ocropy processors: put derived images under output fileGrp…

binarize/clip/denoise/deskew/dewarp/segment/resegment:
instead of writing AlternativeImages under a separate
fileGrp (named `OCR-D-IMG-*`), place them under the
output fileGrp (along with PAGE-XML) as well.
Differentiate their file ID further by adding
suffixes (named `.IMG-*`).

Author: Robert Sachunsky

ocrd_cis/ocropy/binarize.py

@@ -31,7 +31,6 @@
 
 TOOL = 'ocrd-cis-ocropy-binarize'
 LOG = getLogger('processor.OcropyBinarize')
-FALLBACK_FILEGRP_IMG = 'OCR-D-IMG-BIN'
 
 def binarize(pil_image, method='ocropy', maxskew=2, threshold=0.5, nrm=False):
     LOG.debug('binarizing %dx%d image with method=%s', pil_image.width, pil_image.height, method)
@@ -81,12 +80,6 @@ def __init__(self, *args, **kwargs):
                 LOG.critical('requested method %s does not support grayscale normalized output',
                              self.parameter['method'])
                 raise Exception('only method=ocropy allows grayscale=true')
-            try:
-                self.page_grp, self.image_grp = self.output_file_grp.split(',')
-            except ValueError:
-                self.page_grp = self.output_file_grp
-                self.image_grp = FALLBACK_FILEGRP_IMG
-                LOG.info("No output file group for images specified, falling back to '%s'", FALLBACK_FILEGRP_IMG)
 
     def process(self):
         """Binarize (and optionally deskew/despeckle) the pages/regions/lines of the workspace.
@@ -102,21 +95,24 @@ def process(self):
         by removing connected components smaller than ``noise_maxsize``.
         Finally, apply results to the image and export it as an image file.
 
-        Add the new image file to the workspace with the fileGrp USE given
-        in the second position of the output fileGrp, or ``OCR-D-IMG-BIN``,
-        and an ID based on input file and input element.
+        Add the new image file to the workspace along with the output fileGrp,
+        and using a file ID with suffix ``.IMG-BIN`` along with further
+        identification of the input element.
 
         Reference each new image in the AlternativeImage of the element.
 
         Produce a new output file by serialising the resulting hierarchy.
         """
         level = self.parameter['level-of-operation']
+        assert len(self.output_file_grp.split(',')) == 1, \
+            "Expected exactly one output file group, but '%s' has %d" % (
+                self.output_file_grp, len(self.output_file_grp.split(',')))
 
         for (n, input_file) in enumerate(self.input_files):
             LOG.info("INPUT FILE %i / %s", n, input_file.pageId or input_file.ID)
-            file_id = input_file.ID.replace(self.input_file_grp, self.image_grp)
+            file_id = input_file.ID.replace(self.input_file_grp, self.output_file_grp)
             if file_id == input_file.ID:
-                file_id = concat_padded(self.image_grp, n)
+                file_id = concat_padded(self.output_file_grp, n)
 
             pcgts = page_from_file(self.workspace.download_file(input_file))
             page_id = pcgts.pcGtsId or input_file.pageId or input_file.ID # (PageType has no id)
@@ -164,19 +160,19 @@ def process(self):
                                           file_id + '_' + region.id + '_' + line.id)
 
             # update METS (add the PAGE file):
-            file_id = input_file.ID.replace(self.input_file_grp, self.page_grp)
+            file_id = input_file.ID.replace(self.input_file_grp, self.output_file_grp)
             if file_id == input_file.ID:
-                file_id = concat_padded(self.page_grp, n)
-            file_path = os.path.join(self.page_grp, file_id + '.xml')
+                file_id = concat_padded(self.output_file_grp, n)
+            file_path = os.path.join(self.output_file_grp, file_id + '.xml')
             out = self.workspace.add_file(
                 ID=file_id,
-                file_grp=self.page_grp,
+                file_grp=self.output_file_grp,
                 pageId=input_file.pageId,
                 local_filename=file_path,
                 mimetype=MIMETYPE_PAGE,
                 content=to_xml(pcgts))
             LOG.info('created file ID: %s, file_grp: %s, path: %s',
-                     file_id, self.page_grp, out.local_filename)
+                     file_id, self.output_file_grp, out.local_filename)
 
     def process_page(self, page, page_image, page_xywh, page_id, file_id):
         LOG.info("About to binarize page '%s'", page_id)
@@ -207,15 +203,16 @@ def process_page(self, page, page_image, page_xywh, page_id, file_id):
         page.set_orientation(orientation)
         # update METS (add the image file):
         if self.parameter['grayscale']:
-            file_id += '.nrm'
+            file_id += '.IMG-NRM'
             features += ',grayscale_normalized'
         else:
+            file_id += '.IMG-BIN'
             features += ',binarized'
         file_path = self.workspace.save_image_file(
             bin_image,
             file_id,
             page_id=page_id,
-            file_grp=self.image_grp)
+            file_grp=self.output_file_grp)
         # update PAGE (reference the image file):
         page.add_AlternativeImage(AlternativeImageType(
             filename=file_path,
@@ -251,15 +248,16 @@ def process_region(self, region, region_image, region_xywh, page_id, file_id):
         region.set_orientation(orientation)
         # update METS (add the image file):
         if self.parameter['grayscale']:
-            file_id += '.nrm'
+            file_id += '.IMG-NRM'
             features += ',grayscale_normalized'
         else:
+            file_id += '.IMG-BIN'
             features += ',binarized'
         file_path = self.workspace.save_image_file(
             bin_image,
             file_id,
             page_id=page_id,
-            file_grp=self.image_grp)
+            file_grp=self.output_file_grp)
         # update PAGE (reference the image file):
         region.add_AlternativeImage(AlternativeImageType(
             filename=file_path,
@@ -289,15 +287,16 @@ def process_line(self, line, line_image, line_xywh, page_id, region_id, file_id)
             features += ',despeckled'
         # update METS (add the image file):
         if self.parameter['grayscale']:
-            file_id += '.nrm'
+            file_id += '.IMG-NRM'
             features += ',grayscale_normalized'
         else:
+            file_id += '.IMG-BIN'
             features += ',binarized'
         file_path = self.workspace.save_image_file(
             bin_image,
             file_id,
             page_id=page_id,
-            file_grp=self.image_grp)
+            file_grp=self.output_file_grp)
         # update PAGE (reference the image file):
         line.add_AlternativeImage(AlternativeImageType(
             filename=file_path,

ocrd_cis/ocropy/clip.py

@@ -35,7 +35,6 @@
 
 TOOL = 'ocrd-cis-ocropy-clip'
 LOG = getLogger('processor.OcropyClip')
-FALLBACK_FILEGRP_IMG = 'OCR-D-IMG-CLIP'
 
 class OcropyClip(Processor):
 
@@ -44,13 +43,6 @@ def __init__(self, *args, **kwargs):
         kwargs['ocrd_tool'] = self.ocrd_tool['tools'][TOOL]
         kwargs['version'] = self.ocrd_tool['version']
         super(OcropyClip, self).__init__(*args, **kwargs)
-        if hasattr(self, 'output_file_grp'):
-            try:
-                self.page_grp, self.image_grp = self.output_file_grp.split(',')
-            except ValueError:
-                self.page_grp = self.output_file_grp
-                self.image_grp = FALLBACK_FILEGRP_IMG
-                LOG.info("No output file group for images specified, falling back to '%s'", FALLBACK_FILEGRP_IMG)
 
     def process(self):
         """Clip text regions / lines of the workspace at intersections with neighbours.
@@ -69,9 +61,9 @@ def process(self):
         which are only contained in the neighbour by clipping them to white (background),
         and export the (final) result as image file.
 
-        Add the new image file to the workspace with the fileGrp USE given
-        in the second position of the output fileGrp, or ``OCR-D-IMG-CLIP``,
-        and an ID based on the input file and input element.
+        Add the new image file to the workspace along with the output fileGrp,
+        and using a file ID with suffix ``.IMG-CLIP`` along with further
+        identification of the input element.
 
         Reference each new image in the AlternativeImage of the element.
 
@@ -88,12 +80,15 @@ def process(self):
         # deskewing, because that would make segments incomensurable with their
         # neighbours.
         level = self.parameter['level-of-operation']
+        assert len(self.output_file_grp.split(',')) == 1, \
+            "Expected exactly one output file group, but '%s' has %d" % (
+                self.output_file_grp, len(self.output_file_grp.split(',')))
 
         for (n, input_file) in enumerate(self.input_files):
             LOG.info("INPUT FILE %i / %s", n, input_file.pageId or input_file.ID)
-            file_id = input_file.ID.replace(self.input_file_grp, self.image_grp)
+            file_id = input_file.ID.replace(self.input_file_grp, self.output_file_grp)
             if file_id == input_file.ID:
-                file_id = concat_padded(self.image_grp, n)
+                file_id = concat_padded(self.output_file_grp, n)
 
             pcgts = page_from_file(self.workspace.download_file(input_file))
             page_id = pcgts.pcGtsId or input_file.pageId or input_file.ID # (PageType has no id)
@@ -218,19 +213,19 @@ def process(self):
                                              input_file.pageId, file_id + '_' + region.id + '_' + line.id)
 
             # update METS (add the PAGE file):
-            file_id = input_file.ID.replace(self.input_file_grp, self.page_grp)
+            file_id = input_file.ID.replace(self.input_file_grp, self.output_file_grp)
             if file_id == input_file.ID:
-                file_id = concat_padded(self.page_grp, n)
-            file_path = os.path.join(self.page_grp, file_id + '.xml')
+                file_id = concat_padded(self.output_file_grp, n)
+            file_path = os.path.join(self.output_file_grp, file_id + '.xml')
             out = self.workspace.add_file(
                 ID=file_id,
-                file_grp=self.page_grp,
+                file_grp=self.output_file_grp,
                 pageId=input_file.pageId,
                 local_filename=file_path,
                 mimetype=MIMETYPE_PAGE,
                 content=to_xml(pcgts))
             LOG.info('created file ID: %s, file_grp: %s, path: %s',
-                     file_id, self.page_grp, out.local_filename)
+                     file_id, self.output_file_grp, out.local_filename)
 
     def process_segment(self, segment, segment_mask, segment_polygon, neighbours,
                         background_image, parent_image, parent_coords, parent_bin,
@@ -270,9 +265,9 @@ def process_segment(self, segment, segment_mask, segment_polygon, neighbours,
         # update METS (add the image file):
         file_path = self.workspace.save_image_file(
             segment_image,
-            file_id=file_id,
+            file_id=file_id + '.IMG-CLIP',
             page_id=page_id,
-            file_grp=self.image_grp)
+            file_grp=self.output_file_grp)
         # update PAGE (reference the image file):
         segment.add_AlternativeImage(AlternativeImageType(
             filename=file_path,

ocrd_cis/ocropy/denoise.py

@@ -22,7 +22,6 @@
 
 TOOL = 'ocrd-cis-ocropy-denoise'
 LOG = getLogger('processor.OcropyDenoise')
-FALLBACK_FILEGRP_IMG = 'OCR-D-IMG-DESPECK'
 
 class OcropyDenoise(Processor):
 
@@ -31,13 +30,6 @@ def __init__(self, *args, **kwargs):
         kwargs['ocrd_tool'] = self.ocrd_tool['tools'][TOOL]
         kwargs['version'] = self.ocrd_tool['version']
         super(OcropyDenoise, self).__init__(*args, **kwargs)
-        if hasattr(self, 'output_file_grp'):
-            try:
-                self.page_grp, self.image_grp = self.output_file_grp.split(',')
-            except ValueError:
-                self.page_grp = self.output_file_grp
-                self.image_grp = FALLBACK_FILEGRP_IMG
-                LOG.info("No output file group for images specified, falling back to '%s'", FALLBACK_FILEGRP_IMG)
 
     def process(self):
         """Despeckle the pages / regions / lines of the workspace.
@@ -52,21 +44,24 @@ def process(self):
         smaller than ``noise_maxsize``. Apply results to the image and export
         it as an image file.
 
-        Add the new image file to the workspace with the fileGrp USE given
-        in the second position of the output fileGrp, or ``OCR-D-IMG-DESPECK``,
-        and an ID based on input file and input element.
+        Add the new image file to the workspace along with the output fileGrp,
+        and using a file ID with suffix ``.IMG-DESPECK`` along with further
+        identification of the input element.
 
         Reference each new image in the AlternativeImage of the element.
 
         Produce a new output file by serialising the resulting hierarchy.
         """
         level = self.parameter['level-of-operation']
+        assert len(self.output_file_grp.split(',')) == 1, \
+            "Expected exactly one output file group, but '%s' has %d" % (
+                self.output_file_grp, len(self.output_file_grp.split(',')))
 
         for (n, input_file) in enumerate(self.input_files):
             LOG.info("INPUT FILE %i / %s", n, input_file.pageId or input_file.ID)
-            file_id = input_file.ID.replace(self.input_file_grp, self.image_grp)
+            file_id = input_file.ID.replace(self.input_file_grp, self.output_file_grp)
             if file_id == input_file.ID:
-                file_id = concat_padded(self.image_grp, n)
+                file_id = concat_padded(self.output_file_grp, n)
 
             pcgts = page_from_file(self.workspace.download_file(input_file))
             page_id = pcgts.pcGtsId or input_file.pageId or input_file.ID # (PageType has no id)
@@ -127,19 +122,19 @@ def process(self):
                                              file_id + '_' + region.id + '_' + line.id)
 
             # update METS (add the PAGE file):
-            file_id = input_file.ID.replace(self.input_file_grp, self.page_grp)
+            file_id = input_file.ID.replace(self.input_file_grp, self.output_file_grp)
             if file_id == input_file.ID:
-                file_id = concat_padded(self.page_grp, n)
-            file_path = os.path.join(self.page_grp, file_id + '.xml')
+                file_id = concat_padded(self.output_file_grp, n)
+            file_path = os.path.join(self.output_file_grp, file_id + '.xml')
             out = self.workspace.add_file(
                 ID=file_id,
-                file_grp=self.page_grp,
+                file_grp=self.output_file_grp,
                 pageId=input_file.pageId,
                 local_filename=file_path,
                 mimetype=MIMETYPE_PAGE,
                 content=to_xml(pcgts))
             LOG.info('created file ID: %s, file_grp: %s, path: %s',
-                     file_id, self.page_grp, out.local_filename)
+                     file_id, self.output_file_grp, out.local_filename)
 
     def process_segment(self, segment, segment_image, segment_xywh, zoom, page_id, file_id):
         LOG.info("About to despeckle '%s'", file_id)
@@ -148,9 +143,9 @@ def process_segment(self, segment, segment_image, segment_xywh, zoom, page_id, f
         # update METS (add the image file):
         file_path = self.workspace.save_image_file(
             bin_image,
-            file_id,
+            file_id + '.IMG-DESPECK',
             page_id=page_id,
-            file_grp=self.image_grp)
+            file_grp=self.output_file_grp)
         # update PAGE (reference the image file):
         segment.add_AlternativeImage(AlternativeImageType(
             filename=file_path,

ocrd_cis/ocropy/deskew.py

@@ -25,7 +25,6 @@
 
 TOOL = 'ocrd-cis-ocropy-deskew'
 LOG = getLogger('processor.OcropyDeskew')
-FALLBACK_FILEGRP_IMG = 'OCR-D-IMG-DESKEW'
 
 def deskew(pil_image, maxskew=2):
     array = pil2array(pil_image)
@@ -39,13 +38,6 @@ def __init__(self, *args, **kwargs):
         kwargs['ocrd_tool'] = ocrd_tool['tools'][TOOL]
         kwargs['version'] = ocrd_tool['version']
         super(OcropyDeskew, self).__init__(*args, **kwargs)
-        if hasattr(self, 'output_file_grp'):
-            try:
-                self.page_grp, self.image_grp = self.output_file_grp.split(',')
-            except ValueError:
-                self.page_grp = self.output_file_grp
-                self.image_grp = FALLBACK_FILEGRP_IMG
-                LOG.info("No output file group for images specified, falling back to '%s'", FALLBACK_FILEGRP_IMG)
 
     def process(self):
         """Deskew the regions of the workspace.
@@ -59,19 +51,22 @@ def process(self):
         the deskewing angle of the region (up to ``maxskew``). Annotate the
         angle in the region.
 
-        Add a new image file to the workspace with the fileGrp USE given
-        in the second position of the output fileGrp, or ``OCR-D-IMG-DESKEW``,
-        and an ID based on input file and input element.
+        Add the new image file to the workspace along with the output fileGrp,
+        and using a file ID with suffix ``.IMG-DESKEW`` along with further
+        identification of the input element.
 
         Produce a new output file by serialising the resulting hierarchy.
         """
         level = self.parameter['level-of-operation']
+        assert len(self.output_file_grp.split(',')) == 1, \
+            "Expected exactly one output file group, but '%s' has %d" % (
+                self.output_file_grp, len(self.output_file_grp.split(',')))
 
         for (n, input_file) in enumerate(self.input_files):
             LOG.info("INPUT FILE %i / %s", n, input_file.pageId or input_file.ID)
-            file_id = input_file.ID.replace(self.input_file_grp, self.image_grp)
+            file_id = input_file.ID.replace(self.input_file_grp, self.output_file_grp)
             if file_id == input_file.ID:
-                file_id = concat_padded(self.image_grp, n)
+                file_id = concat_padded(self.output_file_grp, n)
 
             pcgts = page_from_file(self.workspace.download_file(input_file))
             page_id = pcgts.pcGtsId or input_file.pageId or input_file.ID # (PageType has no id)
@@ -117,19 +112,19 @@ def process(self):
                                           file_id + '_' + region.id)
 
             # update METS (add the PAGE file):
-            file_id = input_file.ID.replace(self.input_file_grp, self.page_grp)
+            file_id = input_file.ID.replace(self.input_file_grp, self.output_file_grp)
             if file_id == input_file.ID:
-                file_id = concat_padded(self.page_grp, n)
-            file_path = os.path.join(self.page_grp, file_id + '.xml')
+                file_id = concat_padded(self.output_file_grp, n)
+            file_path = os.path.join(self.output_file_grp, file_id + '.xml')
             out = self.workspace.add_file(
                 ID=file_id,
-                file_grp=self.page_grp,
+                file_grp=self.output_file_grp,
                 pageId=input_file.pageId,
                 local_filename=file_path,
                 mimetype=MIMETYPE_PAGE,
                 content=to_xml(pcgts))
             LOG.info('created file ID: %s, file_grp: %s, path: %s',
-                     file_id, self.page_grp, out.local_filename)
+                     file_id, self.output_file_grp, out.local_filename)
 
     def _process_segment(self, segment, segment_image, segment_coords, segment_id, page_id, file_id):
         angle0 = segment_coords['angle'] # deskewing (w.r.t. top image) already applied to segment_image
@@ -149,9 +144,9 @@ def _process_segment(self, segment, segment_image, segment_coords, segment_id, p
         # update METS (add the image file):
         file_path = self.workspace.save_image_file(
             segment_image,
-            file_id,
+            file_id + '.IMG-DESKEW',
             page_id=page_id,
-            file_grp=self.image_grp)
+            file_grp=self.output_file_grp)
         # update PAGE (reference the image file):
         segment.add_AlternativeImage(AlternativeImageType(
             filename=file_path,