2.0

app.py

@@ -6,55 +6,66 @@ def insert_points(text):
     counter = 1
 
     # -----------------------------
-    # 1) COLLAPSE MULTIPLE BLANK LINES INTO ONE
-    #    e.g. 2+ newlines in a row => single newline
+    # 0) NORMALIZE LINE ENDINGS AND REMOVE EXCESS WHITESPACE
     # -----------------------------
+    text = text.strip()
+    # Replace Windows-style line endings
+    text = text.replace('\r\n', '\n')
+    # Collapse multiple blank lines into one
     text = re.sub(r'\n\s*\n+', '\n', text)
 
     # -----------------------------
-    # 2) PROCESS HEADINGS FIRST (before dealing with newlines)
-    #    This ensures headings get properly tagged
+    # 1) FIRST PASS: INSERT NEWLINES BEFORE HEADINGS
+    #    This separates headings that might be in running text
+    # -----------------------------
+    text = re.sub(r'(\s+)(#+\s+)', r'\1\n\2', text)
+
+    # Special case for heading at the start without newline
+    if text.startswith('#'):
+        text = '\n' + text
+
+    # -----------------------------
+    # 2) TAG ALL HEADINGS
+    #    Now that headings are on their own lines, tag them
     # -----------------------------
     def heading_replacer(m):
         nonlocal counter
-        # Get the full heading match
-        full_heading = m.group(0)
-        # Extract just the hash symbols
-        hashes = m.group(1)
-        # Insert a newline before and after the tag
-        replacement = f"\n[INSERT_POINT_{counter:03d}]\n{full_heading}"
+        heading = m.group(1)  # The entire heading including hash symbols
+        tag = f"[INSERT_POINT_{counter:03d}]\n"
         counter += 1
-        return replacement
+        return f"\n{tag}{heading}"
 
-    # Match headings that are at the start of a line (preceded by newline or start of string)
-    # We capture the full heading pattern including content after the hashes
-    text = re.sub(r'(^|\n)(\#{1,6}[ \t].+?)(?=\n|$)', heading_replacer, text)
+    # Look for any heading pattern (one or more # followed by space and text)
+    text = re.sub(r'\n(#+\s+.+?)(?=\n|$)', heading_replacer, text)
 
     # -----------------------------
-    # 3) INSERT A TAG BEFORE EACH PARAGRAPH BREAK
-    #    These are newlines that aren't already tagged
+    # 3) TAG PARAGRAPH BREAKS
+    #    Add tags at meaningful paragraph breaks
     # -----------------------------
-    def newline_replacer(_match):
+    def paragraph_replacer(m):
         nonlocal counter
-        # Insert tag on its own line
-        out = f"\n[INSERT_POINT_{counter:03d}]\n"
+        tag = f"[INSERT_POINT_{counter:03d}]\n"
         counter += 1
-        return out
+        return f"\n{tag}"
 
-    # Replace newlines that aren't already preceded by a tag
-    text = re.sub(r'(?<!\[INSERT_POINT_\d{3}]\n)\n', newline_replacer, text)
+    # Find newlines that aren't already followed by a tag
+    text = re.sub(r'\n(?!\[INSERT_POINT_)', paragraph_replacer, text)
 
     # -----------------------------
-    # 4) CLEAN UP: Remove any extra newlines before tags
-    #    (Sometimes we might have created double newlines)
+    # 4) ADD TAG AT THE BEGINNING IF NEEDED
     # -----------------------------
-    text = re.sub(r'\n\n(\[INSERT_POINT_\d{3}])', r'\n\1', text)
-
-    # If text doesn't start with a newline + tag, add one at the beginning
     if not text.startswith('\n[INSERT_POINT_'):
         text = f"\n[INSERT_POINT_{counter:03d}]\n" + text
         counter += 1
 
+    # -----------------------------
+    # 5) 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)
+
     return text
 
 
@@ -71,10 +82,11 @@ demo = gr.Interface(
     outputs=gr.Textbox(label="Processed Text with Tags"),
     title="Insert Point Tagger",
     description=(
-        "1) Collapses multiple blank lines into one.\n"
-        "2) Inserts [INSERT_POINT_XXX] tags on their own lines.\n"
-        "3) Properly tags headings and paragraph breaks.\n"
-        "4) Maintains sequential numbering for all tags."
+        "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."
     ),
 )