bug-fixes

Dockerfile

@@ -8,6 +8,8 @@ COPY . .
 # If ARG is not provided, ENV will be empty and config.ts will use defaults
 ARG VITE_DEFAULT_RELAYS
 ARG VITE_THREAD_TIMEOUT_DAYS
+ARG BUILD_DATE
+# BUILD_DATE is used to force rebuilds - can be set to current timestamp
 ENV VITE_DEFAULT_RELAYS=${VITE_DEFAULT_RELAYS}
 ENV VITE_THREAD_TIMEOUT_DAYS=${VITE_THREAD_TIMEOUT_DAYS}
 RUN npm run build

README.md

@@ -2,7 +2,7 @@
 
 A decentralized messageboard built on the Nostr protocol.
 
-![Aitherboard logo](/static/og-image.png)
+![Aitherboard logo](/og-image.png)
 
 **About**: [https://aitherboard.imwald.eu/about](https://aitherboard.imwald.eu/about)
 

docker-compose.yml

@@ -1,16 +1,19 @@
 services:
   aitherboard:
     container_name: aitherboard
-    # Using pre-built image (recommended for production)
-    image: silberengel/aitherboard:latest
-    # Alternative: Build from source (for development)
-    # build:
-    #   context: .
-    #   # Optional: override defaults from config.ts if needed
-    #   # Uncomment and modify if you want custom values:
-    #   # args:
-    #   #   VITE_DEFAULT_RELAYS: "wss://theforest.nostr1.com,wss://nostr21.com,wss://nostr.land"
-    #   #   VITE_THREAD_TIMEOUT_DAYS: "30"
+    # Build from source (always rebuilds - use --no-cache flag when running)
+    build:
+      context: .
+      dockerfile: Dockerfile
+      # Optional: override defaults from config.ts if needed
+      # Uncomment and modify if you want custom values:
+      # args:
+      #   VITE_DEFAULT_RELAYS: "wss://theforest.nostr1.com,wss://nostr21.com,wss://nostr.land"
+      #   VITE_THREAD_TIMEOUT_DAYS: "30"
+      # Note: To force rebuild: docker-compose build --no-cache aitherboard && docker-compose up -d
+    # Alternative: Using pre-built image (for production)
+    # Uncomment the image line below and comment out the build section above
+    # image: silberengel/aitherboard:latest
     ports:
       - "9876:9876"
     environment:

docker-rebuild.sh

@@ -0,0 +1,9 @@
+#!/bin/bash
+# Script to rebuild aitherboard container without cache
+# Usage: ./docker-rebuild.sh
+
+echo "Rebuilding aitherboard container (no cache)..."
+docker-compose build --no-cache aitherboard
+echo "Starting containers..."
+docker-compose up -d
+echo "Done! aitherboard has been rebuilt and restarted."

public/healthz.json

@@ -2,7 +2,7 @@
   "status": "ok",
   "service": "aitherboard",
   "version": "0.3.1",
-  "buildTime": "2026-02-13T04:20:13.671Z",
+  "buildTime": "2026-02-14T07:05:05.090Z",
   "gitCommit": "unknown",
-  "timestamp": 1770956413671
+  "timestamp": 1771052705090
 }

src/lib/components/content/FileExplorer.svelte

@@ -105,7 +105,8 @@
           const apiBase = baseHost.includes('gitlab.com') 
             ? 'https://gitlab.com/api/v4'
             : `${baseHost}/api/v4`;
-          apiUrl = `${apiBase}/projects/${projectPath}/repository/files/${encodeURIComponent(file.path)}/raw?ref=${repoInfo.defaultBranch}`;
+          // Use proxy endpoint to avoid CORS issues
+          apiUrl = `/api/gitea-proxy/projects/${projectPath}/repository/files/${encodeURIComponent(file.path)}/raw?baseUrl=${encodeURIComponent(apiBase)}&ref=${repoInfo.defaultBranch}`;
         }
       } else if (url.includes('onedev')) {
         // OneDev API: similar to Gitea, uses /api/v1/repos/{owner}/{repo}/contents/{path}
@@ -113,7 +114,9 @@
         if (match) {
           const [, baseUrl, owner, repo] = match;
           const cleanRepo = repo.replace(/\.git$/, '');
-          apiUrl = `${baseUrl}/api/v1/repos/${owner}/${cleanRepo}/contents/${file.path}?ref=${repoInfo.defaultBranch}`;
+          const baseApiUrl = `${baseUrl}/api/v1`;
+          // Use proxy endpoint to avoid CORS issues
+          apiUrl = `/api/gitea-proxy/repos/${owner}/${cleanRepo}/contents/${file.path}?baseUrl=${encodeURIComponent(baseApiUrl)}&ref=${repoInfo.defaultBranch}`;
         }
       } else {
         // Try Gitea/Forgejo pattern (handles .git suffix) - generic fallback
@@ -121,7 +124,9 @@
         if (match) {
           const [, baseUrl, owner, repo] = match;
           const cleanRepo = repo.replace(/\.git$/, '');
-          apiUrl = `${baseUrl}/api/v1/repos/${owner}/${cleanRepo}/contents/${file.path}?ref=${repoInfo.defaultBranch}`;
+          const baseApiUrl = `${baseUrl}/api/v1`;
+          // Use proxy endpoint to avoid CORS issues
+          apiUrl = `/api/gitea-proxy/repos/${owner}/${cleanRepo}/contents/${file.path}?baseUrl=${encodeURIComponent(baseApiUrl)}&ref=${repoInfo.defaultBranch}`;
         }
       }
 

src/lib/components/content/MarkdownRenderer.svelte

@@ -520,6 +520,7 @@
     }
     
     // 3. Match plain media URLs (using the SAME pattern as getMediaAttachmentUrls)
+    // IMPORTANT: This must match the exact same regex pattern used in FeedPost.mediaAttachmentUrls
     const urlRegex = /(https?:\/\/[^\s<>"{}|\\^`\[\]]+\.(jpg|jpeg|png|gif|webp|svg|bmp|mp4|webm|ogg|mov|avi|mkv|mp3|wav|flac|aac|m4a)(\?[^\s<>"{}|\\^`\[\]]*)?)/gi;
     urlRegex.lastIndex = 0;
     while ((match = urlRegex.exec(text)) !== null) {
@@ -540,15 +541,16 @@
       }
       
       const normalizedUrl = normalizeUrl(url);
+      // Check exclusion - this is critical for preventing duplicate images
       if (normalizedExcludeUrls.has(normalizedUrl)) {
-        // Remove excluded plain URLs
+        // Remove excluded plain URLs - they're already displayed by MediaAttachments
         if (typeof console !== 'undefined') {
-          console.debug('MarkdownRenderer: Found excluded URL, removing:', url, 'normalized:', normalizedUrl);
+          console.debug('MarkdownRenderer: Removing excluded URL from content:', url, 'normalized:', normalizedUrl);
         }
         replacements.push({ fullMatch, replacement: '', index });
       } else {
         if (typeof console !== 'undefined' && normalizedExcludeUrls.size > 0) {
-          console.debug('MarkdownRenderer: URL not excluded, will convert:', url, 'normalized:', normalizedUrl, 'excluded set:', Array.from(normalizedExcludeUrls));
+          console.debug('MarkdownRenderer: URL not in exclusion list, will convert to HTML:', url, 'normalized:', normalizedUrl, 'excluded set:', Array.from(normalizedExcludeUrls));
         }
         // Convert non-excluded plain URLs to HTML tags
         const ext = match[2].toLowerCase();
@@ -767,7 +769,9 @@
     const normalizedExcludeUrls = new Set(excludeMediaUrls.map(url => normalizeUrl(url)));
     let filtered = htmlContent;
     
-    // Remove ALL img tags with excluded URLs
+    // Remove ALL img tags with excluded URLs (aggressive cleanup)
+    // This is critical for kind 11 events to prevent duplicate images
+    const beforeCount = (filtered.match(/<img[^>]*>/gi) || []).length;
     filtered = filtered.replace(/<img[^>]*>/gi, (match) => {
       const srcMatch = match.match(/src=["']([^"']+)["']/i);
       if (srcMatch) {
@@ -775,13 +779,17 @@
         const normalizedSrc = normalizeUrl(src);
         if (normalizedExcludeUrls.has(normalizedSrc)) {
           if (typeof console !== 'undefined') {
-            console.debug('MarkdownRenderer: Removing excluded img tag from content:', src);
+            console.debug('MarkdownRenderer: Removing excluded img tag from HTML:', src, 'normalized:', normalizedSrc, 'excluded set:', Array.from(normalizedExcludeUrls));
           }
-          return '';
+          return ''; // Remove the img tag completely
         }
       }
       return match;
     });
+    const afterCount = (filtered.match(/<img[^>]*>/gi) || []).length;
+    if (typeof console !== 'undefined' && beforeCount !== afterCount) {
+      console.debug('MarkdownRenderer: Exclusion filtering removed', beforeCount - afterCount, 'img tags');
+    }
     
     // Remove links pointing to excluded image URLs
     filtered = filtered.replace(/<a\s+([^>]*?)href=["']([^"']+)["']([^>]*?)>([^<]*?)<\/a>/gi, (match, beforeAttrs, url, afterAttrs, linkText) => {

src/lib/modules/discussions/DiscussionCard.svelte

@@ -70,6 +70,77 @@
   let showReplyForm = $state(false);
   let isLoggedIn = $derived(sessionManager.isLoggedIn());
 
+  // Normalize URL for comparison (same logic as MediaAttachments and MarkdownRenderer)
+  function normalizeUrl(url: string): string {
+    if (!url || typeof url !== 'string') return url;
+    try {
+      const parsed = new URL(url);
+      const normalized = `${parsed.protocol}//${parsed.host.toLowerCase()}${parsed.pathname}`.replace(/\/$/, '');
+      return normalized;
+    } catch {
+      return url.trim().replace(/\/$/, '').toLowerCase();
+    }
+  }
+
+  // Get media URLs that MediaAttachments will display (for fullView)
+  // This matches the logic in MediaAttachments.extractMedia() and FeedPost.mediaAttachmentUrls
+  const mediaAttachmentUrls = $derived.by(() => {
+    const urls: string[] = [];
+    const seen = new Set<string>();
+    const forceRender = isMediaKind;
+
+    // 1. Image tag (NIP-23)
+    const imageTag = thread.tags.find((t) => t[0] === 'image');
+    if (imageTag && imageTag[1]) {
+      const normalized = normalizeUrl(imageTag[1]);
+      if (!seen.has(normalized)) {
+        urls.push(imageTag[1]);
+        seen.add(normalized);
+      }
+    }
+
+    // 2. Extract from markdown content (images in markdown syntax)
+    const imageRegex = /!\[.*?\]\((.*?)\)/g;
+    let match;
+    while ((match = imageRegex.exec(thread.content)) !== null) {
+      const url = match[1];
+      const normalized = normalizeUrl(url);
+      if (!seen.has(normalized)) {
+        urls.push(url);
+        seen.add(normalized);
+      }
+    }
+
+    // 3. Extract from AsciiDoc content (images in AsciiDoc syntax)
+    const asciidocImageRegex = /image::?([^\s\[\]]+)(?:\[[^\]]*\])?/g;
+    asciidocImageRegex.lastIndex = 0;
+    let asciidocMatch;
+    while ((asciidocMatch = asciidocImageRegex.exec(thread.content)) !== null) {
+      const url = asciidocMatch[1];
+      if (url.startsWith('http://') || url.startsWith('https://')) {
+        const normalized = normalizeUrl(url);
+        if (!seen.has(normalized)) {
+          urls.push(url);
+          seen.add(normalized);
+        }
+      }
+    }
+
+    // 4. Extract plain image URLs from content (matching MediaAttachments logic)
+    const urlRegex = /(https?:\/\/[^\s<>"{}|\\^`\[\]]+\.(jpg|jpeg|png|gif|webp|svg|bmp|mp4|webm|ogg|mov|avi|mkv|mp3|wav|flac|aac|m4a)(\?[^\s<>"{}|\\^`\[\]]*)?)/gi;
+    urlRegex.lastIndex = 0;
+    while ((match = urlRegex.exec(thread.content)) !== null) {
+      const url = match[1];
+      const normalized = normalizeUrl(url);
+      if (!seen.has(normalized)) {
+        urls.push(url);
+        seen.add(normalized);
+      }
+    }
+
+    return urls;
+  });
+
   onMount(async () => {
     await loadStats();
   });
@@ -239,7 +310,7 @@
         {#if shouldAutoRenderMedia}
           <MediaAttachments event={thread} forceRender={isMediaKind} onMediaClick={handleMediaUrlClick} isFeedView={false} />
         {/if}
-        <MarkdownRenderer content={thread.content} event={thread} />
+        <MarkdownRenderer content={thread.content} event={thread} excludeMediaUrls={mediaAttachmentUrls} />
       </div>
       </div>
     </a>
@@ -289,7 +360,7 @@
 
       <div class="post-content mb-2">
         <MediaAttachments event={thread} forceRender={isMediaKind} onMediaClick={handleMediaUrlClick} />
-        <MarkdownRenderer content={thread.content} event={thread} />
+        <MarkdownRenderer content={thread.content} event={thread} excludeMediaUrls={mediaAttachmentUrls} />
       </div>
 
       {#if getTopics().length > 0}

src/lib/modules/feed/FeedPost.svelte

@@ -726,6 +726,7 @@
   // Get media URLs that MediaAttachments will display (for fullView)
   // This matches the logic in MediaAttachments.extractMedia()
   // Use $derived to ensure it updates reactively
+  // For kind 11 (discussion threads), we always extract ALL media URLs to ensure they're excluded from markdown
   const mediaAttachmentUrls = $derived.by(() => {
     const urls: string[] = [];
     const seen = new Set<string>();
@@ -997,7 +998,7 @@
         <PollCard pollEvent={post} />
       {:else if post.content && post.content.trim()}
         {#if typeof console !== 'undefined' && fullView}
-          {console.debug('FeedPost fullView: Media attachment URLs to exclude:', mediaAttachmentUrls, 'for post:', post.id)}
+          {console.debug('FeedPost fullView: Media attachment URLs to exclude:', mediaAttachmentUrls, 'for post:', post.id, 'kind:', post.kind)}
         {/if}
         <MarkdownRenderer content={post.content} event={post} excludeMediaUrls={mediaAttachmentUrls} />
       {:else if !isMediaKind && post.kind !== KIND.POLL}

src/lib/services/content/git-repo-fetcher.ts

@@ -320,11 +320,31 @@ async function fetchFromGitLab(owner: string, repo: string, baseUrl: string): Pr
     const projectPath = `${owner}/${repo}`;
     const encodedPath = encodeURIComponent(projectPath);
     
-    const [repoData, branchesData, commitsData] = await Promise.all([
-      fetch(`${baseUrl}/projects/${encodedPath}`).then(r => r.json()),
-      fetch(`${baseUrl}/projects/${encodedPath}/repository/branches`).then(r => r.json()),
-      fetch(`${baseUrl}/projects/${encodedPath}/repository/commits?per_page=10`).then(r => r.json())
+    // Use proxy endpoint to avoid CORS issues
+    const [repoResponse, branchesResponse, commitsResponse] = await Promise.all([
+      fetch(`/api/gitea-proxy/projects/${encodedPath}?baseUrl=${encodeURIComponent(baseUrl)}`),
+      fetch(`/api/gitea-proxy/projects/${encodedPath}/repository/branches?baseUrl=${encodeURIComponent(baseUrl)}`),
+      fetch(`/api/gitea-proxy/projects/${encodedPath}/repository/commits?baseUrl=${encodeURIComponent(baseUrl)}&per_page=10`)
     ]);
+    
+    if (!repoResponse.ok) {
+      console.warn(`GitLab API error for repo ${owner}/${repo}: ${repoResponse.status} ${repoResponse.statusText}`);
+      return null;
+    }
+    
+    const repoData = await repoResponse.json();
+    let branchesData: any[] = [];
+    let commitsData: any[] = [];
+    
+    if (branchesResponse.ok) {
+      const data = await branchesResponse.json();
+      branchesData = Array.isArray(data) ? data : [];
+    }
+    
+    if (commitsResponse.ok) {
+      const data = await commitsResponse.json();
+      commitsData = Array.isArray(data) ? data : [];
+    }
 
     const branches: GitBranch[] = branchesData.map((b: any) => ({
       name: b.name,
@@ -346,7 +366,7 @@ async function fetchFromGitLab(owner: string, repo: string, baseUrl: string): Pr
     // Fetch file tree
     let files: GitFile[] = [];
     try {
-      const treeData = await fetch(`${baseUrl}/projects/${encodedPath}/repository/tree?recursive=true&per_page=100`).then(r => r.json());
+      const treeData = await fetch(`/api/gitea-proxy/projects/${encodedPath}/repository/tree?baseUrl=${encodeURIComponent(baseUrl)}&recursive=true&per_page=100`).then(r => r.json());
       files = treeData.map((item: any) => ({
         name: item.name,
         path: item.path,
@@ -363,7 +383,7 @@ async function fetchFromGitLab(owner: string, repo: string, baseUrl: string): Pr
     const readmeFiles = ['README.adoc', 'README.md', 'README.rst', 'README.txt'];
     for (const readmeFile of readmeFiles) {
       try {
-        const fileData = await fetch(`${baseUrl}/projects/${encodedPath}/repository/files/${encodeURIComponent(readmeFile)}/raw?ref=${repoData.default_branch}`).then(r => {
+        const fileData = await fetch(`/api/gitea-proxy/projects/${encodedPath}/repository/files/${encodeURIComponent(readmeFile)}/raw?baseUrl=${encodeURIComponent(baseUrl)}&ref=${repoData.default_branch}`).then(r => {
           if (!r.ok) throw new Error('Not found');
           return r.text();
         });
@@ -398,7 +418,7 @@ async function fetchFromGitLab(owner: string, repo: string, baseUrl: string): Pr
       // If found in tree, fetch it
       if (readmePath) {
         try {
-          const fileData = await fetch(`${baseUrl}/projects/${encodedPath}/repository/files/${encodeURIComponent(readmePath)}/raw?ref=${repoData.default_branch}`).then(r => {
+          const fileData = await fetch(`/api/gitea-proxy/projects/${encodedPath}/repository/files/${encodeURIComponent(readmePath)}/raw?baseUrl=${encodeURIComponent(baseUrl)}&ref=${repoData.default_branch}`).then(r => {
             if (!r.ok) throw new Error('Not found');
             return r.text();
           });
@@ -436,8 +456,7 @@ async function fetchFromGitLab(owner: string, repo: string, baseUrl: string): Pr
 async function fetchFromGitea(owner: string, repo: string, baseUrl: string): Promise<GitRepoInfo | null> {
   try {
     // Use proxy endpoint to avoid CORS issues
-    const proxyBaseUrl = encodeURIComponent(baseUrl);
-    const repoResponse = await fetch(`/api/gitea-proxy/${proxyBaseUrl}/repos/${owner}/${repo}`);
+    const repoResponse = await fetch(`/api/gitea-proxy/repos/${owner}/${repo}?baseUrl=${encodeURIComponent(baseUrl)}`);
     if (!repoResponse.ok) {
       console.warn(`Gitea API error for repo ${owner}/${repo}: ${repoResponse.status} ${repoResponse.statusText}`);
       return null;
@@ -447,8 +466,8 @@ async function fetchFromGitea(owner: string, repo: string, baseUrl: string): Pro
     const defaultBranch = repoData.default_branch || 'master';
     
     const [branchesResponse, commitsResponse] = await Promise.all([
-      fetch(`/api/gitea-proxy/${proxyBaseUrl}/repos/${owner}/${repo}/branches`).catch(() => null),
-      fetch(`/api/gitea-proxy/${proxyBaseUrl}/repos/${owner}/${repo}/commits?limit=10`).catch(() => null)
+      fetch(`/api/gitea-proxy/repos/${owner}/${repo}/branches?baseUrl=${encodeURIComponent(baseUrl)}`).catch(() => null),
+      fetch(`/api/gitea-proxy/repos/${owner}/${repo}/commits?baseUrl=${encodeURIComponent(baseUrl)}&limit=10`).catch(() => null)
     ]);
     
     let branchesData: any[] = [];
@@ -501,7 +520,7 @@ async function fetchFromGitea(owner: string, repo: string, baseUrl: string): Pro
     let files: GitFile[] = [];
     try {
       // Try the git/trees endpoint first (more complete)
-      const treeResponse = await fetch(`/api/gitea-proxy/${proxyBaseUrl}/repos/${owner}/${repo}/git/trees/${defaultBranch}?recursive=1`).catch(() => null);
+      const treeResponse = await fetch(`/api/gitea-proxy/repos/${owner}/${repo}/git/trees/${defaultBranch}?baseUrl=${encodeURIComponent(baseUrl)}&recursive=1`).catch(() => null);
       if (treeResponse && treeResponse.ok) {
         const treeData = await treeResponse.json();
         if (treeData.tree && Array.isArray(treeData.tree)) {
@@ -516,7 +535,7 @@ async function fetchFromGitea(owner: string, repo: string, baseUrl: string): Pro
         }
       } else {
         // Fallback to contents endpoint (only root directory)
-        const contentsResponse = await fetch(`/api/gitea-proxy/${proxyBaseUrl}/repos/${owner}/${repo}/contents?ref=${defaultBranch}`).catch(() => null);
+        const contentsResponse = await fetch(`/api/gitea-proxy/repos/${owner}/${repo}/contents?baseUrl=${encodeURIComponent(baseUrl)}&ref=${defaultBranch}`).catch(() => null);
         if (contentsResponse && contentsResponse.ok) {
           const contentsData = await contentsResponse.json();
           if (Array.isArray(contentsData)) {
@@ -539,7 +558,7 @@ async function fetchFromGitea(owner: string, repo: string, baseUrl: string): Pro
     const readmeFiles = ['README.adoc', 'README.md', 'README.rst', 'README.txt'];
     for (const readmeFile of readmeFiles) {
       try {
-        const fileResponse = await fetch(`/api/gitea-proxy/${proxyBaseUrl}/repos/${owner}/${repo}/contents/${readmeFile}?ref=${defaultBranch}`);
+        const fileResponse = await fetch(`/api/gitea-proxy/repos/${owner}/${repo}/contents/${readmeFile}?baseUrl=${encodeURIComponent(baseUrl)}&ref=${defaultBranch}`);
         if (!fileResponse.ok) throw new Error('Not found');
         const fileData = await fileResponse.json();
         if (fileData.content) {
@@ -578,7 +597,7 @@ async function fetchFromGitea(owner: string, repo: string, baseUrl: string): Pro
       // If found in tree, fetch it
       if (readmePath) {
         try {
-          const fileResponse = await fetch(`/api/gitea-proxy/${proxyBaseUrl}/repos/${owner}/${repo}/contents/${readmePath}?ref=${defaultBranch}`);
+          const fileResponse = await fetch(`/api/gitea-proxy/repos/${owner}/${repo}/contents/${readmePath}?baseUrl=${encodeURIComponent(baseUrl)}&ref=${defaultBranch}`);
           if (!fileResponse.ok) throw new Error('Not found');
           const fileData = await fileResponse.json();
           if (fileData.content) {

src/routes/api/gitea-proxy/[...path]/+server.ts

@@ -1,35 +1,109 @@
 import type { RequestHandler } from '@sveltejs/kit';
 
 /**
- * Proxy endpoint for Gitea API requests to avoid CORS issues
- * Usage: /api/gitea-proxy/{baseUrl}/repos/{owner}/{repo}/...
- * Example: /api/gitea-proxy/https%3A%2F%2Fgit.imwald.eu/api/v1/repos/silberengel/aitherboard
+ * Proxy endpoint for Git hosting API requests (Gitea, GitLab, OneDev, etc.) to avoid CORS issues
+ * Usage: /api/gitea-proxy/{apiPath}?baseUrl={baseUrl}
+ * Examples:
+ *   - Gitea: /api/gitea-proxy/repos/silberengel/aitherboard/contents/README.adoc?baseUrl=https://git.imwald.eu/api/v1&ref=master
+ *   - GitLab: /api/gitea-proxy/projects/owner%2Frepo/repository/files/path/raw?baseUrl=https://gitlab.com/api/v4&ref=master
+ *   - OneDev: /api/gitea-proxy/repos/owner/repo/contents/README.adoc?baseUrl=https://onedev.example.com/api/v1&ref=master
  */
-export const GET: RequestHandler = async ({ params, url }: { params: { path?: string }; url: URL }) => {
-  try {
-    // Reconstruct the full path from params
-    const pathParts = params.path?.split('/') || [];
+
+const CORS_HEADERS = {
+  'Access-Control-Allow-Origin': '*',
+  'Access-Control-Allow-Methods': 'GET, OPTIONS',
+  'Access-Control-Allow-Headers': 'Content-Type'
+} as const;
+
+function createErrorResponse(message: string, status: number): Response {
+  return new Response(JSON.stringify({ error: message }), {
+    status,
+    headers: {
+      'Content-Type': 'application/json',
+      ...CORS_HEADERS
+    }
+  });
+}
+
+function buildTargetUrl(baseUrl: string, apiPath: string, searchParams: URLSearchParams): string {
+  // Ensure baseUrl doesn't have a trailing slash
+  const cleanBaseUrl = baseUrl.endsWith('/') ? baseUrl.slice(0, -1) : baseUrl;
+  
+  // For GitLab, both the project path and file paths need to be re-encoded
+  // SvelteKit splits URL-encoded paths into separate segments
+  let processedPath = apiPath;
+  if (apiPath.startsWith('projects/')) {
+    const parts = apiPath.split('/');
     
-    // Extract base URL (first part should be the encoded base URL)
-    if (pathParts.length < 1) {
-      return new Response(JSON.stringify({ error: 'Missing base URL' }), {
-        status: 400,
-        headers: { 'Content-Type': 'application/json' }
-      });
+    // GitLab project paths are: projects/{owner}/{repo}/...
+    // If we have at least 3 parts (projects, owner, repo), combine owner and repo
+    if (parts.length >= 3) {
+      const projectPath = `${parts[1]}/${parts[2]}`;
+      const encodedProjectPath = encodeURIComponent(projectPath);
+      
+      // Check if this is a file path: projects/{owner}/{repo}/repository/files/{file_path}/raw
+      const filesIndex = parts.indexOf('files');
+      if (filesIndex !== -1 && filesIndex < parts.length - 1) {
+        // Found /repository/files/, encode the file path (everything between 'files' and 'raw')
+        const filePathParts = parts.slice(filesIndex + 1, parts.length - 1); // Exclude 'raw' at the end
+        const filePath = filePathParts.join('/');
+        // GitLab API requires the file path to be URL-encoded
+        // encodeURIComponent will encode slashes as %2F, which is what GitLab expects
+        const encodedFilePath = encodeURIComponent(filePath);
+        
+        // Debug logging
+        console.log('[Gitea Proxy] File path parts:', filePathParts);
+        console.log('[Gitea Proxy] File path (joined):', filePath);
+        console.log('[Gitea Proxy] Encoded file path:', encodedFilePath);
+        
+        // Reconstruct: projects/{encodedProjectPath}/repository/files/{encodedFilePath}/raw
+        // Rebuild the path up to 'files', then add encoded file path and 'raw'
+        const beforeFiles = `projects/${encodedProjectPath}/repository/files`;
+        processedPath = `${beforeFiles}/${encodedFilePath}/${parts[parts.length - 1]}`;
+      } else {
+        // Not a file path, just re-encode the project path
+        processedPath = `projects/${encodedProjectPath}${parts.length > 3 ? '/' + parts.slice(3).join('/') : ''}`;
+      }
+    }
+  }
+  
+  // Ensure processedPath starts with a slash
+  const cleanApiPath = processedPath.startsWith('/') ? processedPath : `/${processedPath}`;
+  
+  // Construct query string (excluding baseUrl)
+  const queryParts: string[] = [];
+  for (const [key, value] of searchParams.entries()) {
+    if (key !== 'baseUrl') {
+      queryParts.push(`${encodeURIComponent(key)}=${encodeURIComponent(value)}`);
     }
+  }
+  const queryString = queryParts.length > 0 ? `?${queryParts.join('&')}` : '';
+  
+  // Construct the full URL manually to preserve encoding
+  // Note: We construct as string because new URL() with pathname assignment would decode %2F
+  return `${cleanBaseUrl}${cleanApiPath}${queryString}`;
+}
 
-    // Decode the base URL (it's URL-encoded)
-    const baseUrl = decodeURIComponent(pathParts[0]);
+export const GET: RequestHandler = async ({ params, url }) => {
+  try {
+    const baseUrl = url.searchParams.get('baseUrl');
+    const apiPath = params.path;
     
-    // Reconstruct the API path (everything after the base URL)
-    const apiPath = pathParts.slice(1).join('/');
+    if (!baseUrl) {
+      return createErrorResponse('Missing baseUrl query parameter', 400);
+    }
     
-    // Add query parameters from the original request
-    const queryString = url.search;
-    const fullUrl = `${baseUrl}/${apiPath}${queryString}`;
+    if (!apiPath) {
+      return createErrorResponse('Missing API path', 400);
+    }
 
-    // Fetch from Gitea API
-    const response = await fetch(fullUrl, {
+    const targetUrl = buildTargetUrl(baseUrl, apiPath, url.searchParams);
+    
+    // Debug logging (remove in production if needed)
+    console.log('[Gitea Proxy] Original path:', apiPath);
+    console.log('[Gitea Proxy] Target URL:', targetUrl);
+    
+    const response = await fetch(targetUrl, {
       method: 'GET',
       headers: {
         'Accept': 'application/json',
@@ -37,41 +111,33 @@ export const GET: RequestHandler = async ({ params, url }: { params: { path?: st
       }
     });
 
-    // Get response body
     const contentType = response.headers.get('content-type') || 'application/json';
     const body = await response.text();
+    
+    // Log error responses for debugging
+    if (!response.ok) {
+      console.error('[Gitea Proxy] Error response:', response.status, response.statusText);
+      console.error('[Gitea Proxy] Response body:', body.substring(0, 500));
+    }
 
-    // Return response with CORS headers
     return new Response(body, {
       status: response.status,
       statusText: response.statusText,
       headers: {
         'Content-Type': contentType,
-        'Access-Control-Allow-Origin': '*',
-        'Access-Control-Allow-Methods': 'GET, OPTIONS',
-        'Access-Control-Allow-Headers': 'Content-Type'
+        ...CORS_HEADERS
       }
     });
   } catch (error) {
     const message = error instanceof Error ? error.message : 'Unknown error';
-    console.error('Gitea proxy error:', message);
-    return new Response(JSON.stringify({ error: message }), {
-      status: 500,
-      headers: {
-        'Content-Type': 'application/json',
-        'Access-Control-Allow-Origin': '*'
-      }
-    });
+    console.error('Git hosting proxy error:', message);
+    return createErrorResponse(message, 500);
   }
 };
 
-export const OPTIONS: RequestHandler = async (): Promise<Response> => {
+export const OPTIONS: RequestHandler = async () => {
   return new Response(null, {
     status: 204,
-    headers: {
-      'Access-Control-Allow-Origin': '*',
-      'Access-Control-Allow-Methods': 'GET, OPTIONS',
-      'Access-Control-Allow-Headers': 'Content-Type'
-    }
+    headers: CORS_HEADERS
   });
 };

svelte.config.js

@@ -15,7 +15,16 @@ const config = {
       precompress: true
     }),
     prerender: {
-      handleUnseenRoutes: 'ignore'
+      handleUnseenRoutes: 'ignore',
+      handleHttpError: ({ path, referrer, message }) => {
+        // Ignore 404s for static assets during prerendering
+        // These will be available at runtime from the public directory
+        if (path === '/favicon.ico' || path === '/og-image.png') {
+          return;
+        }
+        // For other errors, throw to fail the build
+        throw new Error(`${message} (${path})`);
+      }
     }
   }
 };

vite.config.ts

@@ -5,6 +5,7 @@ import { SvelteKitPWA } from '@vite-pwa/sveltekit';
 import compression from 'vite-plugin-compression';
 
 export default defineConfig({
+  publicDir: 'public',
   plugins: [
     sveltekit(),
     compression({
@@ -115,7 +116,8 @@ export default defineConfig({
         ]
       },
       devOptions: {
-        enabled: false
+        enabled: false,
+        suppressWarnings: true  // Suppress service worker warnings in dev mode
       }
     }),
     {
@@ -131,7 +133,11 @@ export default defineConfig({
   ],
   server: {
     port: 5173,
-    strictPort: false
+    strictPort: false,
+    fs: {
+      // Allow serving files from the project root (including static directory if needed)
+      allow: ['..']
+    }
   },
   build: {
     target: 'esnext',