Merge remote-tracking branch 'origin/adapt-tessdata' into update-2022…

…-10-25

.github/workflows/makeall-linux.yaml

@@ -34,8 +34,8 @@ jobs:
       PYTHON_VERSION: ${{ matrix.python-version }}
 
     steps:
-    - uses: actions/checkout@v2
-    - uses: actions/setup-python@v2
+    - uses: actions/checkout@v3
+    - uses: actions/setup-python@v4
       with:
         python-version: ${{ env.PYTHON_VERSION }}
     #    architecture: x64

.github/workflows/makedocker.yml

@@ -57,8 +57,8 @@ jobs:
       PYTHON_VERSION: ${{ github.event.inputs.python-version }}
 
     steps:
-    - uses: actions/checkout@v2
-    - uses: actions/setup-python@v2
+    - uses: actions/checkout@v3
+    - uses: actions/setup-python@v4
       with:
         python-version: ${{ env.PYTHON_VERSION }}
     - name: Show Python3 version

Makefile

@@ -139,7 +139,6 @@ Variables:
 	PIP_OPTIONS: extra options for the `pip install` command like `-q` or `-v` or `-e`
 	TESSERACT_MODELS: list of additional models/languages to download for Tesseract. Default: "$(ALL_TESSERACT_MODELS)"
 	TESSERACT_CONFIG: command line options for Tesseract `configure`. Default: "$(TESSERACT_CONFIG)"
-	TESSDATA: directory path where to install Tesseract models. Default (based on XDG_DATA_HOME): "$(TESSDATA)"
 EOF
 endef
 export HELP
@@ -253,6 +252,8 @@ OCRD_EXECUTABLES += $(OCRD_COR_ASV_ANN)
 OCRD_COR_ASV_ANN := $(BIN)/ocrd-cor-asv-ann-evaluate
 OCRD_COR_ASV_ANN += $(BIN)/ocrd-cor-asv-ann-process
 OCRD_COR_ASV_ANN += $(BIN)/ocrd-cor-asv-ann-align
+OCRD_COR_ASV_ANN += $(BIN)/ocrd-cor-asv-ann-join
+OCRD_COR_ASV_ANN += $(BIN)/ocrd-cor-asv-ann-mark
 OCRD_COR_ASV_ANN += $(BIN)/cor-asv-ann-train
 OCRD_COR_ASV_ANN += $(BIN)/cor-asv-ann-proc
 OCRD_COR_ASV_ANN += $(BIN)/cor-asv-ann-eval
@@ -411,6 +412,7 @@ OCRD_TESSEROCR := $(BIN)/ocrd-tesserocr-binarize
 OCRD_TESSEROCR += $(BIN)/ocrd-tesserocr-crop
 OCRD_TESSEROCR += $(BIN)/ocrd-tesserocr-deskew
 OCRD_TESSEROCR += $(BIN)/ocrd-tesserocr-recognize
+OCRD_TESSEROCR += $(BIN)/ocrd-tesserocr-segment
 OCRD_TESSEROCR += $(BIN)/ocrd-tesserocr-segment-line
 OCRD_TESSEROCR += $(BIN)/ocrd-tesserocr-segment-region
 OCRD_TESSEROCR += $(BIN)/ocrd-tesserocr-segment-word
@@ -462,7 +464,7 @@ install-models-calamari: $(BIN)/ocrd
 	. $(ACTIVATE_VENV) && ocrd resmgr download ocrd-calamari-recognize '*'
 OCRD_EXECUTABLES += $(OCRD_CALAMARI)
 OCRD_CALAMARI := $(BIN)/ocrd-calamari-recognize
-$(OCRD_CALAMARI): ocrd_calamari
+$(OCRD_CALAMARI): ocrd_calamari $(BIN)/ocrd
 	$(pip_install)
 endif
 
@@ -490,7 +492,7 @@ OCRD_ANYBASEOCR += $(BIN)/ocrd-anybaseocr-dewarp
 OCRD_ANYBASEOCR += $(BIN)/ocrd-anybaseocr-tiseg
 OCRD_ANYBASEOCR += $(BIN)/ocrd-anybaseocr-textline
 OCRD_ANYBASEOCR += $(BIN)/ocrd-anybaseocr-layout-analysis
-$(call multirule,$(OCRD_ANYBASEOCR)): ocrd_anybaseocr
+$(call multirule,$(OCRD_ANYBASEOCR)): ocrd_anybaseocr $(BIN)/ocrd
 	$(pip_install)
 endif
 
@@ -517,7 +519,8 @@ install-models-sbb-binarization:
 
 OCRD_EXECUTABLES += $(SBB_BINARIZATION)
 SBB_BINARIZATION := $(BIN)/ocrd-sbb-binarize
-$(SBB_BINARIZATION): sbb_binarization
+SBB_BINARIZATION += $(BIN)/sbb_binarize
+$(call multirule,$(SBB_BINARIZATION)): sbb_binarization $(BIN)/ocrd
 	$(pip_install)
 endif
 
@@ -528,7 +531,7 @@ install-models-eynollah:
 	. $(ACTIVATE_VENV) && ocrd resmgr download ocrd-eynollah-segment '*'
 OCRD_EXECUTABLES += $(EYNOLLAH_SEGMENT)
 EYNOLLAH_SEGMENT := $(BIN)/ocrd-eynollah-segment
-$(EYNOLLAH_SEGMENT): eynollah
+$(EYNOLLAH_SEGMENT): eynollah $(BIN)/ocrd
 	$(pip_install)
 endif
 
@@ -689,11 +692,10 @@ CUSTOM_DEPS += libpango1.0-dev
 
 XDG_DATA_HOME ?= $(if $(HOME),$(HOME)/.local/share,/usr/local/share)
 DEFAULT_RESLOC ?= $(XDG_DATA_HOME)/ocrd-resources
-TESSDATA ?= $(DEFAULT_RESLOC)/ocrd-tesserocr-recognize
+TESSDATA = $(VIRTUAL_ENV)/share/tessdata/
 TESSDATA_RELEASE = 4.1.0
 TESSDATA_URL := https://github.com/tesseract-ocr/tessdata_fast/raw/$(TESSDATA_RELEASE)
 TESSERACT_TRAINEDDATA = $(ALL_TESSERACT_MODELS:%=$(TESSDATA)/%.traineddata)
-TESSERACT_TRAINEDDATA += $(ALL_TESSERACT_MODELS:%=$(VIRTUAL_ENV)/share/tessdata/%.traineddata)
 
 stripdir = $(patsubst %/,%,$(dir $(1)))
 
@@ -715,10 +717,6 @@ $(TESSDATA)/%.traineddata:
 	$(call WGET,$@,$(TESSDATA_URL)/$(notdir $(call stripdir,$@))/$(notdir $@)) || \
 		{ $(RM) $@; false; }
 
-$(VIRTUAL_ENV)/share/tessdata/%.traineddata: $(TESSDATA)/%.traineddata
-	@mkdir -p $(dir $@)
-	cp $< $@
-
 tesseract/Makefile.in: tesseract
 	cd tesseract && ./autogen.sh
 

README.md

@@ -333,14 +333,16 @@ This table lists which tag contains which module:
 | format-converters           | -         | ☑        | ☑         |
 | ocrd_calamari               | -         | ☑        | ☑         |
 | ocrd_keraslm                | -         | ☑        | ☑         |
-| ocrd_olahd_client           | -         | ☑        | ☑         |
+| ocrd_olahd_client           | ☑         | ☑        | ☑         |
 | ocrd_olena                  | -         | ☑        | ☑         |
 | ocrd_segment                | -         | ☑        | ☑         |
 | tesseract                   | -         | ☑        | ☑         |
 | ocrd_anybaseocr             | -         | -        | ☑         |
-| ocrd_kraken                 | -         | -        | -         |
+| ocrd_detectron2             | -         | -        | ☑         |
+| ocrd_doxa                   | -         | -        | ☑         |
+| ocrd_kraken                 | -         | -        | ☑         |
 | ocrd_ocropy                 | -         | -        | -         |
-| ocrd_pc_segmentation        | -         | -        | ☑         |
+| ocrd_pc_segmentation        | -         | -        | -         |
 | ocrd_typegroups_classifier  | -         | -        | ☑         |
 | sbb_binarization            | -         | -        | ☑         |
 | cor-asv-fst                 | -         | -        | -         |
@@ -350,8 +352,6 @@ enabled by explicitly setting `OCRD_MODULES` or `DISABLED_MODULES`:
 
 * cor-asv-fst (runtime issues)
 * ocrd_ocropy (better implementation in ocrd_cis available)
-* ocrd_kraken (currently unmaintained)
-* clstm (required only for ocrd_kraken)
 
 ### Uninstall
 
@@ -373,7 +373,6 @@ This repo offers solutions to the following problems with OCR-D integration.
 
 The following Python modules need an installation from code for different reasons:
 
-- clstm (needs modified code for Python3)
 - cor-asv-ann (not available in PyPI)
 - cor-asv-fst (not available in PyPI)
 - dinglehopper (not available in PyPI)
@@ -417,7 +416,6 @@ _(Solved by managing and delegating to different subsets of venvs.)_
 
 Not all modules advertise their system package requirements via `make deps-ubuntu`.
 
-- `clstm`: depends on `scons libprotobuf-dev protobuf-compiler libpng-dev libeigen3-dev swig`
 - `tesseract` (when installing from source not PPA): depends on `libleptonica-dev` etc
 
 _(Solved by maintaining these requirements under `deps-ubuntu` here.)_

release.sh

@@ -22,7 +22,7 @@ usage () {
     echo ""
     echo "Commands:"
     echo ""
-    echo "  update              Update all submodules to most recent master/dev branch"
+    echo "  update              Update all submodules to most recent default branch"
     echo "  changelog           Generate a changelog for all modified submodules"
     echo "  release-github      Release to GitHub as $version"
     echo "  release-dockerhub   Release ocrd/all:maximum as ocrd/all:${version#v} to DockerHub"
@@ -79,7 +79,7 @@ update_one_submodule () {
         cd $sm
         local branch=$(git remote show origin | sed -n '/HEAD branch/s/.*: //p')
         loginfo "Updating submodule $sm / branch $branch"
-        git pull -q --rebase origin "$branch" 
+        git pull -q --rebase origin "$branch"
         git pull -q --rebase origin "$branch" --tags
         git submodule update --init
      )