Update app.py

app.py

@@ -80,43 +80,46 @@ async def generate_audio(text, voice, filename):
     return filename
 
 def detect_and_convert_links(text):
+    # Convert Markdown links [text](url) to HTML <a> tags
+    md_link_pattern = re.compile(r'\[(.*?)\]\((https?://[^\s\[\]()<>{}]+)\)')
+    text = md_link_pattern.sub(r'<a href="\2">\1</a>', text)
+    
+    # Convert plain URLs to HTML <a> tags, avoiding already tagged links
     url_pattern = re.compile(
-        r'(https?://|www\.)[^\s\[\]()<>{}]+(\.[^\s\[\]()<>{}]+)+(/[^\s\[\]()<>{}]*)?',
+        r'(?<!href=")(https?://[^\s\[\]()<>{}]+)',
         re.IGNORECASE
     )
-    md_link_pattern = re.compile(r'\[(.*?)\]\((https?://[^\s\[\]()<>{}]+)\)')
-    text = md_link_pattern.sub(r'<a href="\2">\1</a>', text)
-    start_idx = 0
-    result = []
-    while start_idx < len(text):
-        match = url_pattern.search(text, start_idx)
-        if not match:
-            result.append(text[start_idx:])
-            break
-        prev_text = text[start_idx:match.start()]
-        tag_balance = prev_text.count('<a') - prev_text.count('</a')
-        if tag_balance > 0:
-            result.append(text[start_idx:match.end()])
-        else:
-            result.append(text[start_idx:match.start()])
-            url = match.group(0)
-            if url.startswith('www.'):
-                url_with_prefix = 'http://' + url
-            else:
-                url_with_prefix = url
-            result.append(f'<a href="{url_with_prefix}">{url}</a>')
-        start_idx = match.end()
-    return ''.join(result)
+    def replace_url(match):
+        url = match.group(1)
+        return f'<a href="{url}">{url}</a>'
+    
+    text = url_pattern.sub(replace_url, text)
+    return text
 
 def apply_emoji_font(text, emoji_font):
-    link_pattern = re.compile(r'<a\s+href="([^"]+)">(.*?)</a>')
-    links = []
-    def save_link(match):
-        link_idx = len(links)
-        links.append((match.group(1), match.group(2)))
-        return f"###LINK_{link_idx}###"
-    text = link_pattern.sub(save_link, text)
-    text = re.sub(r'<b>(.*?)</b>', lambda m: f'###BOLD_START###{m.group(1)}###BOLD_END###', text)
+    # Preserve links and bold tags
+    link_pattern = re.compile(r'(<a\s+href="[^"]+">.*?</a>)')
+    bold_pattern = re.compile(r'(<b>.*?</b>)')
+    
+    # Split text around links and bold tags
+    segments = []
+    last_pos = 0
+    for match in link_pattern.finditer(text):
+        start, end = match.span()
+        if last_pos < start:
+            segments.append(('text', text[last_pos:start]))
+        segments.append(('link', match.group(0)))
+        last_pos = end
+    for match in bold_pattern.finditer(text[last_pos:]):
+        start, end = match.span()
+        if last_pos < start + last_pos:
+            segments.append(('text', text[last_pos:start + last_pos]))
+        segments.append(('bold', match.group(0)))
+        last_pos = start + end
+    if last_pos < len(text):
+        segments.append(('text', text[last_pos:]))
+    
+    # Apply emoji font to text segments only
     emoji_pattern = re.compile(
         r"([\U0001F300-\U0001F5FF"
         r"\U0001F600-\U0001F64F"
@@ -131,28 +134,32 @@ def apply_emoji_font(text, emoji_font):
         r"\u2600-\u26FF"
         r"\u2700-\u27BF]+)"
     )
+    
     def replace_emoji(match):
         emoji = match.group(1)
         emoji = unicodedata.normalize('NFC', emoji)
         return f'<font face="{emoji_font}">{emoji}</font>'
-    segments = []
-    last_pos = 0
-    for match in emoji_pattern.finditer(text):
-        start, end = match.span()
-        if last_pos < start:
-            segments.append(f'<font face="DejaVuSans">{text[last_pos:start]}</font>')
-        segments.append(replace_emoji(match))
-        last_pos = end
-    if last_pos < len(text):
-        segments.append(f'<font face="DejaVuSans">{text[last_pos:]}</font>')
-    combined_text = ''.join(segments)
-    combined_text = combined_text.replace('###BOLD_START###', '</font><b><font face="DejaVuSans">')
-    combined_text = combined_text.replace('###BOLD_END###', '</font></b><font face="DejaVuSans">')
-    for i, (url, label) in enumerate(links):
-        placeholder = f"###LINK_{i}###"
-        link_html = f'<a href="{url}"><font face="DejaVuSans">{label}</font></a>'
-        combined_text = combined_text.replace(placeholder, link_html)
-    return combined_text
+    
+    result = []
+    for seg_type, content in segments:
+        if seg_type == 'text':
+            # Apply font to non-emoji text and emoji separately
+            parts = []
+            last_pos = 0
+            for match in emoji_pattern.finditer(content):
+                start, end = match.span()
+                if last_pos < start:
+                    parts.append(f'<font face="DejaVuSans">{content[last_pos:start]}</font>')
+                parts.append(replace_emoji(match))
+                last_pos = end
+            if last_pos < len(content):
+                parts.append(f'<font face="DejaVuSans">{content[last_pos:]}</font>')
+            result.append(''.join(parts))
+        else:
+            # Keep links and bold tags unchanged
+            result.append(content)
+    
+    return ''.join(result)
 
 def markdown_to_pdf_content(markdown_text, render_with_bold, auto_bold_numbers, add_space_before_numbered, headings_to_fonts):
     lines = markdown_text.strip().split('\n')
@@ -500,9 +507,8 @@ with st.sidebar:
     enlarge_numbered = st.checkbox("Enlarge Font Size for Numbered Lines", value=True, key="enlarge_numbered")
     add_space_before_numbered = st.checkbox("Add Space Ahead of Numbered Lines", value=False, key="add_space_before_numbered")
 
-    # Here we use a font that has more emojis
     headings_to_fonts = st.checkbox("Headings to Fonts", value=False, key="headings_to_fonts", 
-                               help="Convert Markdown headings (# Heading) and emphasis (*word*) to appropriate font styles")
+                                   help="Convert Markdown headings (# Heading) and emphasis (*word*) to appropriate font styles")
     
     auto_columns = st.checkbox("AutoColumns", value=False, key="auto_columns")