From e0d925069578507dc951dcd48c9e4626397197b0 Mon Sep 17 00:00:00 2001 From: Philipp Zumstein Date: Wed, 1 Jan 2020 21:50:05 +0100 Subject: [PATCH] :arrow_up: Upgrade to new version of hOCR-to-ALTO The transformations alto2.0__hocr and alto2.1__hocr are temporarily kept in order to make it backwards compatible. --- Makefile | 15 +++++++++------ example/Makefile | 2 +- vendor/Makefile | 2 +- 3 files changed, 11 insertions(+), 8 deletions(-) diff --git a/Makefile b/Makefile index 00298a2..f6e14f3 100644 --- a/Makefile +++ b/Makefile @@ -74,12 +74,15 @@ xsd: vendor xslt: vendor $(MKDIR) xslt # symlink hocr<->alto as well as the language codes lookup xml - cd xslt && $(LN) ../vendor/hOCR-to-ALTO/hocr2alto2.0.xsl hocr__alto2.0.xsl - cd xslt && $(LN) ../vendor/hOCR-to-ALTO/hocr2alto2.1.xsl hocr__alto2.1.xsl - cd xslt && $(LN) ../vendor/hOCR-to-ALTO/alto2hocr.xsl alto2.0__hocr.xsl - cd xslt && $(LN) ../vendor/hOCR-to-ALTO/alto2hocr.xsl alto2.1__hocr.xsl - cd xslt && $(LN) ../vendor/hOCR-to-ALTO/hocr2text.xsl hocr__text.xsl - cd xslt && $(LN) ../vendor/hOCR-to-ALTO/alto2text.xsl alto__text.xsl + cd xslt && $(LN) ../vendor/hOCR-to-ALTO/hocr__alto2.0.xsl hocr__alto2.0.xsl + cd xslt && $(LN) ../vendor/hOCR-to-ALTO/hocr__alto2.1.xsl hocr__alto2.1.xsl + cd xslt && $(LN) ../vendor/hOCR-to-ALTO/hocr__alto3.xsl hocr__alto3.0.xsl + cd xslt && $(LN) ../vendor/hOCR-to-ALTO/hocr__alto4.xsl hocr__alto4.0.xsl + cd xslt && $(LN) ../vendor/hOCR-to-ALTO/alto__hocr.xsl alto__hocr.xsl + cd xslt && $(LN) alto__hocr.xsl alto2.0__hocr.xsl + cd xslt && $(LN) alto__hocr.xsl alto2.1__hocr.xsl + cd xslt && $(LN) ../vendor/hOCR-to-ALTO/hocr__text.xsl hocr__text.xsl + cd xslt && $(LN) ../vendor/hOCR-to-ALTO/alto__text.xsl alto__text.xsl cd xslt && $(LN) ../vendor/hOCR-to-ALTO/codes_lookup.xml codes_lookup.xml cd xslt && $(LN) ../vendor/format-converters/page2hocr.xsl page__hocr.xsl cd xslt && $(LN) ../vendor/format-converters/abbyy2hocr.xsl abbyy__hocr.xsl diff --git a/example/Makefile b/example/Makefile index 7a8fef1..2a2c194 100644 --- a/example/Makefile +++ b/example/Makefile @@ -45,7 +45,7 @@ $(BASENAME).alto.page.alto : $(BASENAME).alto.page $(OCR_TRANSFORM) page alto $< | $(XMLLINT) - > $@ $(BASENAME).roundtrip.hocr : $(BASENAME).alto - $(OCR_TRANSFORM) alto2.0 hocr $< | $(XMLLINT) - > $@ + $(OCR_TRANSFORM) alto hocr $< | $(XMLLINT) - > $@ clean: $(RM) $(BASENAME)*.hocr $(BASENAME)*.alto diff --git a/vendor/Makefile b/vendor/Makefile index 8fb9f7b..bfb9557 100644 --- a/vendor/Makefile +++ b/vendor/Makefile @@ -28,7 +28,7 @@ ABBYY_SCHEMA_VERSIONS = 6-schema-v1 8-schema-v2 9-schema-v1 10-schema-v1 HOCR2ALTO_REPO = hOCR-to-ALTO HOCR2ALTO_URL = https://github.com/filak/$(HOCR2ALTO_REPO) -HOCR2ALTO_COMMITID = 43f9d9dcdd38d37071f8ef1998d377d21c9cadd2 +HOCR2ALTO_COMMITID = 7b21c4727a099117ed00c4f335aa83d0392c919f HOCR_SPEC_REPO = hocr-spec-python HOCR_SPEC_URL = https://github.com/kba/$(HOCR_SPEC_REPO)