handle zero area chars when reconstructing spans

VikParuchuri · Jan 25, 2025 · 1452098 · 1452098
1 parent b2a6ee9
commit 1452098
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 2 deletions.
diff --git a/pdftext/pdf/links.py b/pdftext/pdf/links.py
@@ -187,7 +187,10 @@ def _reconstruct_spans(orig_span: dict, links: List[Link]) -> List[Span]:
         char_bbox = Bbox(char['bbox'].bbox)
         intersections: List[Tuple[float, Link]] = []
         for i, link_bbox in enumerate(link_bboxes):
-            area = link_bbox.intersection_area(char_bbox)
+            if char_bbox.area > 0:
+                area = link_bbox.intersection_area(char_bbox)
+            else:
+                area = link_bbox.intersection_area(Bbox(char['bbox'].bbox, ensure_nonzero_area=True))
             if area > 0:
                 intersections.append((area, links[i]))
 

diff --git a/pdftext/schema.py b/pdftext/schema.py
@@ -5,8 +5,12 @@
 
 
 class Bbox:
-    def __init__(self, bbox: List[float]):
+    def __init__(self, bbox: List[float], ensure_nonzero_area=False):
+        if ensure_nonzero_area:
+            bbox[2] = max(bbox[0], bbox[2] + 1)
+            bbox[3] = max(bbox[1], bbox[3] + 1)
         self.bbox = bbox
+        self.ensure_nonzero_area = ensure_nonzero_area
 
     def __getitem__(self, item):
         return self.bbox[item]