Update app.py

app.py

@@ -115,15 +115,6 @@ def normalise_hyphens(text):
     # Replace hyphen variants with U+002D for internal consistency
     return text.replace('\u2011', '-').replace('\u2013', '-').replace('\u2014', '-')
 
-def select_longest_segment(text):
-    # Split text by various dashes (hyphen, en dash, em dash)
-    dash_variants = ['\u002D', '\u2011', '\u2013', '\u2014']
-    segments = text
-    for dash in dash_variants:
-        segments = segments.split(dash)
-    # Return the longest segment, or original text if no dashes
-    return max(segments, key=len, default=text).strip()
-
 def encode_text_fragment(text):
     # Encode text for W3C Text Fragments, preserving only regular hyphens (U+002D)
     # Non-breaking hyphens (U+2011) are encoded as %E2%80%91
@@ -131,12 +122,13 @@ def encode_text_fragment(text):
     # Em dashes (U+2014) are encoded as %E2%80%94
     return urllib.parse.quote(text, safe='-')
 
-def format_citation_html(url, fragment_text, author, year, scc_hash):
-    # Select the longest segment for the text fragment to avoid breaking the link
-    selected_fragment = select_longest_segment(fragment_text)
-    encoded_fragment = encode_text_fragment(selected_fragment)
-    full_url = f"{url}#:~:text={encoded_fragment}"
-    return f'<a href="{full_url}" data-hash="{scc_hash}">{author} ({year})</a>'
+def generate_citation_hash(author, year, url, fragment_text, cited_text, username, task_name, current_date, current_time):
+    # Normalise hyphens for consistent hash generation
+    normalised_fragment_text = normalise_hyphens(fragment_text)
+    normalised_cited_text = normalise_hyphens(cited_text)
+    normalised_task_name = normalise_hyphens(task_name)
+    data = f"{author}, {year} | {url} | {normalised_fragment_text} | {normalised_cited_text} | {username} | {normalised_task_name} | {current_date} | {current_time}"
+    return hashlib.sha256(data.encode('utf-8')).hexdigest()
 
 def format_citation_html(url, fragment_text, author, year, scc_hash):
     # Use original fragment_text for text fragment URL to match external source