new approach

app.py

@@ -1,49 +1,44 @@
 import re
 import gradio as gr
 
-
 def insert_points(text):
-    # Split the input by newline to handle multi-line text
-    lines = text.split('\n')
-
+    # Initialize a counter accessible by the replacer function
     counter = 1
-    output = []
-
-    for line in lines:
-        # This pattern captures sequences of '#' (1 or more) plus optional trailing spaces
-        # We'll split the line on these sequences, but keep them so we can re-insert with tags.
-        segments = re.split(r'(\#+\s*)', line)
-
-        # We'll rebuild this line piece by piece
-        rebuilt_line = []
-
-        for seg in segments:
-            if re.match(r'^\#+\s*$', seg):
-                # This segment is a run of '#' (with optional trailing space)
-                # Insert a tag BEFORE it
-                tag = f"[INSERT_POINT_{counter:03d}]"
-                counter += 1
-                rebuilt_line.append(tag)
-
-                # Then append the actual '#' segment
-                rebuilt_line.append(seg)
-            else:
-                # Normal text
-                rebuilt_line.append(seg)
-
-        # After processing all segments in this line,
-        # append a final tag to match your "insert a tag after every line" requirement
-        line_tag = f"[INSERT_POINT_{counter:03d}]"
-        counter += 1
-        rebuilt_line.append(line_tag)
-
-        # Join the processed segments for this line
-        output.append("".join(rebuilt_line))
-
-    # Finally, join all lines with newlines
-    return "\n".join(output)
-
 
+    # This function will be called for each match found by re.sub
+    # It inserts the tag *before* the matched text (newline or hash sequence)
+    def replacer(match):
+        nonlocal counter  # Use the counter from the outer scope
+        tag = f"[INSERT_POINT_{counter:03d}]"
+        counter += 1
+        # match.group(0) contains the actual matched string ('\n' or '##' etc.)
+        return tag + match.group(0)
+
+    # The pattern looks for either:
+    # 1) A newline character ('\n')
+    # 2) A sequence of one or more '#' characters ('\#+')
+    #    '#' needs to be escaped ('\#') because it's a special regex character.
+    # The parentheses create capturing groups, but match.group(0) gives the whole match anyway.
+    pattern = r'(\n|\#+)'
+
+    # Use re.sub to find all matches of the pattern and replace them
+    # by calling the 'replacer' function for each match.
+    processed_text = re.sub(pattern, replacer, text)
+
+    # One edge case: If the *very beginning* of the text starts with '#',
+    # the regex above won't match anything *before* it.
+    # We need to check if the text starts with hashes (possibly after whitespace)
+    # and prepend the first tag if necessary.
+    # However, the current `re.sub(pattern, replacer, text)` already handles this
+    # correctly because it finds the '#' sequence at the beginning and the
+    # replacer adds the tag *before* it. Let's re-verify.
+    # Example: If text is "## Title", pattern finds "##" at index 0.
+    # Replacer runs, returns "[INSERT_POINT_001]##". Result is correct.
+    # So, no special handling for the beginning is needed with this pattern.
+
+    return processed_text
+
+# --- Gradio Interface Code (Unchanged from your original) ---
 demo = gr.Interface(
     fn=insert_points,
     inputs=gr.Textbox(
@@ -51,13 +46,14 @@ demo = gr.Interface(
         placeholder="Paste your text here...",
         label="Your Input Text"
     ),
-    outputs="text",
+    outputs=gr.Textbox(label="Processed Text with Tags"), # Changed output type to Textbox for better viewing
     title="Insert Point Tagger",
+    # Updated description for clarity
     description=(
         "Paste a block of text and get '[INSERT_POINT_###]' tags added "
-        "1) after every newline, and 2) before every '#' sequence."
+        "1) **before** every newline, and 2) **before** every '#' sequence (heading)."
     ),
 )
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()