new linear approach

app.py

@@ -25,46 +25,60 @@ def insert_points(text):
         text = '\n' + text
 
     # -----------------------------
-    # 2) TAG ALL HEADINGS
-    #    Now that headings are on their own lines, tag them
+    # 2) SPLIT TEXT INTO SEGMENTS AND TAG
+    #    Process the text in a single pass to avoid duplicate tags
     # -----------------------------
-    def heading_replacer(m):
-        nonlocal counter
-        heading = m.group(1)  # The entire heading including hash symbols
-        tag = f"[INSERT_POINT_{counter:03d}]\n"
-        counter += 1
-        return f"\n{tag}{heading}"
 
-    # Look for any heading pattern (one or more # followed by space and text)
-    text = re.sub(r'\n(#+\s+.+?)(?=\n|$)', heading_replacer, text)
+    # Split the text into segments (paragraphs and headings)
+    segments = re.split(r'(\n#+\s+.+?(?=\n|$)|\n)', text)
+    segments = [s for s in segments if s]  # Remove empty segments
 
-    # -----------------------------
-    # 3) TAG PARAGRAPH BREAKS
-    #    Add tags at meaningful paragraph breaks
-    # -----------------------------
-    def paragraph_replacer(m):
-        nonlocal counter
-        tag = f"[INSERT_POINT_{counter:03d}]\n"
-        counter += 1
-        return f"\n{tag}"
+    result = []
+    previous_was_tag = False
+
+    for segment in segments:
+        # If this is a newline, add a tag (but not after another tag)
+        if segment == '\n':
+            if not previous_was_tag:
+                result.append(f"\n[INSERT_POINT_{counter:03d}]\n")
+                counter += 1
+                previous_was_tag = True
+            else:
+                result.append('\n')  # Just add the newline without a tag
+
+        # If this is a heading, add a tag before it
+        elif segment.startswith('\n#'):
+            if not previous_was_tag:
+                result.append(f"\n[INSERT_POINT_{counter:03d}]")
+                counter += 1
+                previous_was_tag = True
+            result.append(segment)
+            previous_was_tag = False
+
+        # Regular text segment
+        else:
+            result.append(segment)
+            previous_was_tag = False
 
-    # Find newlines that aren't already followed by a tag
-    text = re.sub(r'\n(?!\[INSERT_POINT_)', paragraph_replacer, text)
+    # Join all segments back together
+    text = ''.join(result)
 
     # -----------------------------
-    # 4) ADD TAG AT THE BEGINNING IF NEEDED
+    # 3) ADD TAG AT THE BEGINNING IF NEEDED
     # -----------------------------
-    if not text.startswith('\n[INSERT_POINT_'):
-        text = f"\n[INSERT_POINT_{counter:03d}]\n" + text
+    if not text.startswith('[INSERT_POINT_'):
+        text = f"[INSERT_POINT_{counter:03d}]\n" + text
         counter += 1
 
     # -----------------------------
-    # 5) CLEAN UP: Remove excess newlines
+    # 4) CLEAN UP: Remove excess newlines
     # -----------------------------
     # Remove extra blank lines before tags
     text = re.sub(r'\n\n+(\[INSERT_POINT_)', r'\n\1', text)
     # Remove extra blank lines at the beginning
     text = re.sub(r'^\n+', '', text)
+    # Ensure no consecutive tags
+    text = re.sub(r'(\[INSERT_POINT_\d{3}]\n)\s*\[INSERT_POINT_\d{3}]', r'\1', text)
 
     return text
 
@@ -83,10 +97,10 @@ demo = gr.Interface(
     title="Insert Point Tagger",
     description=(
         "This processor:\n"
-        "1) Separates and tags headings (text starting with #).\n"
-        "2) Tags paragraph breaks with sequential numbers.\n"
-        "3) Places each tag on its own line.\n"
-        "4) Works with heading patterns that may be embedded in running text."
+        "1) Tags headings and paragraph breaks with sequential numbers.\n"
+        "2) Places each tag on its own line.\n"
+        "3) Ensures consistent, sequential numbering (001, 002, etc.).\n"
+        "4) Avoids consecutive tags - never two tags in a row."
     ),
 )