Merge pull request #66 from kba/core-common

Adapt to utils moved to core, #49

ocrd_tesserocr/binarize.py

@@ -21,12 +21,6 @@
 from ocrd import Processor
 
 from .config import TESSDATA_PREFIX, OCRD_TOOL
-from .common import (
-    image_from_page,
-    image_from_segment,
-    save_image_file,
-    membername
-)
 
 TOOL = 'ocrd-tesserocr-binarize'
 LOG = getLogger('processor.TesserocrBinarize')
@@ -52,6 +46,7 @@ def process(self):
         
         Produce a new output file by serialising the resulting hierarchy.
         """
+        # pylint: disable=attribute-defined-outside-init
         try:
             self.page_grp, self.image_grp = self.output_file_grp.split(',')
         except ValueError:
@@ -77,16 +72,16 @@ def process(self):
                                                                          value=self.parameter[name])
                                                                for name in self.parameter.keys()])]))
                 page = pcgts.get_Page()
-                page_image, page_xywh, _ = image_from_page(
-                    self.workspace, page, page_id)
+                page_image, page_xywh, _ = self.workspace.image_from_page(
+                    page, page_id)
                 LOG.info("Binarizing on '%s' level in page '%s'", oplevel, page_id)
 
                 regions = page.get_TextRegion() + page.get_TableRegion()
                 if not regions:
                     LOG.warning("Page '%s' contains no text regions", page_id)
                 for region in regions:
-                    region_image, region_xywh = image_from_segment(
-                        self.workspace, region, page_image, page_xywh)
+                    region_image, region_xywh = self.workspace.image_from_segment(
+                        region, page_image, page_xywh)
                     if oplevel == 'region':
                         tessapi.SetPageSegMode(PSM.SINGLE_BLOCK)
                         self._process_segment(tessapi, RIL.BLOCK, region, region_image, region_xywh,
@@ -98,8 +93,8 @@ def process(self):
                             LOG.warning("Page '%s' region '%s' contains no text lines",
                                         page_id, region.id)
                         for line in lines:
-                            line_image, line_xywh = image_from_segment(
-                                self.workspace, line, region_image, region_xywh)
+                            line_image, line_xywh = self.workspace.image_from_segment(
+                                line, region_image, region_xywh)
                             tessapi.SetPageSegMode(PSM.SINGLE_LINE)
                             self._process_segment(tessapi, RIL.TEXTLINE, line, line_image, line_xywh,
                                                   "line '%s'" % line.id, input_file.pageId,
@@ -129,7 +124,7 @@ def _process_segment(self, tessapi, ril, segment, image, xywh, where, page_id, f
             LOG.error('Cannot binarize %s', where)
             return
         # update METS (add the image file):
-        file_path = save_image_file(self.workspace, image_bin,
+        file_path = self.workspace.save_image_file(image_bin,
                                     file_id,
                                     page_id=page_id,
                                     file_grp=self.image_grp)

ocrd_tesserocr/crop.py

@@ -4,6 +4,7 @@
 import tesserocr
 from ocrd_utils import (
     getLogger, concat_padded,
+    bbox_from_points, points_from_bbox, bbox_from_xywh,
     MIMETYPE_PAGE
 )
 from ocrd_modelfactory import page_from_file
@@ -18,10 +19,6 @@
 from ocrd import Processor
 
 from .config import TESSDATA_PREFIX, OCRD_TOOL
-from .common import (
-    bbox_from_points, points_from_bbox,
-    bbox_from_xywh, save_image_file
-)
 
 TOOL = 'ocrd-tesserocr-crop'
 LOG = getLogger('processor.TesserocrCrop')
@@ -113,7 +110,7 @@ def process(self):
                 # iterate over all text blocks and compare their
                 # bbox extent to the running min and max values
                 for component in tessapi.GetComponentImages(tesserocr.RIL.BLOCK, True):
-                    image, xywh, index, para = component
+                    image, xywh, index, _ = component
                     #
                     # the region reference in the reading order element
                     #
@@ -163,7 +160,7 @@ def process(self):
                     file_id = input_file.ID.replace(self.input_file_grp, FILEGRP_IMG)
                     if file_id == input_file.ID:
                         file_id = concat_padded(FILEGRP_IMG, n)
-                    file_path = save_image_file(self.workspace, page_image,
+                    file_path = self.workspace.save_image_file(page_image,
                                                 file_id,
                                                 page_id=page_id,
                                                 file_grp=FILEGRP_IMG)

ocrd_tesserocr/deskew.py

@@ -12,6 +12,7 @@
 
 from ocrd_utils import (
     getLogger, concat_padded,
+    membername,
     MIMETYPE_PAGE
 )
 from ocrd_modelfactory import page_from_file
@@ -25,12 +26,6 @@
 from ocrd import Processor
 
 from .config import TESSDATA_PREFIX, OCRD_TOOL
-from .common import (
-    image_from_page,
-    image_from_segment,
-    save_image_file,
-    membername
-)
 
 TOOL = 'ocrd-tesserocr-deskew'
 LOG = getLogger('processor.TesserocrDeskew')
@@ -45,19 +40,19 @@ def __init__(self, *args, **kwargs):
 
     def process(self):
         """Performs deskewing of the page / region with Tesseract on the workspace.
-        
+
         Open and deserialise PAGE input files and their respective images,
         then iterate over the element hierarchy down to the region level
         for all text and table regions.
-        
+
         Set up Tesseract to recognise the region image's orientation, skew
         and script (with both OSD and AnalyseLayout). Rotate the image
         accordingly, and annotate the angle, readingDirection and textlineOrder.
-        
+
         Create a corresponding image file, and reference it as AlternativeImage
         in the region element and as file with a fileGrp USE `OCR-D-IMG-DESKEW`
         in the workspace.
-        
+
         Produce a new output file by serialising the resulting hierarchy.
         """
         oplevel = self.parameter['operation_level']
@@ -84,8 +79,8 @@ def process(self):
                                                                          value=self.parameter[name])
                                                                for name in self.parameter.keys()])]))
                 page = pcgts.get_Page()
-                page_image, page_xywh, page_image_info = image_from_page(
-                    self.workspace, page, page_id)
+                page_image, page_xywh, page_image_info = self.workspace.image_from_page(
+                    page, page_id)
                 if page_image_info.xResolution != 1:
                     dpi = page_image_info.xResolution
                     if page_image_info.resolutionUnit == 'cm':
@@ -102,8 +97,8 @@ def process(self):
                     if not regions:
                         LOG.warning("Page '%s' contains no text regions", page_id)
                     for region in regions:
-                        region_image, region_xywh = image_from_segment(
-                            self.workspace, region, page_image, page_xywh)
+                        region_image, region_xywh = self.workspace.image_from_segment(
+                            region, page_image, page_xywh)
                         self._process_segment(tessapi, region, region_image, region_xywh,
                                               "region '%s'" % region.id, input_file.pageId,
                                               file_id + '_' + region.id)
@@ -269,7 +264,7 @@ def _process_segment(self, tessapi, segment, image, xywh, where, page_id, file_i
             #     points = points_from_x0y0x1y1(list(baseline[0]) + list(baseline[1]))
             #     segment.add_Baseline(BaselineType(points=points))
         # update METS (add the image file):
-        file_path = save_image_file(self.workspace, image,
+        file_path = self.workspace.save_image_file(image,
                                     file_id,
                                     page_id=page_id,
                                     file_grp=FILEGRP_IMG)

ocrd_tesserocr/recognize.py

@@ -6,10 +6,13 @@
     PyTessBaseAPI, get_languages)
 
 from ocrd_utils import (
-    getLogger, concat_padded,
-    points_from_x0y0x1y1,
-    xywh_from_points, points_from_xywh,
-    MIMETYPE_PAGE)
+    getLogger,
+    concat_padded,
+    points_from_polygon,
+    polygon_from_x0y0x1y1,
+    coordinates_for_segment,
+    MIMETYPE_PAGE
+)
 from ocrd_models.ocrd_page import (
     CoordsType,
     GlyphType, WordType,
@@ -21,14 +24,6 @@
 from ocrd import Processor
 
 from .config import TESSDATA_PREFIX, OCRD_TOOL
-from .common import (
-    points_from_polygon,
-    xywh_from_polygon,
-    polygon_from_x0y0x1y1,
-    coordinates_for_segment,
-    image_from_page,
-    image_from_segment
-)
 
 TOOL = 'ocrd-tesserocr-recognize'
 LOG = getLogger('processor.TesserocrRecognize')
@@ -133,8 +128,8 @@ def process(self):
                                                                          value=self.parameter[name])
                                                                for name in self.parameter.keys()])]))
                 page = pcgts.get_Page()
-                page_image, page_xywh, page_image_info = image_from_page(
-                    self.workspace, page, page_id)
+                page_image, page_xywh, page_image_info = self.workspace.image_from_page(
+                    page, page_id)
                 if page_image_info.xResolution != 1:
                     dpi = page_image_info.xResolution
                     if page_image_info.resolutionUnit == 'cm':
@@ -165,8 +160,8 @@ def process(self):
 
     def _process_regions(self, tessapi, regions, page_image, page_xywh):
         for region in regions:
-            region_image, region_xywh = image_from_segment(
-                self.workspace, region, page_image, page_xywh)
+            region_image, region_xywh = self.workspace.image_from_segment(
+                region, page_image, page_xywh)
             if self.parameter['textequiv_level'] == 'region':
                 tessapi.SetImage(region_image)
                 tessapi.SetPageSegMode(PSM.SINGLE_BLOCK)
@@ -191,8 +186,8 @@ def _process_lines(self, tessapi, textlines, region_image, region_xywh):
         for line in textlines:
             if self.parameter['overwrite_words']:
                 line.set_Word([])
-            line_image, line_xywh = image_from_segment(
-                self.workspace, line, region_image, region_xywh)
+            line_image, line_xywh = self.workspace.image_from_segment(
+                line, region_image, region_xywh)
             # todo: Tesseract works better if the line images have a 5px margin everywhere
             tessapi.SetImage(line_image)
             # RAW_LINE fails with pre-LSTM models, but sometimes better with LSTM models
@@ -268,8 +263,8 @@ def _process_words_in_line(self, result_it, line, line_xywh):
 
     def _process_existing_words(self, tessapi, words, line_image, line_xywh):
         for word in words:
-            word_image, word_xywh = image_from_segment(
-                self.workspace, word, line_image, line_xywh)
+            word_image, word_xywh = self.workspace.image_from_segment(
+                word, line_image, line_xywh)
             tessapi.SetImage(word_image)
             tessapi.SetPageSegMode(PSM.SINGLE_WORD)
             if self.parameter['textequiv_level'] == 'word':
@@ -296,8 +291,8 @@ def _process_existing_words(self, tessapi, words, line_image, line_xywh):
 
     def _process_existing_glyphs(self, tessapi, glyphs, word_image, word_xywh):
         for glyph in glyphs:
-            glyph_image, glyph_xywh = image_from_segment(
-                self.workspace, glyph, word_image, word_xywh)
+            glyph_image, _ = self.workspace.image_from_segment(
+                glyph, word_image, word_xywh)
             tessapi.SetImage(glyph_image)
             tessapi.SetPageSegMode(PSM.SINGLE_CHAR)
             LOG.debug("Recognizing text in glyph '%s'", glyph.id)

ocrd_tesserocr/segment_line.py

@@ -19,10 +19,6 @@
 )
 
 from .config import TESSDATA_PREFIX, OCRD_TOOL
-from .common import (
-    image_from_page,
-    image_from_segment
-)
 
 TOOL = 'ocrd-tesserocr-segment-line'
 LOG = getLogger('processor.TesserocrSegmentLine')
@@ -70,8 +66,8 @@ def process(self):
                                                                          value=self.parameter[name])
                                                                for name in self.parameter.keys()])]))
                 page = pcgts.get_Page()
-                page_image, page_xywh, page_image_info = image_from_page(
-                    self.workspace, page, page_id)
+                page_image, page_xywh, page_image_info = self.workspace.image_from_page(
+                    page, page_id)
                 if page_image_info.xResolution != 1:
                     dpi = page_image_info.xResolution
                     if page_image_info.resolutionUnit == 'cm':
@@ -86,8 +82,8 @@ def process(self):
                         else:
                             LOG.warning('keeping existing TextLines in region "%s"', region.id)
                     LOG.debug("Detecting lines in region '%s'", region.id)
-                    region_image, region_xywh = image_from_segment(
-                        self.workspace, region, page_image, page_xywh)
+                    region_image, region_xywh = self.workspace.image_from_segment(
+                        region, page_image, page_xywh)
                     tessapi.SetImage(region_image)
                     for line_no, component in enumerate(tessapi.GetComponentImages(RIL.TEXTLINE, True, raw_image=True)):
                         line_id = '%s_line%04d' % (region.id, line_no)

ocrd_tesserocr/segment_region.py

@@ -7,10 +7,15 @@
 )
 
 from ocrd_utils import (
-    getLogger, concat_padded,
+    getLogger,
+    concat_padded,
     points_from_x0y0x1y1,
-    points_from_xywh, xywh_from_points,
-    MIMETYPE_PAGE)
+    points_from_xywh,
+    xywh_from_points,
+    MIMETYPE_PAGE,
+    points_from_polygon,
+    membername
+)
 from ocrd_modelfactory import page_from_file
 from ocrd_models.ocrd_page import (
     MetadataItemType,
@@ -29,12 +34,6 @@
 from ocrd import Processor
 
 from .config import TESSDATA_PREFIX, OCRD_TOOL
-from .common import (
-    image_from_page,
-    save_image_file,
-    points_from_polygon,
-    membername
-)
 
 TOOL = 'ocrd-tesserocr-segment-region'
 LOG = getLogger('processor.TesserocrSegmentRegion')
@@ -106,7 +105,7 @@ def process(self):
                         page.set_TextRegion([])
                     else:
                         LOG.warning('keeping existing TextRegions')
-                # todo: also make non-text regions protected?
+                # TODO: also make non-text regions protected?
                 page.set_AdvertRegion([])
                 page.set_ChartRegion([])
                 page.set_ChemRegion([])
@@ -126,8 +125,8 @@ def process(self):
                         page.set_ReadingOrder([])
                     else:
                         LOG.warning('keeping existing ReadingOrder')
-                page_image, page_xywh, page_image_info = image_from_page(
-                    self.workspace, page, page_id)
+                page_image, page_xywh, page_image_info = self.workspace.image_from_page(
+                    page, page_id)
                 if page_image_info.xResolution != 1:
                     dpi = page_image_info.xResolution
                     if page_image_info.resolutionUnit == 'cm':
@@ -259,9 +258,9 @@ def _process_page(self, it, page, page_image, page_xywh, page_id, file_id):
                 #  GetBinaryImage).
                 # You have been warned!
                 # get the raw image (masked by white space along the block polygon):
-                region_image, top, left = it.GetImage(RIL.BLOCK, self.parameter['padding'], page_image)
+                region_image, _, _ = it.GetImage(RIL.BLOCK, self.parameter['padding'], page_image)
                 # update METS (add the image file):
-                file_path = save_image_file(self.workspace, region_image,
+                file_path = self.workspace.save_image_file(region_image,
                                             file_id + '_' + ID,
                                             page_id=page_id,
                                             file_grp=FILEGRP_IMG)