Merge branch 'port-to-v3' of https://github.com/OCR-D/ocrd_kraken into port-to-v3

# Conflicts:
#	ocrd_kraken/binarize.py

Author: kba

ocrd_kraken/binarize.py

@@ -21,8 +21,6 @@ def executable(self):
 
     def setup(self):
         self.logger = getLogger('processor.KrakenBinarize')
-        assert_file_grp_cardinality(self.input_file_grp, 1)
-        assert_file_grp_cardinality(self.output_file_grp, 1)
 
     def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[str] = None, page_id: Optional[str] = None) -> OcrdProcessResult:
         """Binarize the pages/regions/lines with Kraken.

ocrd_kraken/ocrd-tool.json

@@ -4,8 +4,8 @@
   "tools": {
     "ocrd-kraken-binarize": {
       "executable": "ocrd-kraken-binarize",
-      "input_file_grp": ["OCR-D-IMG", "OCR-D-PRE-CROP", "OCR-D-SEG-REGION", "OCR-D-SEG-LINE"],
-      "output_file_grp": ["OCR-D-PRE-BIN"],
+      "input_file_grp_cardinality": 1,
+      "output_file_grp_cardinality": 1,
       "categories": [
         "Image preprocessing"
       ],
@@ -24,8 +24,8 @@
     },
     "ocrd-kraken-segment": {
       "executable": "ocrd-kraken-segment",
-      "input_file_grp": ["OCR-D-IMG", "OCR-D-PRE-CROP", "OCR-D-PRE-BIN"],
-      "output_file_grp": ["OCR-D-SEG-REGION", "OCR-D-SEG-LINE"],
+      "input_file_grp_cardinality": 1,
+      "output_file_grp_cardinality": 1,
       "categories": [
         "Layout analysis"
       ],
@@ -128,8 +128,8 @@
     },
     "ocrd-kraken-recognize": {
       "executable": "ocrd-kraken-recognize",
-      "input_file_grp": ["OCR-D-SEG-LINE"],
-      "output_file_grp": ["OCR-D-OCR-KRAK"],
+      "input_file_grp_cardinality": 1,
+      "output_file_grp_cardinality": 1,
       "categories": ["Text recognition and optimization"],
       "steps": ["recognition/text-recognition"],
       "description": "Text recognition with Kraken",

ocrd_kraken/recognize.py

@@ -55,10 +55,8 @@ def executable(self):
 
     def setup(self):
         """
-        Assert filegrp cardinality, load model, set predict function
+        Load model, set predict function
         """
-        assert_file_grp_cardinality(self.input_file_grp, 1)
-        assert_file_grp_cardinality(self.output_file_grp, 1)
 
         self.logger = getLogger('processor.KrakenRecognize')
         import torch

ocrd_kraken/segment.py

@@ -46,8 +46,6 @@ def setup(self):
         Load models
         """
         self.logger = getLogger('processor.KrakenSegment')
-        assert_file_grp_cardinality(self.input_file_grp, 1)
-        assert_file_grp_cardinality(self.output_file_grp, 1)
         kwargs = {}
         kwargs['text_direction'] = self.parameter['text_direction']
         self.use_legacy = self.parameter['use_legacy']

tests/test_binarize.py

@@ -6,7 +6,7 @@
 
 from tests.base import assets, main
 
-from ocrd import Resolver
+from ocrd import Resolver, run_processor
 from ocrd_kraken.binarize import KrakenBinarize
 from ocrd_utils.logging import setOverrideLogLevel
 
@@ -37,24 +37,24 @@ def workspace(tmpdir):
 #      )
 
 def test_binarize_regions(workspace):
-    proc = KrakenBinarize(
-        workspace,
-        input_file_grp="OCR-D-GT-PAGE",
-        output_file_grp="OCR-D-IMG-BIN-KRAKEN",
-        parameter={'level-of-operation': 'region'}
+    run_processor(KrakenBinarize,
+                  workspace=workspace,
+                  input_file_grp="OCR-D-GT-PAGE",
+                  output_file_grp="OCR-D-IMG-BIN-KRAKEN",
+                  parameter={'level-of-operation': 'region'}
     )
-    proc.process()
     workspace.save_mets()
+    # FIXME: add result assertions (find_files, parsing PAGE etc)
 
 def test_binarize_lines(workspace):
-    proc = KrakenBinarize(
-        workspace,
-        input_file_grp="OCR-D-GT-PAGE",
-        output_file_grp="OCR-D-IMG-BIN-KRAKEN",
-        parameter={'level-of-operation': 'line'}
+    run_processor(KrakenBinarize,
+                  workspace=workspace,
+                  input_file_grp="OCR-D-GT-PAGE",
+                  output_file_grp="OCR-D-IMG-BIN-KRAKEN",
+                  parameter={'level-of-operation': 'line'}
     )
-    proc.process()
     workspace.save_mets()
+    # FIXME: add result assertions (find_files, parsing PAGE etc)
 
 if __name__ == "__main__":
     main(__file__)

tests/test_recognize.py

@@ -20,13 +20,13 @@ def test_recognize(self):
         with pushd_popd(tempdir=True) as tempdir:
             workspace = resolver.workspace_from_url(assets.path_to('communist_manifesto/data/mets.xml'), dst_dir=tempdir, download=True)
             workspace.overwrite_mode = True
-            proc = KrakenRecognize(
-                workspace,
-                input_file_grp="OCR-D-SEG-KRAKEN",
-                output_file_grp="OCR-D-OCR-KRAKEN",
+            run_processor(KrakenRecognize,
+                          workspace=workspace,
+                          input_file_grp="OCR-D-SEG-KRAKEN",
+                          output_file_grp="OCR-D-OCR-KRAKEN",
             )
-            proc.process()
             workspace.save_mets()
+            # FIXME: add result assertions (find_files, parsing PAGE etc)
 
 if __name__ == "__main__":
     main(__file__)

tests/test_segment.py

@@ -5,7 +5,7 @@
 
 from tests.base import TestCase, assets, main
 
-from ocrd import Resolver
+from ocrd import Resolver, run_processor
 from ocrd_utils import initLogging, pushd_popd
 from ocrd_kraken.segment import KrakenSegment
 
@@ -18,42 +18,45 @@ def test_run_blla(self):
         resolver = Resolver()
         with pushd_popd(tempdir=True) as tempdir:
             workspace = resolver.workspace_from_url(assets.path_to('communist_manifesto/data/mets.xml'), dst_dir=tempdir, download=True)
-            proc = KrakenSegment(
-                workspace,
+            run_processor(
+                KrakenSegment,
+                workspace=workspace,
                 input_file_grp="OCR-D-IMG-BIN",
                 output_file_grp="OCR-D-SEG-LINE-KRAKEN",
                 parameter={'maxcolseps': 0, 'use_legacy': False}
             )
-            proc.process()
             workspace.save_mets()
+            # FIXME: add result assertions (find_files, parsing PAGE etc)
 
     def test_run_blla_regionlevel(self):
         resolver = Resolver()
         with pushd_popd(tempdir=True) as tempdir:
             workspace = resolver.workspace_from_url(assets.path_to('kant_aufklaerung_1784-page-region/data/mets.xml'), dst_dir=tempdir, download=True)
-            proc = KrakenSegment(
-                workspace,
+            run_processor(
+                KrakenSegment,
+                workspace=workspace,
                 input_file_grp="OCR-D-GT-SEG-REGION",
                 output_file_grp="OCR-D-SEG-LINE-KRAKEN",
                 page_id="phys_0005",
                 parameter={'maxcolseps': 0, 'use_legacy': False}
             )
-            proc.process()
             workspace.save_mets()
+            # FIXME: add result assertions (find_files, parsing PAGE etc)
 
     def test_run_legacy(self):
         resolver = Resolver()
         # with pushd_popd('/tmp/kraken-test') as tempdir:
         with pushd_popd(tempdir=True) as tempdir:
             workspace = resolver.workspace_from_url(assets.path_to('communist_manifesto/data/mets.xml'), dst_dir=tempdir, download=True)
-            proc = KrakenSegment(
-                workspace,
+            run_processor(
+                KrakenSegment,
+                workspace=workspace,
                 input_file_grp="OCR-D-IMG-BIN",
                 output_file_grp="OCR-D-SEG-LINE-KRAKEN",
                 parameter={'maxcolseps': 0, 'use_legacy': True}
             )
-            proc.process()
             workspace.save_mets()
+            # FIXME: add result assertions (find_files, parsing PAGE etc)
 
 if __name__ == "__main__":
     main(__file__)