Update TextProcessor.py

TextProcessor.py

@@ -16,7 +16,7 @@ def paragraphs(folderpath):
     for (i, pg_ind) in enumerate(pg_indices):
         lines = eval(df[df['Pg Ind'] == i]['Lines'].tolist()[0])
         pg_x1s, pg_x2s = [], []
-        for (j,n) in enumerate(lines):
+        for (j, n) in enumerate(lines):
             x1s.append(n[0])
             y1s.append(n[1])
             x2s.append(n[2])
@@ -34,42 +34,42 @@ def paragraphs(folderpath):
         if j == 0:
             para = []
             continue
-        
+
         if len(line_texts[j]) > 0:
-            prior_median = (baselines[pg_inds[j-1]] + rights[pg_inds[j-1]])/2
-            current_median = (baselines[pg_inds[j]] + rights[pg_inds[j]])/2
-            
-            prior_endswith_period = re.search('[:\.]([^A-z]{0,2})$', line_texts[j-1].strip()) is not None # Include colon?
-            prior_is_section_header = re.search('^([ABCDEIVX]+)$', line_texts[j-1].strip()) is not None
+            prior_median = (baselines[pg_inds[j - 1]] + rights[pg_inds[j - 1]]) / 2
+            current_median = (baselines[pg_inds[j]] + rights[pg_inds[j]]) / 2
+
+            prior_endswith_period = re.search('[:\.]([^A-z]{0,2})$',line_texts[j - 1].strip()) is not None  # Include colon?
+            prior_is_section_header = re.search('^([ABCDEIVX]+)$', line_texts[j - 1].strip()) is not None
             current_is_section_header = re.search('^([ABCDEIVX]+)$', line_texts[j].strip()) is not None
-            prior_is_asterisk = re.search('^([\s\*]+)$', line_texts[j-1].strip()) is not None
+            prior_is_asterisk = re.search('^([\s\*]+)$', line_texts[j - 1].strip()) is not None
             current_is_asterisk = re.search('^([\s\*]+)$', line_texts[j].strip()) is not None
-            prior_is_date = re.search('(\[[A-z\s0-9]*,\s[0-9]*]+)$', line_texts[j-1].strip()) is not None
-            
-            current_tabbed = x1s[j]-baselines[pg_inds[j]] > 7
-            prior_tabbed = x1s[j-1]-baselines[pg_inds[j-1]] > 7
-            prior_supertabbed = x1s[j-1]-baselines[pg_inds[j-1]] > 18
-            current_supertabbed = x1s[j]-baselines[pg_inds[j]] > 18
-            prior_more_left = (x1s[j]-baselines[pg_inds[j]])-(x1s[j-1]-baselines[pg_inds[j-1]]) > 7
-            prior_right_margin = x1s[j-1] > prior_median
+            prior_is_date = re.search('(\[[A-z\s0-9]*,\s[0-9]*]+)$', line_texts[j - 1].strip()) is not None
+
+            current_tabbed = x1s[j] - baselines[pg_inds[j]] > 7
+            prior_tabbed = x1s[j - 1] - baselines[pg_inds[j - 1]] > 7
+            prior_supertabbed = x1s[j - 1] - baselines[pg_inds[j - 1]] >= 11
+            current_supertabbed = x1s[j] - baselines[pg_inds[j]] >= 11
+            prior_more_left = (x1s[j] - baselines[pg_inds[j]]) - (x1s[j - 1] - baselines[pg_inds[j - 1]]) > 7
+            prior_right_margin = x1s[j - 1] > prior_median
 
             is_section_header = (prior_is_section_header or current_is_section_header or prior_is_asterisk or current_is_asterisk or prior_is_date)
-            is_the_classic = (prior_endswith_period and current_tabbed and prior_more_left and not prior_supertabbed) # Note: Supertabbing oofs stuff
+            is_the_classic = (prior_endswith_period and current_tabbed and prior_more_left and not prior_supertabbed)  # Note: Supertabbing oofs stuff
             is_start_blockquote = (prior_endswith_period and current_supertabbed and prior_more_left and not is_inblock)
             is_after_blockquote = (prior_endswith_period and not current_supertabbed and is_inblock)
             is_after_disposition = (prior_right_margin and current_tabbed)
             is_after_oneline_paragraph = (prior_tabbed and current_tabbed and not prior_supertabbed and not current_supertabbed and not is_inblock)
-            
+
             if is_start_blockquote and not is_section_header:
                 is_inblock = True
             if is_after_blockquote:
                 is_inblock = False
-                
+
             if is_section_header or is_the_classic or is_after_oneline_paragraph or is_start_blockquote or is_after_blockquote or is_after_disposition:
                 paras.append(para)
                 para = []
-                print('\n')
-            print(str([j, pg_inds[j]]) + ':\t' + str(is_the_classic) + '\t' + str(is_start_blockquote)  + '\t' + str(is_after_blockquote)+ '\t' + str(is_after_disposition) + '\t' + line_text)
+            #     print('\n')
+            # print(str([j, pg_inds[j]]) + ':\t' + str(is_the_classic) + '\t' + str(is_start_blockquote)  + '\t' + str(is_after_blockquote)+ '\t' + str(is_after_disposition) + '\t' + str(is_after_oneline_paragraph) + '\t' + line_text)
             para.append((pg_inds[j], line_inds[j], line_text))
 
     paras.append(para)
@@ -95,3 +95,4 @@ def process_file(folderpath):
         image = cv2.imread(folderpath + '/' + str(pg_ind) + '-processed.png')
         cv2.circle(image, (x1 - 15, int(0.5 * (y1 + y2))), radius=1, color=(240, 32, 160), thickness=2)
         cv2.imwrite(folderpath + '/' + str(pg_ind) + '-processed.png', image)
+