update to OcrdPageResult change

Author: kba

ocrd_kraken/binarize.py

@@ -1,9 +1,10 @@
 from __future__ import absolute_import
-import os
 from os.path import join
 from typing import Optional
 
-from ocrd_models.ocrd_process_result import OcrdProcessResult
+from ocrd.processor.base import OcrdPageResult
+from ocrd.processor.ocrd_page_result import OcrdPageResultImage
+
 import kraken.binarization
 from ocrd import Processor
 from ocrd_utils import assert_file_grp_cardinality, getLogger, make_file_id, MIMETYPE_PAGE
@@ -22,7 +23,7 @@ def executable(self):
     def setup(self):
         self.logger = getLogger('processor.KrakenBinarize')
 
-    def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[str] = None, page_id: Optional[str] = None) -> OcrdProcessResult:
+    def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[str] = None, page_id: Optional[str] = None) -> OcrdPageResult:
         """Binarize the pages/regions/lines with Kraken.
 
         Iterate over the input PAGE element hierarchy down to the requested
@@ -50,14 +51,14 @@ def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[st
         assert page
         page_image, page_xywh, _ = self.workspace.image_from_page(
             page, page_id, feature_filter='binarized')
-        images = []
+        result = OcrdPageResult(pcgts)
         if self.parameter['level-of-operation'] == 'page':
             self.logger.info("Binarizing page '%s'", page_id)
             bin_image = kraken.binarization.nlbin(page_image)
             bin_image_id = f'{output_file_id}.IMG-BIN'
             bin_image_path = join(self.output_file_grp, f'{bin_image_id}.png')
             page.add_AlternativeImage(AlternativeImageType(filename=bin_image_path, comments=f'{page_xywh["features"]},binarized'))
-            images.append((bin_image, bin_image_id, bin_image_path))
+            result.images.append(OcrdPageResultImage(bin_image, bin_image_id, bin_image_path))
         else:
             for region in page.get_AllRegions(classes=['Text']):
                 region_image, region_xywh = self.workspace.image_from_segment(
@@ -68,7 +69,7 @@ def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[st
                     bin_image_id = f'{output_file_id}_{region.id}.IMG-BIN'
                     bin_image_path = join(self.output_file_grp, f'{bin_image_id}.png')
                     region.add_AlternativeImage(AlternativeImageType(filename=bin_image_path, comments=f'{region_xywh["features"]},binarized'))
-                    images.append((bin_image, bin_image_id, bin_image_path))
+                    result.images.append(OcrdPageResultImage(bin_image, bin_image_id, bin_image_path))
                 else:
                     for line in region.get_TextLine():
                         line_image, line_xywh = self.workspace.image_from_segment(
@@ -78,5 +79,5 @@ def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[st
                         bin_image_id = f'{output_file_id}_{region.id}_{line.id}.IMG-BIN'
                         bin_image_path = join(self.output_file_grp, f'{bin_image_id}.png')
                         line.add_AlternativeImage(AlternativeImageType(filename=bin_image_path, comments=f'{page_xywh["features"]},binarized'))
-                        images.append((bin_image, bin_image_id, bin_image_path))
-        return OcrdProcessResult(pcgts, images)
+                        result.images.append(OcrdPageResultImage(bin_image, bin_image_id, bin_image_path))
+        return result

ocrd_kraken/recognize.py

@@ -1,6 +1,5 @@
-from os.path import join
 from typing import Optional, Union
-from ocrd_models import OcrdProcessResult
+from ocrd.processor.base import OcrdPageResult
 import regex
 import itertools
 import numpy as np
@@ -11,20 +10,15 @@
 from ocrd import Processor
 from ocrd_utils import (
     getLogger,
-    make_file_id,
-    assert_file_grp_cardinality,
     coordinates_of_segment,
     coordinates_for_segment,
     bbox_from_polygon,
     points_from_polygon,
     points_from_bbox,
     polygon_from_points,
     xywh_from_points,
-    bbox_from_points,
     transform_coordinates,
-    MIMETYPE_PAGE,
 )
-from ocrd_modelfactory import page_from_file
 from ocrd_models.ocrd_page import (
     OcrdPage,
     RegionRefType,
@@ -45,8 +39,6 @@
     TextLineOrderSimpleType
 )
 
-from ocrd_kraken.config import OCRD_TOOL
-
 class KrakenRecognize(Processor):
 
     @property
@@ -76,7 +68,7 @@ def predict(page_image, segmentation):
                          self.parameter['bidi_reordering'])
         self.predict = predict
 
-    def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[str] = None, page_id: Optional[str] = None) -> OcrdProcessResult:
+    def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[str] = None, page_id: Optional[str] = None) -> OcrdPageResult:
         """Recognize text on lines with Kraken.
 
         Open the parsed PAGE-XML file, then iterate over the element hierarchy
@@ -222,7 +214,7 @@ def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[st
             page_update_higher_textequiv_levels('line', pcgts)
 
         self.logger.info("Finished recognition, serializing")
-        return OcrdProcessResult(pcgts)
+        return OcrdPageResult(pcgts)
 
 # zzz should go into core ocrd_utils
 def baseline_of_segment(segment, coords):

ocrd_kraken/segment.py

@@ -1,20 +1,15 @@
 from typing import Optional
 from PIL import ImageOps
-from os.path import join
 
 from ocrd import Processor
-from ocrd_models import OcrdProcessResult
+from ocrd.processor.ocrd_page_result import OcrdPageResult
 from ocrd_utils import (
     getLogger,
-    assert_file_grp_cardinality,
-    make_file_id,
-    concat_padded,
     polygon_from_x0y0x1y1,
     points_from_polygon,
     polygon_mask,
     coordinates_for_segment,
     coordinates_of_segment,
-    MIMETYPE_PAGE
 )
 import ocrd_models.ocrd_page
 from ocrd_models.ocrd_page import (
@@ -25,16 +20,12 @@
     TextLineType,
     CoordsType,
     BaselineType,
-    to_xml
 )
-from ocrd_modelfactory import page_from_file
 
 import shapely.geometry as geom
 from shapely.prepared import prep as geom_prep
 import torch
 
-from .config import OCRD_TOOL
-
 class KrakenSegment(Processor):
 
     @property
@@ -71,7 +62,7 @@ def segmenter(img, mask=None):
             return segment(img, mask=mask, **kwargs)
         self.segmenter = segmenter
 
-    def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[str] = None, page_id: Optional[str] = None) -> OcrdProcessResult:
+    def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[str] = None, page_id: Optional[str] = None) -> OcrdPageResult:
         """Segment into (regions and) lines with Kraken.
 
         Iterate over the element hierarchy of the PAGE-XML down to the
@@ -142,7 +133,7 @@ def process_page_pcgts(self, *input_pcgts: OcrdPage, output_file_id: Optional[st
                     self.logger.warning('Keeping %d lines in region "%s"', len(region.TextLine or []), region.id)
                 self._process_region(page_image, page_coords, region, zoom)
 
-        return OcrdProcessResult(pcgts)
+        return OcrdPageResult(pcgts)
 
     def _process_page(self, page_image, page_coords, page, zoom=1.0):
         def getmask():