Fix font issue with spans

VikParuchuri · May 7, 2024 · 50398d5 · 50398d5
1 parent ebace90
commit 50398d5
Show file tree

Hide file tree

Showing 3 changed files with 5 additions and 3 deletions.
diff --git a/pdftext/inference.py b/pdftext/inference.py
@@ -128,7 +128,9 @@ def infer_single_page(text_chars, block_threshold=settings.BLOCK_THRESHOLD):
             prediction_probs = yield training_row
             # First item is probability of same line/block, second is probability of new line, third is probability of new block
             if prediction_probs[0] >= .5:
-                pass
+                # Ensure we update spans properly for font info when predicting no new line
+                if prev_font_info != font_info:
+                    span = update_span(line, span)
             elif prediction_probs[2] > block_threshold:
                 span = update_span(line, span)
                 line = update_line(block, line)

diff --git a/pdftext/pdf/chars.py b/pdftext/pdf/chars.py
@@ -75,7 +75,7 @@ def get_pdfium_chars(pdf, fontname_sample_freq=settings.FONTNAME_SAMPLE_FREQ, pa
 
             rotation = pdfium_c.FPDFText_GetCharAngle(text_page, i)
             rotation = rotation * 180 / math.pi # convert from radians to degrees
-            coords = text_page.get_charbox(i, loose=rotation == 0) # Loose doesn't work properly when page is rotated
+            coords = text_page.get_charbox(i, loose=rotation == 0) # Loose doesn't work properly when charbox is rotated
             device_coords = page_bbox_to_device_bbox(page, coords, page_width, page_height, bl_origin, page_rotation, normalize=True)
 
             char_info = {

diff --git a/pyproject.toml b/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "pdftext"
-version = "0.3.6"
+version = "0.3.7"
 description = "Extract structured text from pdfs quickly"
 authors = ["Vik Paruchuri <[email protected]>"]
 license = "Apache-2.0"