ocrd_page: PcGts from image

.pylintrc

@@ -5,6 +5,7 @@ ignored-modules=cv2,tesserocr,ocrd.model
 [MESSAGES CONTROL]
 ignore-patterns='.*generateds.*'
 disable =
+    inconsistent-return-statements,
     ungrouped-imports,
     missing-docstring,
     no-self-use,

ocrd/model/ocrd_file.py

@@ -1,5 +1,5 @@
 import os
-from ocrd.constants import NAMESPACES as NS, TAG_METS_FLOCAT
+from ocrd.constants import NAMESPACES as NS, TAG_METS_FLOCAT, TAG_METS_FILE
 
 from .ocrd_xml_base import ET
 
@@ -12,9 +12,13 @@ class OcrdFile(object):
     #  def create(mimetype, ID, url, local_filename):
     #      el_fileGrp.SubElement('file')
 
-    def __init__(self, el, instance=None, local_filename=None, workspace=None):
+    def __init__(self, el, mimetype=None, instance=None, local_filename=None, workspace=None):
+        if el is None:
+            el = ET.Element(TAG_METS_FILE)
         self._el = el
+        self.mimetype = mimetype
         self.local_filename = local_filename
+
         self._instance = instance
         self.workspace = workspace
 

ocrd/model/ocrd_page.py

@@ -20,18 +20,24 @@ def to_xml(el):
     el.export(sio, 0, name_='PcGts', namespacedef_='xmlns:pc="%s"' % NAMESPACES['page'])
     return '<?xml version="1.0" encoding="UTF-8"?>\n' + sio.getvalue()
 
+def page_from_image(input_file):
+    if input_file.local_filename is None:
+        raise Exception("input_file must have 'local_filename' property")
+    exif = OcrdExif.from_filename(input_file.local_filename)
+    content = PAGE_XML_EMPTY.replace('<Page>', '<Page imageWidth="%d" imageHeight="%i" imageFilename="%s">' % (
+        exif.width,
+        exif.height,
+        input_file.url
+    ))
+    return content
+
 def from_file(input_file):
     """
     Create a new PAGE-XML from a METS file representing a PAGE-XML or an image.
     """
     #  print("PARSING PARSING '%s'" % input_file)
     if input_file.mimetype.startswith('image'):
-        exif = OcrdExif.from_filename(input_file.local_filename)
-        content = PAGE_XML_EMPTY.replace('<Page>', '<Page imageWidth="%d" imageHeight="%i" imageFilename="%s">' % (
-            exif.width,
-            exif.height,
-            input_file.url
-        ))
+        content = page_from_image(input_file)
         return parseString(content.encode('utf-8'), silence=True)
     elif input_file.mimetype == 'text/page+xml':
         return parse(input_file.local_filename, silence=True)

test/model/test_ocrd_mets.py

@@ -1,12 +1,11 @@
-from ocrd.model import OcrdMets
-
 from test.base import TestCase, main, assets
-METS_HEROLD = assets.url_of('SBB0000F29300010000/mets.xml')
+
+from ocrd.model import OcrdMets
 
 class TestOcrdMets(TestCase):
 
     def setUp(self):
-        self.mets = OcrdMets(filename=METS_HEROLD)
+        self.mets = OcrdMets(filename=assets.url_of('SBB0000F29300010000/mets.xml'))
 
     def test_unique_identifier(self):
         self.assertEqual(self.mets.unique_identifier, 'http://resolver.staatsbibliothek-berlin.de/SBB0000F29300010000', 'Right identifier')

test/model/test_ocrd_page.py

@@ -1,18 +1,26 @@
-
-import sys
-
 from test.base import TestCase, main, assets
+import ocrd.model.ocrd_file as ocrd_file
 import ocrd.model.ocrd_page as ocrd_page
 
 # pylint: disable=protected-access
 
 class TestOcrdPage(TestCase):
 
     def setUp(self):
-        with open(assets.url_of('page-with-glyphs.xml').replace('file://', ''), 'rb') as f:
+        with open(assets.url_of('page-with-glyphs.xml', remove_file=True), 'rb') as f:
             self.xml_as_str = f.read()
             self.pcgts = ocrd_page.parseString(self.xml_as_str, silence=True)
 
+    def test_from_file(self):
+        f = ocrd_file.OcrdFile(
+            None,
+            mimetype='image/tif',
+            local_filename=assets.url_of('kant_aufklaerung_1784/kant_aufklaerung_1784_0017.tif', remove_file=True)
+        )
+        self.assertEqual(f.mimetype, 'image/tif')
+        p = ocrd_page.from_file(f)
+        print(ocrd_page.to_xml(p))
+
     def test_pcGtsId(self):
         self.assertEqual(self.pcgts.pcGtsId, 'glyph-test')