feat: server-side favicon proxy

Author: Algorithm5838

backend/open_webui/routers/utils.py

@@ -1,6 +1,15 @@
 import black
+import hashlib
+import ipaddress
 import logging
 import markdown
+import socket
+import time
+from collections import OrderedDict
+from urllib.parse import urljoin, urlparse
+
+import aiohttp
+from bs4 import BeautifulSoup
 
 from open_webui.models.chats import ChatTitleMessagesForm
 from open_webui.config import DATA_DIR, ENABLE_ADMIN_EXPORT
@@ -121,3 +130,268 @@ async def download_db(user=Depends(get_admin_user)):
         media_type='application/octet-stream',
         filename='webui.db',
     )
+
+
+_FAVICON_CACHE: OrderedDict = OrderedDict()
+_FAVICON_CACHE_MAX = 1000
+_FAVICON_CACHE_TTL = 7 * 24 * 3600
+_FAVICON_TIMEOUT = aiohttp.ClientTimeout(total=5)
+_MAX_ICON_BYTES = 256 * 1024   # 256 KiB
+_MAX_HTML_BYTES = 250 * 1024   # 250 KiB
+_MAX_REDIRECTS = 3
+_ALLOWED_ICON_TYPES = {
+    'image/x-icon',
+    'image/vnd.microsoft.icon',
+    'image/png',
+    'image/jpeg',
+    'image/gif',
+    'image/webp',
+}
+_BADGE_PALETTE = [
+    '#e74c3c', '#e67e22', '#2ecc71', '#3498db',
+    '#9b59b6', '#1abc9c', '#e91e63', '#f39c12',
+]
+
+
+def _svg_badge(hostname: str) -> bytes:
+    if hostname:
+        letter = hostname[0].upper()
+        idx = hashlib.sha256(hostname.encode()).digest()[0] % len(_BADGE_PALETTE)
+        color = _BADGE_PALETTE[idx]
+    else:
+        letter = '?'
+        color = '#95a5a6'
+    svg = (
+        f'<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 32 32">'
+        f'<circle cx="16" cy="16" r="16" fill="{color}"/>'
+        f'<text x="16" y="21" font-size="16" font-family="sans-serif" '
+        f'fill="white" text-anchor="middle">{letter}</text>'
+        f'</svg>'
+    )
+    return svg.encode()
+
+
+def _normalize_hostname(url: str) -> str | None:
+    try:
+        if '://' not in url:
+            url = 'https://' + url
+        parsed = urlparse(url)
+        hostname = (parsed.hostname or '').lower().rstrip('.')
+        if not hostname:
+            return None
+        try:
+            hostname = hostname.encode('idna').decode('ascii')
+        except (UnicodeError, UnicodeDecodeError):
+            pass
+        return hostname or None
+    except Exception:
+        return None
+
+
+def _is_global_host(hostname: str) -> bool:
+    """Return True only if every resolved IP for hostname is globally routable."""
+    try:
+        addrs = socket.getaddrinfo(hostname, None)
+        if not addrs:
+            return False
+        for _, _, _, _, sockaddr in addrs:
+            if not ipaddress.ip_address(sockaddr[0]).is_global:
+                return False
+        return True
+    except Exception:
+        return False
+
+
+def _is_safe_url(url: str) -> bool:
+    """Return True if url has an http/https scheme and resolves to a global IP."""
+    try:
+        parsed = urlparse(url)
+        if parsed.scheme not in ('http', 'https'):
+            return False
+        hostname = parsed.hostname
+        if not hostname:
+            return False
+        return _is_global_host(hostname)
+    except Exception:
+        return False
+
+
+async def _fetch_bounded(
+    session: aiohttp.ClientSession,
+    url: str,
+    max_bytes: int,
+    allowed_types: set[str] | None,
+) -> tuple[bytes, str] | None:
+    """GET with capped size and per-hop SSRF validation."""
+    current = url
+    for _ in range(_MAX_REDIRECTS + 1):
+        try:
+            async with session.get(
+                current, allow_redirects=False, timeout=_FAVICON_TIMEOUT
+            ) as resp:
+                if resp.status in (301, 302, 303, 307, 308):
+                    location = resp.headers.get('Location', '')
+                    if not location:
+                        return None
+                    next_url = urljoin(current, location)
+                    if not _is_safe_url(next_url):
+                        return None
+                    current = next_url
+                    continue
+
+                if resp.status != 200:
+                    return None
+
+                ct = resp.headers.get('Content-Type', '').split(';')[0].strip().lower()
+                if allowed_types is not None and ct not in allowed_types:
+                    return None
+
+                data = b''
+                async for chunk in resp.content.iter_chunked(8192):
+                    data += chunk
+                    if len(data) > max_bytes:
+                        return None
+                return data, ct
+        except Exception:
+            return None
+    return None
+
+
+async def _discover_favicon(hostname: str) -> tuple[bytes, str] | None:
+    if not _is_global_host(hostname):
+        return None
+
+    async with aiohttp.ClientSession() as session:
+        # try /favicon.ico
+        result = await _fetch_bounded(
+            session,
+            f'https://{hostname}/favicon.ico',
+            _MAX_ICON_BYTES,
+            _ALLOWED_ICON_TYPES,
+        )
+        if result:
+            return result
+
+        # fall back to HTML <link> discovery
+        html_result = await _fetch_bounded(
+            session,
+            f'https://{hostname}/',
+            _MAX_HTML_BYTES,
+            {'text/html'},
+        )
+        if not html_result:
+            return None
+
+        html_bytes, _ = html_result
+        try:
+            soup = BeautifulSoup(html_bytes, 'html.parser')
+        except Exception:
+            return None
+
+        candidates: list[tuple[int, int, str]] = []
+        for tag in soup.find_all('link'):
+            rel = tag.get('rel') or []
+            if isinstance(rel, str):
+                rel = [rel]
+            rel_lower = [r.lower() for r in rel]
+
+            href = tag.get('href') or ''
+            if not href or href.startswith('data:'):
+                continue
+            tag_type = (tag.get('type') or '').lower()
+            if 'svg' in tag_type:
+                continue
+
+            if 'apple-touch-icon' in rel_lower and 'precomposed' not in rel_lower:
+                priority, size = 0, 180
+            elif 'icon' in rel_lower:
+                priority = 1
+                size = 0
+                sizes_str = (tag.get('sizes') or '').strip()
+                if sizes_str and sizes_str.lower() != 'any':
+                    try:
+                        size = max(
+                            int(p.lower().split('x')[0])
+                            for p in sizes_str.split()
+                            if 'x' in p.lower()
+                        )
+                    except Exception:
+                        size = 0
+            elif 'shortcut' in rel_lower:
+                priority, size = 2, 0
+            elif 'apple-touch-icon' in rel_lower:  # precomposed
+                priority, size = 3, 0
+            else:
+                continue
+
+            full_href = urljoin(f'https://{hostname}/', href)
+            candidates.append((priority, -size, full_href))
+
+        candidates.sort(key=lambda x: (x[0], x[1]))
+
+        for _, _, href in candidates:
+            if not _is_safe_url(href):
+                continue
+            icon_result = await _fetch_bounded(
+                session, href, _MAX_ICON_BYTES, _ALLOWED_ICON_TYPES
+            )
+            if icon_result:
+                return icon_result
+
+    return None
+
+
+def _cache_get(hostname: str) -> tuple[bytes, str] | None:
+    entry = _FAVICON_CACHE.get(hostname)
+    if entry is None:
+        return None
+    content, ct, expires_at = entry
+    if time.monotonic() > expires_at:
+        del _FAVICON_CACHE[hostname]
+        return None
+    _FAVICON_CACHE.move_to_end(hostname)
+    return content, ct
+
+
+def _cache_set(hostname: str, content: bytes, ct: str) -> None:
+    if hostname in _FAVICON_CACHE:
+        _FAVICON_CACHE.move_to_end(hostname)
+    _FAVICON_CACHE[hostname] = (content, ct, time.monotonic() + _FAVICON_CACHE_TTL)
+    while len(_FAVICON_CACHE) > _FAVICON_CACHE_MAX:
+        _FAVICON_CACHE.popitem(last=False)
+
+
+_FAVICON_RESPONSE_HEADERS = {
+    'Cache-Control': 'public, max-age=604800',
+    'X-Content-Type-Options': 'nosniff',
+}
+
+
+@router.get('/favicon')
+async def get_favicon(url: str = ''):
+    """Favicon proxy; always returns a response."""
+    hostname = _normalize_hostname(url)
+
+    if hostname:
+        cached = _cache_get(hostname)
+        if cached:
+            content, ct = cached
+            return Response(content=content, media_type=ct, headers=_FAVICON_RESPONSE_HEADERS)
+
+    icon: tuple[bytes, str] | None = None
+    if hostname:
+        try:
+            icon = await _discover_favicon(hostname)
+        except Exception:
+            pass
+
+    if icon:
+        content, ct = icon
+        _cache_set(hostname, content, ct)
+        return Response(content=content, media_type=ct, headers=_FAVICON_RESPONSE_HEADERS)
+
+    content = _svg_badge(hostname or '')
+    ct = 'image/svg+xml'
+    if hostname:
+        _cache_set(hostname, content, ct)
+    return Response(content=content, media_type=ct, headers=_FAVICON_RESPONSE_HEADERS)

src/lib/components/chat/Messages/Citations.svelte

@@ -1,5 +1,6 @@
 <script lang="ts">
 	import { getContext } from 'svelte';
+	import { WEBUI_API_BASE_URL } from '$lib/constants';
 	import { embed, showControls, showEmbeds } from '$lib/stores';
 
 	import CitationModal from './Citations/CitationModal.svelte';
@@ -175,7 +176,7 @@
 				<div class="flex -space-x-1 items-center">
 					{#each urlCitations.slice(0, 3) as citation, idx}
 						<img
-							src="https://www.google.com/s2/favicons?sz=32&domain={citation.source.name}"
+							src={`${WEBUI_API_BASE_URL}/utils/favicon?url=${encodeURIComponent(citation.source.name)}`}
 							alt="favicon"
 							class="size-4 rounded-full shrink-0 border border-white dark:border-gray-850 bg-white dark:bg-gray-900"
 							on:error={(e) => {

src/lib/components/chat/Messages/ResponseMessage/WebSearchResults.svelte

@@ -1,4 +1,6 @@
 <script lang="ts">
+	import { WEBUI_API_BASE_URL } from '$lib/constants';
+
 	import ChevronDown from '$lib/components/icons/ChevronDown.svelte';
 	import ChevronUp from '$lib/components/icons/ChevronUp.svelte';
 	import Search from '$lib/components/icons/Search.svelte';
@@ -66,7 +68,7 @@
 					<div class=" flex justify-center items-center gap-3">
 						<div class="w-fit">
 							<img
-								src="https://www.google.com/s2/favicons?sz=32&domain={item.link}"
+								src={`${WEBUI_API_BASE_URL}/utils/favicon?url=${encodeURIComponent(item.link)}`}
 								alt="{item?.title ?? item.link} favicon"
 								class="size-3.5"
 							/>
@@ -106,7 +108,7 @@
 					<div class=" flex justify-center items-center gap-3">
 						<div class="w-fit">
 							<img
-								src="https://www.google.com/s2/favicons?sz=32&domain={url}"
+								src={`${WEBUI_API_BASE_URL}/utils/favicon?url=${encodeURIComponent(url)}`}
 								alt="{url} favicon"
 								class="size-3.5"
 							/>