Update app.py

app.py

@@ -115,33 +115,31 @@ def normalise_hyphens(text):
     # Replace hyphen variants with U+002D for internal consistency
     return text.replace('\u2011', '-').replace('\u2013', '-').replace('\u2014', '-')
 
-def get_longest_segment(text):
-    # Split text by various dash types (hyphen, en dash, em dash, non-breaking hyphen)
-    dash_pattern = r'[\u002D\u2011\u2013\u2014]'
-    segments = re.split(dash_pattern, text)
-    # Remove empty segments and strip whitespace
-    segments = [segment.strip() for segment in segments if segment.strip()]
-    if not segments:
-        return text  # Return original text if no valid segments
-    # Return the longest segment
-    return max(segments, key=len)
+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):
-    # Get the longest segment if text contains dashes
-    fragment_text = get_longest_segment(text)
     # Encode text for W3C Text Fragments, preserving only regular hyphens (U+002D)
-    return urllib.parse.quote(fragment_text, safe='-')
+    # Non-breaking hyphens (U+2011) are encoded as %E2%80%91
+    # En dashes (U+2013) are encoded as %E2%80%93
+    # Em dashes (U+2014) are encoded as %E2%80%94
+    return urllib.parse.quote(text, safe='-')
 
-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):
+    # 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 format_citation_html(url, fragment_text, author, year, scc_hash):
-    # Use the longest segment for the text fragment URL
+    # Use original fragment_text for text fragment URL to match external source
     encoded_fragment = encode_text_fragment(fragment_text)
     full_url = f"{url}#:~:text={encoded_fragment}"
     return f'<a href="{full_url}" data-hash="{scc_hash}">{author} ({year})</a>'