fix(layout,table): update e2e test

Signed-off-by: Clément Doumouro <[email protected]>

Author: ClemDoum

docling/models/layout_model.py

@@ -102,22 +102,17 @@ def download_models(
         )
 
     def draw_clusters_and_cells_side_by_side(
-        self,
-        conv_res,
-        page,
-        clusters,
-        mode_prefix: str,
-        show: bool = False,
+        self, conv_res, page, clusters, mode_prefix: str, show: bool = False
     ):
         """
         Draws a page image side by side with clusters filtered into two categories:
         - Left: Clusters excluding FORM, KEY_VALUE_REGION, and PICTURE.
         - Right: Clusters including FORM, KEY_VALUE_REGION, and PICTURE.
         Includes label names and confidence scores for each cluster.
         """
-        page_image = deepcopy(page.image)
-        scale_x = page_image.width / page.size.width
-        scale_y = page_image.height / page.size.height
+        scale_x = page.image.width / page.size.width
+        scale_y = page.image.height / page.size.height
+
         # Filter clusters for left and right images
         exclude_labels = {
             DocItemLabel.FORM,
@@ -127,8 +122,8 @@ def draw_clusters_and_cells_side_by_side(
         left_clusters = [c for c in clusters if c.label not in exclude_labels]
         right_clusters = [c for c in clusters if c.label in exclude_labels]
         # Create a deep copy of the original image for both sides
-        left_image = page_image
-        right_image = copy.deepcopy(left_image)
+        left_image = copy.deepcopy(page.image)
+        right_image = copy.deepcopy(page.image)
 
         # Draw clusters on both images
         draw_clusters(left_image, left_clusters, scale_x, scale_y)
@@ -191,10 +186,7 @@ def __call__(
 
                     if settings.debug.visualize_raw_layout:
                         self.draw_clusters_and_cells_side_by_side(
-                            conv_res,
-                            page,
-                            clusters,
-                            mode_prefix="raw",
+                            conv_res, page, clusters, mode_prefix="raw"
                         )
 
                     # Apply postprocessing
@@ -228,10 +220,7 @@ def __call__(
 
                 if settings.debug.visualize_layout:
                     self.draw_clusters_and_cells_side_by_side(
-                        conv_res,
-                        page,
-                        processed_clusters,
-                        mode_prefix="postprocessed",
+                        conv_res, page, processed_clusters, mode_prefix="postprocessed"
                     )
 
                 yield page

docling/models/ocr_mac_model.py

@@ -107,10 +107,10 @@ def __call__(
                             x2 = x1 + w * im_width
                             y1 = y2 - h * im_height
 
-                            left = x1 / self.scale
-                            top = y1 / self.scale
-                            right = x2 / self.scale
-                            bottom = y2 / self.scale
+                            left = x1 / self.scale + ocr_rect.l
+                            top = y1 / self.scale + ocr_rect.t
+                            right = x2 / self.scale + ocr_rect.l
+                            bottom = y2 / self.scale + ocr_rect.t
 
                             cells.append(
                                 TextCell(

docling/models/table_structure_model.py

@@ -260,7 +260,7 @@ def __call__(
                                     )
                                     new_bbox = _rotate_bbox(
                                         new_cell.to_bounding_box(),
-                                        orientation=-cells_orientation,
+                                        orientation=cells_orientation,
                                         im_size=scaled_page_im.size,
                                     ).model_dump()
                                     tokens.append(

tests/data_scanned/groundtruth/docling_v1/ocr_test.doctags.txt

@@ -1,11 +1,8 @@
 <document>
-<subtitle-level-1><location><page_1><loc_33><loc_87><loc_68><loc_91></location>This is a table test</subtitle-level-1>
-<paragraph><location><page_1><loc_12><loc_83><loc_61><loc_84></location>The test starts with some random text and then a table image:</paragraph>
-<paragraph><location><page_1><loc_45><loc_76><loc_56><loc_77></location>Some column</paragraph>
-<paragraph><location><page_1><loc_62><loc_76><loc_78><loc_77></location>Some other column</paragraph>
-<paragraph><location><page_1><loc_29><loc_70><loc_37><loc_71></location>Some row</paragraph>
-<paragraph><location><page_1><loc_47><loc_70><loc_54><loc_71></location>some cell</paragraph>
-<paragraph><location><page_1><loc_65><loc_70><loc_76><loc_71></location>have content</paragraph>
-<paragraph><location><page_1><loc_26><loc_64><loc_39><loc_65></location>Some other row</paragraph>
-<paragraph><location><page_1><loc_46><loc_64><loc_55><loc_65></location>other don't</paragraph>
+<table>
+<location><page_1><loc_9><loc_45><loc_70><loc_86></location>
+<row_0><col_0><col_header>Vertically merged</col_0><col_1><col_header>Other merged column</col_1><col_2><col_header>Yet another column</col_2></row_0>
+<row_1><col_0><body>value</col_0><col_1><body>Some other value</col_1><col_2><body>Yet another value</col_2></row_1>
+<row_2><col_0><body>value</col_0><col_1><body>Some other value</col_1><col_2><body>Yet another value</col_2></row_2>
+</table>
 </document>