coverage      Run unit tests and determine test coverage"
-	@echo "    test-cli      Test the command line tools"
-	@echo "    test/assets   Setup test assets"
-	@echo "    repo/assets   Clone OCR-D/assets to ./repo/assets"
-	@echo "    assets-clean  Remove symlinks in test/assets"
+	@echo "    deps-ubuntu       Install system dependencies in an Ubuntu/Debian Linux"
+	@echo "    install-tesseract Compile and install Tesseract"
+	@echo "    install-tesseract-training Compile and install training utilities for Tesseract"
+	@echo "    install-tesserocr Compile and install Tesserocr"
+	@echo "    deps              Install Python dependencies for install via pip"
+	@echo "    install           Install this package via pip"
+	@echo "    deps-test         Install Python deps for test via pip"
+	@echo "    test              Run unit tests"
+	@echo "    coverage          Run unit tests and determine test coverage"
+	@echo "    test-cli          Test the command line tools"
+	@echo "    test/assets       Setup test assets"
+	@echo "    repo/assets       Clone OCR-D/assets to ./repo/assets"
+	@echo "    repo/tesseract    Checkout Tesseract ./repo/tesseract"
+	@echo "    repo/tesserocr    Checkout Tesserocr to ./repo/tesserocr"
+	@echo "    docker            Build docker image"
+	@echo "    assets-clean      Remove symlinks in test/assets"
 	@echo ""
 	@echo "  Variables"
 	@echo ""
-	@echo "    PYTEST_ARGS     pytest args. Set to '-s' to see log output during test execution, '--verbose' to see individual tests. Default: '$(PYTEST_ARGS)'"
-	@echo "    DOCKER_TAG      Docker container tag"
-	@echo "    TESSDATA_PREFIX search path for recognition models (overriding Tesseract compile-time default)"
-
-# END-EVAL
+	@echo "    PYTEST_ARGS     pytest args. Set to '-s' to see log output during test execution, '--verbose' to see individual tests. [$(PYTEST_ARGS)]"
+	@echo "    DOCKER_TAG      Docker container tag [$(DOCKER_TAG)]"
+	@echo "    TESSDATA_PREFIX search path for recognition models (overriding Tesseract compile-time default) [$(TESSDATA_PREFIX)]"
 
-# Dependencies for deployment in an ubuntu/debian linux
-# (lib*-dev merely for building tesserocr with pip)
+# Dependencies for deployment in an Ubuntu/Debian Linux
+# (lib*-dev merely for building Tesseract and tesserocr from sources)
 deps-ubuntu:
 	apt-get update && apt-get install -y --no-install-recommends \
 		apt-utils \
@@ -108,16 +103,27 @@ docker:
 install-tesserocr: repo/tesserocr
 	$(PIP) install ./$<
 
-install-tesseract: repo/tesseract
-	cd $<; ./autogen.sh
-	mkdir -p build_tesseract
-	cd build_tesseract && $(CURDIR)/repo/tesseract/configure \
+install-tesseract: $(TESSERACT_PREFIX)/bin/tesseract
+
+install-tesseract-training: $(TESSERACT_PREFIX)/bin/lstmtraining
+
+$(TESSERACT_PREFIX)/bin/tesseract: build_tesseract/Makefile
+	$(MAKE) -C build_tesseract install
+	if [[ "$(TESSERACT_PREFIX)" = "/usr"* ]]; then ldconfig; fi
+
+$(TESSERACT_PREFIX)/bin/lstmtraining: build_tesseract/Makefile
+	$(MAKE) -C build_tesseract training-install
+
+build_tesseract/Makefile: repo/tesseract/Makefile.in
+	mkdir -p $(@D)
+	cd $(@D) && $(CURDIR)/repo/tesseract/configure \
 				--prefix=$(TESSERACT_PREFIX) \
 				--disable-openmp \
 				--disable-shared \
-				'CXXFLAGS=-g -O2 -fno-math-errno -Wall -Wextra -Wpedantic -fPIC' && \
-				$(MAKE) install
-	if [[ "$(TESSERACT_PREFIX)" = "/usr"* ]];then ldconfig ;fi
+				'CXXFLAGS=-g -O2 -fno-math-errno -Wall -Wextra -Wpedantic -fPIC'
+
+repo/tesseract/Makefile.in: repo/tesseract
+	cd $<; ./autogen.sh
 
 repo/tesserocr repo/tesseract:
 	git submodule sync $@

From 21f2ad370f4ce1bf09906d9770d4ff80c9ddd015 Mon Sep 17 00:00:00 2001
From: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
Date: Wed, 14 Feb 2024 12:59:35 +0100
Subject: [PATCH 12/13] docker: make tesseract-training, too

and alias tessdata to /models for easier persistence
---
 Dockerfile | 23 ++++++++++++++++-------
 1 file changed, 16 insertions(+), 7 deletions(-)

diff --git a/Dockerfile b/Dockerfile
index 971c556..f1f0b31 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -13,9 +13,6 @@ LABEL \
 
 ENV PYTHONIOENCODING utf8
 
-# set TESSDATA_PREFIX
-ENV TESSDATA_PREFIX /usr/local/share/tessdata
-
 # set frontend non-interactive to silence interactive tzdata config
 ARG DEBIAN_FRONTEND=noninteractive
 
@@ -31,6 +28,8 @@ RUN dpkg-reconfigure -f noninteractive tzdata
 # so let XDG_DATA_HOME coincide with fixed system location
 # (can still be overridden by derived stages)
 ENV XDG_DATA_HOME /usr/local/share
+ENV XDG_CONFIG_HOME /usr/local/share/ocrd-resources
+ENV TESSDATA_PREFIX $XDG_DATA_HOME/tessdata
 
 WORKDIR /build
 COPY setup.py .
@@ -42,18 +41,28 @@ COPY ocrd_tesserocr ./ocrd_tesserocr
 COPY repo/tesserocr ./repo/tesserocr
 COPY repo/tesseract ./repo/tesseract
 COPY Makefile .
-RUN make deps-ubuntu deps install-tesseract install-tesserocr install \
+RUN make deps-ubuntu deps install-tesseract install-tesseract-training install-tesserocr install \
     && rm -rf /build \
     && apt-get -y remove --auto-remove g++ libtesseract-dev make
 
-# PPA tessdata prefix (= ocrd_tesserocr moduledir) is owned by root
-# next line causes failure because tesseract-ocr-eng not existing. Not sure if needed, so skipping
-# RUN sudo chmod go+w `dpkg-query -L tesseract-ocr-eng | sed -n s,/eng.traineddata,,p`
 RUN ocrd resmgr download ocrd-tesserocr-recognize Fraktur.traineddata
 RUN ocrd resmgr download ocrd-tesserocr-recognize deu.traineddata
 RUN ocrd resmgr download ocrd-tesserocr-recognize eng.traineddata
 RUN ocrd resmgr download ocrd-tesserocr-recognize equ.traineddata
 RUN ocrd resmgr download ocrd-tesserocr-recognize osd.traineddata
 
+# as discussed in ocrd_all#378, we do not want to manage more than one resource location
+# to mount for model persistence; 
+# with named volumes, the preinstalled models will be copied to the host and complemented
+# by downloaded models; 
+# tessdata is the only problematic module location
+RUN mkdir -p $XDG_CONFIG_HOME
+RUN mv $TESSDATA_PREFIX $XDG_CONFIG_HOME/ocrd-tesserocr-recognize
+RUN ln -s $XDG_CONFIG_HOME/ocrd-tesserocr-recognize $TESSDATA_PREFIX
+# finally, alias/symlink all ocrd-resources to /models for shorter mount commands
+RUN mv $XDG_CONFIG_HOME /models && ln -s /models $XDG_CONFIG_HOME
+
+
+# finally, alias/symlink all ocrd-resources to /models for shorter mount commands
 WORKDIR /data
 VOLUME /data

From dab081e1c27067e50787128cbd302c35b6693760 Mon Sep 17 00:00:00 2001
From: Robert Sachunsky <38561704+bertsky@users.noreply.github.com>
Date: Wed, 14 Feb 2024 13:20:11 +0100
Subject: [PATCH 13/13] CI: add CD via Dockerhub

---
 .circleci/config.yml | 26 ++++++++++++++++++++++++--
 1 file changed, 24 insertions(+), 2 deletions(-)

diff --git a/.circleci/config.yml b/.circleci/config.yml
index be6f770..1b5dcc6 100644
--- a/.circleci/config.yml
+++ b/.circleci/config.yml
@@ -6,7 +6,7 @@ orbs:
 
 jobs:
 
-  build-python:
+  test-python:
     parameters:
       python-version:
         type: string
@@ -30,10 +30,32 @@ jobs:
       - run: make coverage
       - codecov/upload
 
+  deploy-docker:
+    docker:
+      - image: circleci/buildpack-deps:stretch
+    environment:
+      DOCKER_TAG: ocrd/tesserocr
+    steps:
+      - checkout
+      - setup_remote_docker: # https://circleci.com/docs/2.0/building-docker-images/
+          docker_layer_caching: true
+      - run: make docker DOCKER_TAG=$DOCKER_TAG
+      - run:
+          name: Login to Docker Hub
+          command: echo "$DOCKERHUB_PASS" | docker login --username "$DOCKERHUB_USER" --password-stdin
+      - run: docker push $DOCKER_TAG
+
+
 workflows:
   build:
     jobs:
-      - build-python:
+      - test-python:
           matrix:
             parameters:
               python-version: ['3.7', '3.8', '3.9', '3.10']
+  deploy:
+    jobs:
+      - deploy-docker:
+          filters:
+            branches:
+              only: master