add XSD validators for PAGE and METS, OCR-D#449

bertsky · Apr 29, 2020 · 11e2733 · 11e2733
1 parent 0cded79
commit 11e2733
Show file tree

Hide file tree

Showing 15 changed files with 4,755 additions and 5 deletions.
diff --git a/.envrc b/.envrc
@@ -0,0 +1 @@
+export PYLINTRC=$PWD/.pylintrc
diff --git a/.vimrc b/.vimrc
@@ -0,0 +1,5 @@
+set path+=ocrd
+set path+=ocrd_models
+set path+=ocrd_modelfactory
+set path+=ocrd_validators
+set path+=ocrd_utils
diff --git a/ocrd_validators/ocrd_validators/__init__.py b/ocrd_validators/ocrd_validators/__init__.py
@@ -7,7 +7,10 @@
     'PageValidator',
     'OcrdToolValidator',
     'OcrdZipValidator',
-    'ValidationReport'
+    'ValidationReport',
+    'XsdValidator',
+    'XsdMetsValidator',
+    'XsdPageValidator',
 ]
 
 from .report import ValidationReport
@@ -16,3 +19,6 @@
 from .page_validator import PageValidator
 from .ocrd_tool_validator import OcrdToolValidator
 from .ocrd_zip_validator import OcrdZipValidator
+from .xsd_validator import XsdValidator
+from .xsd_mets_validator import XsdMetsValidator
+from .xsd_page_validator import XsdPageValidator
diff --git a/ocrd_validators/ocrd_validators/bagit-profile.yml b/ocrd_validators/ocrd_validators/bagit-profile.yml
@@ -1,5 +1,5 @@
 BagIt-Profile-Info:
-  BagIt-Profile-Identifier: https://ocr-d.de/bagit-profile.json
+  BagIt-Profile-Identifier: https://ocr-d.github.io/bagit-profile.json
   BagIt-Profile-Version: '1.2.0'
   Source-Organization: OCR-D
   External-Description: BagIt profile for OCR data

diff --git a/ocrd_validators/ocrd_validators/constants.py b/ocrd_validators/ocrd_validators/constants.py
@@ -2,7 +2,7 @@
 Constants for ocrd_validators.
 """
 import yaml
-from pkg_resources import resource_string
+from pkg_resources import resource_string, resource_filename
 
 OCRD_TOOL_SCHEMA = yaml.safe_load(resource_string(__name__, 'ocrd_tool.schema.yml'))
 OCRD_BAGIT_PROFILE = yaml.safe_load(resource_string(__name__, 'bagit-profile.yml'))
@@ -12,3 +12,8 @@
 FILE_GROUP_CATEGORIES = ['IMG', 'SEG', 'OCR', 'COR', 'GT']
 TMP_BAGIT_PREFIX = 'ocrd-bagit-'
 OCRD_BAGIT_PROFILE_URL = 'https://ocr-d.github.io/bagit-profile.json'
+XSD_METS_URL = 'https://www.loc.gov/standards/mets/mets.xsd'
+XSD_PAGE_URL = 'http://www.primaresearch.org/schema/PAGE/gts/pagecontent/2019-07-15/pagecontent.xsd'
+XSD_PATHS = {}
+XSD_PATHS[XSD_METS_URL] = resource_filename(__name__, 'xsd/mets.xsd')
+XSD_PATHS[XSD_PAGE_URL] = resource_filename(__name__, 'xsd/page.xsd')
diff --git a/ocrd_validators/ocrd_validators/xsd/__init__.py b/ocrd_validators/ocrd_validators/xsd/__init__.py
diff --git a/ocrd_validators/ocrd_validators/xsd/mets.xsd b/ocrd_validators/ocrd_validators/xsd/mets.xsd