feat: add architecture sync script and cheerio dependency

- Added sync:arch npm script to synchronize architecture documentation
- Installed cheerio for HTML parsing capabilities
- Alphabetized package.json dependencies for better maintainability

package.json

@@ -9,38 +9,40 @@
     "lint": "next lint",
     "type-check": "tsc --noEmit",
     "test:db": "node scripts/test-db.js",
-    "init:backend": "chmod +x scripts/init-backend.sh && ./scripts/init-backend.sh"
+    "init:backend": "chmod +x scripts/init-backend.sh && ./scripts/init-backend.sh",
+    "sync:arch": "npx tsx scripts/sync-arch.ts"
   },
   "dependencies": {
-    "next": "14.0.4",
-    "react": "^18",
-    "react-dom": "^18",
-    "lucide-react": "^0.294.0",
-    "@radix-ui/react-select": "^2.0.0",
     "@radix-ui/react-checkbox": "^1.0.4",
     "@radix-ui/react-dialog": "^1.0.5",
-    "@radix-ui/react-tabs": "^1.0.4",
     "@radix-ui/react-label": "^2.0.2",
+    "@radix-ui/react-select": "^2.0.0",
     "@radix-ui/react-slot": "^1.0.2",
+    "@radix-ui/react-tabs": "^1.0.4",
+    "cheerio": "^1.1.2",
     "class-variance-authority": "^0.7.0",
     "clsx": "^2.0.0",
+    "lucide-react": "^0.294.0",
+    "next": "14.0.4",
+    "pg": "^8.11.3",
+    "react": "^18",
+    "react-dom": "^18",
+    "react-query": "^3.39.3",
     "tailwind-merge": "^2.1.0",
     "tailwindcss-animate": "^1.0.7",
-    "react-query": "^3.39.3",
+    "undici": "^6.6.2",
-    "zustand": "^4.4.7",
+    "zustand": "^4.4.7"
-    "pg": "^8.11.3",
-    "undici": "^6.6.2"
   },
   "devDependencies": {
-    "typescript": "^5",
     "@types/node": "^20",
+    "@types/pg": "^8.10.9",
     "@types/react": "^18",
     "@types/react-dom": "^18",
-    "@types/pg": "^8.10.9",
     "autoprefixer": "^10.0.1",
+    "eslint": "^8",
+    "eslint-config-next": "14.0.4",
     "postcss": "^8",
     "tailwindcss": "^3.3.0",
-    "eslint": "^8",
+    "typescript": "^5"
-    "eslint-config-next": "14.0.4"
   }
 }

pnpm-lock.yaml

@@ -26,6 +26,9 @@ importers:
       '@radix-ui/react-tabs':
         specifier: ^1.0.4
         version: 1.1.13(@types/[email protected](@types/[email protected]))(@types/[email protected])([email protected]([email protected]))([email protected])
+      cheerio:
+        specifier: ^1.1.2
+        version: 1.1.2
       class-variance-authority:
         specifier: ^0.7.0
         version: 0.7.1
@@ -899,6 +902,9 @@ packages:
     resolution: {integrity: sha512-Ceh+7ox5qe7LJuLHoY0feh3pHuUDHAcRUeyL2VYghZwfpkNIy/+8Ocg0a3UuSoYzavmylwuLWQOf3hl0jjMMIw==}
     engines: {node: '>=8'}
 
+  [email protected]:
+    resolution: {integrity: sha512-JZOSA7Mo9sNGB8+UjSgzdLtokWAky1zbztM3WRLCbZ70/3cTANmQmOdR7y2g+J0e2WXywy1yS468tY+IruqEww==}
+
   [email protected]:
     resolution: {integrity: sha512-9T9UjW3r0UW5c1Q7GTwllptXwhvYmEzFhzMfZ9H7FQWt+uZePjZPjBP/W1ZEyZ1twGWom5/56TF4lPcqjnDHcg==}
 
@@ -948,6 +954,13 @@ packages:
     resolution: {integrity: sha512-oKnbhFyRIXpUuez8iBMmyEa4nbj4IOQyuhc/wy9kY7/WVPcwIO9VA668Pu8RkO7+0G76SLROeyw9CpQ061i4mA==}
     engines: {node: '>=10'}
 
+  [email protected]:
+    resolution: {integrity: sha512-9v9kG0LvzrlcungtnJtpGNxY+fzECQKhK4EGJX2vByejiMX84MFNQw4UxPJl3bFbTMw+Dfs37XaIkCwTZfLh4g==}
+
+  [email protected]:
+    resolution: {integrity: sha512-IkxPpb5rS/d1IiLbHMgfPuS0FgiWTtFIm/Nj+2woXDLTZ7fOT2eqzgYbdMlLweqlHbsZjxEChoVK+7iph7jyQg==}
+    engines: {node: '>=20.18.1'}
+
   [email protected]:
     resolution: {integrity: sha512-7VT13fmjotKpGipCW9JEQAusEPE+Ei8nl6/g4FBAmIm0GOOLMua9NDDo/DWp0ZAxCr3cPq5ZpBqmPAQgDda2Pw==}
     engines: {node: '>= 8.10.0'}
@@ -980,6 +993,13 @@ packages:
     resolution: {integrity: sha512-uV2QOWP2nWzsy2aMp8aRibhi9dlzF5Hgh5SHaB9OiTGEyDTiJJyx0uy51QXdyWbtAHNua4XJzUKca3OzKUd3vA==}
     engines: {node: '>= 8'}
 
+  [email protected]:
+    resolution: {integrity: sha512-TizTzUddG/xYLA3NXodFM0fSbNizXjOKhqiQQwvhlspadZokn1KDy0NZFS0wuEubIYAV5/c1/lAr0TaaFXEXzw==}
+
+  [email protected]:
+    resolution: {integrity: sha512-u/O3vwbptzhMs3L1fQE82ZSLHQQfto5gyZzwteVIEyeaY5Fc7R4dapF/BvRoSYFeqfBk4m0V1Vafq5Pjv25wvA==}
+    engines: {node: '>= 6'}
+
   [email protected]:
     resolution: {integrity: sha512-/Tb/JcjK111nNScGob5MNtsntNM1aCNUDipB/TkwZFhyDrrE47SOx/18wF2bbjgc3ZzCSKW1T5nt5EbFoAz/Vg==}
     engines: {node: '>=4'}
@@ -1055,6 +1075,19 @@ packages:
     resolution: {integrity: sha512-yS+Q5i3hBf7GBkd4KG8a7eBNNWNGLTaEwwYWUijIYM7zrlYDM0BFXHjjPWlWZ1Rg7UaddZeIDmi9jF3HmqiQ2w==}
     engines: {node: '>=6.0.0'}
 
+  [email protected]:
+    resolution: {integrity: sha512-wIkAryiqt/nV5EQKqQpo3SToSOV9J0DnbJqwK7Wv/Trc92zIAYZ4FlMu+JPFW1DfGFt81ZTCGgDEabffXeLyJg==}
+
+  [email protected]:
+    resolution: {integrity: sha512-OLETBj6w0OsagBwdXnPdN0cnMfF9opN69co+7ZrbfPGrdpPVNBUj02spi6B1N7wChLQiPn4CSH/zJvXw56gmHw==}
+
+  [email protected]:
+    resolution: {integrity: sha512-cgwlv/1iFQiFnU96XXgROh8xTeetsnJiDsTc7TYCLFd9+/WNkIqPTxiM/8pSd8VIrhXGTf1Ny1q1hquVqDJB5w==}
+    engines: {node: '>= 4'}
+
+  [email protected]:
+    resolution: {integrity: sha512-6kZKyUajlDuqlHKVX1w7gyslj9MPIXzIFiz/rGu35uC1wMi+kMhQwGhl4lt9unC9Vb9INnY9Z3/ZA3+FhASLaw==}
+
   [email protected]:
     resolution: {integrity: sha512-KIN/nDJBQRcXw0MLVhZE9iQHmG68qAVIBg9CqmUYjmQIhgij9U5MFvrqkUL5FbtyyzZuOeOt0zdeRe4UY7ct+A==}
     engines: {node: '>= 0.4'}
@@ -1071,6 +1104,17 @@ packages:
   [email protected]:
     resolution: {integrity: sha512-L18DaJsXSUk2+42pv8mLs5jJT2hqFkFE4j21wOmgbUqsZ2hL72NsUU785g9RXgo3s0ZNgVl42TiHp3ZtOv/Vyg==}
 
+  [email protected]:
+    resolution: {integrity: sha512-5gvq20T6vfpekVtqrYQsSCFZ1wEg5+wW0/QaZMWkFr6BqD3NfKs0rLCx4rrVlSWJeZb5NBJgVLswK/w2MWU+Gw==}
+
+  [email protected]:
+    resolution: {integrity: sha512-V0hjH4dGPh9Ao5p0MoRY6BVqtwCjhz6vI5LT8AJ55H+4g9/4vbHx1I54fS0XuclLhDHArPQCiMjDxjaL8fPxhw==}
+    engines: {node: '>=0.12'}
+
+  [email protected]:
+    resolution: {integrity: sha512-aN97NXWF6AWBTahfVOIrB/NShkzi5H7F9r1s9mD3cDj4Ko5f2qhhVoYMibXF7GlLveb/D2ioWay8lxI97Ven3g==}
+    engines: {node: '>=0.12'}
+
   [email protected]:
     resolution: {integrity: sha512-WSzPgsdLtTcQwm4CROfS5ju2Wa1QQcVeT37jFjYzdFz1r9ahadC8B8/a4qxJxM+09F18iumCdRmlr96ZYkQvEg==}
     engines: {node: '>= 0.4'}
@@ -1387,6 +1431,13 @@ packages:
     resolution: {integrity: sha512-0hJU9SCPvmMzIBdZFqNPXWa6dqh7WdH0cII9y+CyS8rG3nL48Bclra9HmKhVVUHyPWNH5Y7xDwAB7bfgSjkUMQ==}
     engines: {node: '>= 0.4'}
 
+  [email protected]:
+    resolution: {integrity: sha512-TwAZM+zE5Tq3lrEHvOlvwgj1XLWQCtaaibSN11Q+gGBAS7Y1uZSWwXXRe4iF6OXnaq1riyQAPFOBtYc77Mxq0g==}
+
+  [email protected]:
+    resolution: {integrity: sha512-4fCk79wshMdzMp2rH06qWrJE4iolqLhCUH+OiuIgU++RB0+94NlDL81atO7GX55uUKueo0txHNtvEyI6D7WdMw==}
+    engines: {node: '>=0.10.0'}
+
   [email protected]:
     resolution: {integrity: sha512-hsBTNUqQTDwkWtcdYI2i06Y/nUBEsNEDJKjWdigLvegy8kDuJAS8uRlpkkcQpyEXL0Z/pjDy5HBmMjRCJ2gq+g==}
     engines: {node: '>= 4'}
@@ -1693,6 +1744,9 @@ packages:
     resolution: {integrity: sha512-bdok/XvKII3nUpklnV6P2hxtMNrCboOjAcyBuQnWEhO665FwrSNRxU+AqpsyvO6LgGYPspN+lu5CLtw4jPRKNA==}
     engines: {node: '>=0.10.0'}
 
+  [email protected]:
+    resolution: {integrity: sha512-lqjrjmaOoAnWfMmBPL+XNnynZh2+swxiX3WUE0s4yEHI6m+AwrK2UZOimIRl3X/4QctVqS8AiZjFqyOGrMXb/w==}
+
   [email protected]:
     resolution: {integrity: sha512-rJgTQnkUnH1sFw8yT6VSU3zD3sWmu6sZhIseY8VX+GRu3P6F7Fu+JNDoXfklElbLJSnc3FUQHVe4cU5hj+BcUg==}
     engines: {node: '>=0.10.0'}
@@ -1758,6 +1812,15 @@ packages:
     resolution: {integrity: sha512-GQ2EWRpQV8/o+Aw8YqtfZZPfNRWZYkbidE9k5rpl/hC3vtHHBfGm2Ifi6qWV+coDGkrUKZAxE3Lot5kcsRlh+g==}
     engines: {node: '>=6'}
 
+  [email protected]:
+    resolution: {integrity: sha512-ruw5xyKs6lrpo9x9rCZqZZnIUntICjQAd0Wsmp396Ul9lN/h+ifgVV1x1gZHi8euej6wTfpqX8j+BFQxF0NS/g==}
+
+  [email protected]:
+    resolution: {integrity: sha512-JyeQc9iwFLn5TbvvqACIF/VXG6abODeB3Fwmv/TGdLk2LfbWkaySGY72at4+Ty7EkPZj854u4CrICqNk2qIbow==}
+
+  [email protected]:
+    resolution: {integrity: sha512-IInvU7fabl34qmi9gY8XOVxhYyMyuH2xUNpb2q8/Y+7552KlejkRvqvD19nMoUW/uQGGbqNpA6Tufu5FL5BZgw==}
+
   [email protected]:
     resolution: {integrity: sha512-ak9Qy5Q7jYb2Wwcey5Fpvg2KoAc/ZIhLSLOSBmRmygPsGwkVVt0fZa0qrtMz+m6tJTAHfZQ8FnmB4MG4LWy7/w==}
     engines: {node: '>=8'}
@@ -2031,6 +2094,9 @@ packages:
     resolution: {integrity: sha512-x/+Cz4YrimQxQccJf5mKEbIa1NzeCRNI5Ecl/ekmlYaampdNLPalVyIcCZNNH3MvmqBugV5TMYZXv0ljslUlaw==}
     engines: {node: '>= 0.4'}
 
+  [email protected]:
+    resolution: {integrity: sha512-YZo3K82SD7Riyi0E1EQPojLz7kpepnSQI9IyPbHHg1XXXevb5dJI7tpyN2ADxGcQbHG7vcyRHk0cbwqcQriUtg==}
+
   [email protected]:
     resolution: {integrity: sha512-UOShsPwz7NrMUqhR6t0hWjFduvOzbtv7toDH1/hIrfRNIDBnnBWd0CwJTGvTpngVlmwGCdP9/Zl/tVrDqcuYzQ==}
 
@@ -2265,6 +2331,10 @@ packages:
     resolution: {integrity: sha512-hU/10obOIu62MGYjdskASR3CUAiYaFTtC9Pa6vHyf//mAipSvSQg6od2CnJswq7fvzNS3zJhxoRkgNVaHurWKw==}
     engines: {node: '>=18.17'}
 
+  [email protected]:
+    resolution: {integrity: sha512-QEg3HPMll0o3t2ourKwOeUAZ159Kn9mx5pnzHRQO8+Wixmh88YdZRiIwat0iNzNNXn0yoEtXJqFpyW7eM8BV7g==}
+    engines: {node: '>=20.18.1'}
+
   [email protected]:
     resolution: {integrity: sha512-B60uB5TNBLtN6/LsgAf3udH9saB5p7gqJwcFfbOEZ8BcBHnGwCf6G/TGiEqkRAxX7zAFIUtzdrXQSdL3Q/wqNA==}
 
@@ -2312,6 +2382,14 @@ packages:
     resolution: {integrity: sha512-Lcvm7MGST/4fup+ifyKi2hjyIAwcdI4HRgtvTpIUxBRhB+RFtUh8XtDOxUfctVCnhVi+QQj49i91OyvzkJl6cg==}
     engines: {node: '>=10.13.0'}
 
+  [email protected]:
+    resolution: {integrity: sha512-6qN4hJdMwfYBtE3YBTTHhoeuUrDBPZmbQaxWAqSALV/MeEnR5z1xd8UKud2RAkFoPkmB+hli1TZSnyi84xz1vQ==}
+    engines: {node: '>=18'}
+
+  [email protected]:
+    resolution: {integrity: sha512-QaKxh0eNIi2mE9p2vEdzfagOKHCcj1pJ56EEHGQOVxp8r9/iszLUUV7v89x9O1p/T+NlTM5W7jW6+cz4Fq1YVg==}
+    engines: {node: '>=18'}
+
   [email protected]:
     resolution: {integrity: sha512-TbX3mj8n0odCBFVlY8AxkqcHASw3L60jIuF8jFP78az3C2YhmGvqbHBpAjTRH2/xqYunrJ9g1jSyjCjpoWzIAA==}
     engines: {node: '>= 0.4'}
@@ -3125,6 +3203,8 @@ snapshots:
 
   [email protected]: {}
 
+  [email protected]: {}
+
   [email protected]:
     dependencies:
       balanced-match: 1.0.2
@@ -3189,6 +3269,29 @@ snapshots:
       ansi-styles: 4.3.0
       supports-color: 7.2.0
 
+  [email protected]:
+    dependencies:
+      boolbase: 1.0.0
+      css-select: 5.2.2
+      css-what: 6.2.2
+      domelementtype: 2.3.0
+      domhandler: 5.0.3
+      domutils: 3.2.2
+
+  [email protected]:
+    dependencies:
+      cheerio-select: 2.1.0
+      dom-serializer: 2.0.0
+      domhandler: 5.0.3
+      domutils: 3.2.2
+      encoding-sniffer: 0.2.1
+      htmlparser2: 10.0.0
+      parse5: 7.3.0
+      parse5-htmlparser2-tree-adapter: 7.1.0
+      parse5-parser-stream: 7.1.2
+      undici: 7.16.0
+      whatwg-mimetype: 4.0.0
+
   [email protected]:
     dependencies:
       anymatch: 3.1.3
@@ -3225,6 +3328,16 @@ snapshots:
       shebang-command: 2.0.0
       which: 2.0.2
 
+  [email protected]:
+    dependencies:
+      boolbase: 1.0.0
+      css-what: 6.2.2
+      domhandler: 5.0.3
+      domutils: 3.2.2
+      nth-check: 2.1.1
+
+  [email protected]: {}
+
   [email protected]: {}
 
   [email protected]: {}
@@ -3291,6 +3404,24 @@ snapshots:
     dependencies:
       esutils: 2.0.3
 
+  [email protected]:
+    dependencies:
+      domelementtype: 2.3.0
+      domhandler: 5.0.3
+      entities: 4.5.0
+
+  [email protected]: {}
+
+  [email protected]:
+    dependencies:
+      domelementtype: 2.3.0
+
+  [email protected]:
+    dependencies:
+      dom-serializer: 2.0.0
+      domelementtype: 2.3.0
+      domhandler: 5.0.3
+
   [email protected]:
     dependencies:
       call-bind-apply-helpers: 1.0.2
@@ -3305,6 +3436,15 @@ snapshots:
 
   [email protected]: {}
 
+  [email protected]:
+    dependencies:
+      iconv-lite: 0.6.3
+      whatwg-encoding: 3.1.1
+
+  [email protected]: {}
+
+  [email protected]: {}
+
   [email protected]:
     dependencies:
       array-buffer-byte-length: 1.0.2
@@ -3792,6 +3932,17 @@ snapshots:
     dependencies:
       function-bind: 1.1.2
 
+  [email protected]:
+    dependencies:
+      domelementtype: 2.3.0
+      domhandler: 5.0.3
+      domutils: 3.2.2
+      entities: 6.0.1
+
+  [email protected]:
+    dependencies:
+      safer-buffer: 2.1.2
+
   [email protected]: {}
 
   [email protected]:
@@ -4097,6 +4248,10 @@ snapshots:
 
   [email protected]: {}
 
+  [email protected]:
+    dependencies:
+      boolbase: 1.0.0
+
   [email protected]: {}
 
   [email protected]: {}
@@ -4176,6 +4331,19 @@ snapshots:
     dependencies:
       callsites: 3.1.0
 
+  [email protected]:
+    dependencies:
+      domhandler: 5.0.3
+      parse5: 7.3.0
+
+  [email protected]:
+    dependencies:
+      parse5: 7.3.0
+
+  [email protected]:
+    dependencies:
+      entities: 6.0.1
+
   [email protected]: {}
 
   [email protected]: {}
@@ -4426,6 +4594,8 @@ snapshots:
       es-errors: 1.3.0
       is-regex: 1.2.1
 
+  [email protected]: {}
+
   [email protected]:
     dependencies:
       loose-envify: 1.4.0
@@ -4722,6 +4892,8 @@ snapshots:
 
   [email protected]: {}
 
+  [email protected]: {}
+
   [email protected]:
     dependencies:
       '@babel/runtime': 7.28.4
@@ -4787,6 +4959,12 @@ snapshots:
       glob-to-regexp: 0.4.1
       graceful-fs: 4.2.11
 
+  [email protected]:
+    dependencies:
+      iconv-lite: 0.6.3
+
+  [email protected]: {}
+
   [email protected]:
     dependencies:
       is-bigint: 1.1.0

src/app/api/debug-arch/route.ts

@@ -0,0 +1,36 @@
+import { NextResponse } from 'next/server'
+import { query } from '@/lib/database/config'
+
+export async function GET() {
+  try {
+    // Check if arch platform exists
+    const platformResult = await query(
+      'SELECT * FROM platforms WHERE id = $1',
+      ['arch']
+    )
+    
+    // Count arch packages
+    const countResult = await query(
+      'SELECT COUNT(*) FROM packages WHERE platform_id = $1',
+      ['arch']
+    )
+    
+    // Get sample packages
+    const sampleResult = await query(
+      'SELECT name, version, description FROM packages WHERE platform_id = $1 LIMIT 5',
+      ['arch']
+    )
+    
+    return NextResponse.json({
+      platform: platformResult.rows[0] || null,
+      totalPackages: parseInt(countResult.rows[0].count),
+      samplePackages: sampleResult.rows
+    })
+  } catch (error) {
+    console.error('Debug error:', error)
+    return NextResponse.json(
+      { error: error instanceof Error ? error.message : 'Unknown error' },
+      { status: 500 }
+    )
+  }
+}

src/app/api/sync-arch/route.ts

@@ -0,0 +1,86 @@
+import { NextResponse } from 'next/server'
+import { ArchPackageFetcher } from '@/services/archPackageFetcher'
+
+export const dynamic = 'force-dynamic'
+export const maxDuration = 300 // 5 minutes timeout
+
+let syncInProgress = false
+let syncStatus = {
+  status: 'idle' as 'idle' | 'fetching' | 'storing' | 'complete' | 'error',
+  fetchProgress: 0,
+  fetchTotal: 0,
+  storeProgress: 0,
+  storeTotal: 0,
+  currentPackage: '',
+  error: null as string | null
+}
+
+export async function GET() {
+  return NextResponse.json(syncStatus)
+}
+
+export async function POST() {
+  if (syncInProgress) {
+    return NextResponse.json(
+      { error: 'Sync already in progress' },
+      { status: 409 }
+    )
+  }
+
+  syncInProgress = true
+  syncStatus = {
+    status: 'fetching',
+    fetchProgress: 0,
+    fetchTotal: 0,
+    storeProgress: 0,
+    storeTotal: 0,
+    currentPackage: '',
+    error: null
+  }
+
+  // Start sync in background
+  ;(async () => {
+    try {
+      const fetcher = new ArchPackageFetcher()
+      
+      // Fetch packages
+      console.log('🔄 Starting Arch Linux package fetch...')
+      const packages = await fetcher.fetchAllPackages(
+        (current, total, packageName) => {
+          syncStatus.fetchProgress = current
+          syncStatus.fetchTotal = total
+          syncStatus.currentPackage = packageName
+        }
+      )
+      
+      console.log(`✅ Fetched ${packages.length} Arch packages`)
+      
+      // Store packages
+      syncStatus.status = 'storing'
+      syncStatus.storeTotal = packages.length
+      
+      await fetcher.storePackages(
+        packages,
+        (current, total) => {
+          syncStatus.storeProgress = current
+          syncStatus.storeTotal = total
+        }
+      )
+      
+      syncStatus.status = 'complete'
+      console.log('✅ Arch Linux sync completed successfully')
+      
+    } catch (error) {
+      console.error('❌ Arch sync error:', error)
+      syncStatus.status = 'error'
+      syncStatus.error = error instanceof Error ? error.message : 'Unknown error'
+    } finally {
+      syncInProgress = false
+    }
+  })()
+
+  return NextResponse.json({ 
+    message: 'Arch Linux package sync started',
+    status: syncStatus 
+  })
+}

src/services/archPackageFetcher.ts

@@ -0,0 +1,212 @@
+interface ArchPackage {
+  name: string
+  version: string
+  description: string
+  architecture: string
+  repository: string
+  lastUpdated: string
+}
+
+// Simple HTML parser without external dependencies
+function parseHTML(html: string) {
+  return {
+    querySelector: (selector: string) => {
+      const regex = new RegExp(`<${selector}[^>]*>([\\s\\S]*?)<\\/${selector}>`, 'gi')
+      const match = regex.exec(html)
+      return match ? { textContent: match[1].replace(/<[^>]*>/g, '').trim() } : null
+    },
+    querySelectorAll: (selector: string) => {
+      const results: any[] = []
+      const regex = new RegExp(`<${selector}[^>]*>([\\s\\S]*?)<\\/${selector}>`, 'gi')
+      let match
+      while ((match = regex.exec(html)) !== null) {
+        results.push({ innerHTML: match[1], textContent: match[1].replace(/<[^>]*>/g, '').trim() })
+      }
+      return results
+    }
+  }
+}
+
+export class ArchPackageFetcher {
+  private baseUrl = 'https://archlinux.org/packages'
+  
+  /**
+   * Fetch all Arch Linux packages with progress callback
+   */
+  async fetchAllPackages(
+    onProgress?: (current: number, total: number, packageName: string) => void
+  ): Promise<ArchPackage[]> {
+    console.log('🔍 Starting Arch Linux package fetch...')
+    
+    const allPackages: ArchPackage[] = []
+    let currentPage = 1
+    let totalPages = 1
+    
+    try {
+      // Fetch first page to get total pages
+      const firstPageUrl = `${this.baseUrl}/?page=1`
+      console.log(`📄 Fetching first page: ${firstPageUrl}`)
+      
+      const firstPageResponse = await fetch(firstPageUrl)
+      if (!firstPageResponse.ok) {
+        throw new Error(`HTTP error! status: ${firstPageResponse.status}`)
+      }
+      
+      const firstPageHtml = await firstPageResponse.text()
+      
+      // Parse total packages and pages
+      const statsMatch = firstPageHtml.match(/(\d+)\s+matching packages[\s\S]*?Page \d+ of (\d+)/)
+      const totalPackages = statsMatch ? parseInt(statsMatch[1]) : 0
+      totalPages = statsMatch ? parseInt(statsMatch[2]) : 1
+      
+      console.log(`📦 Total packages: ${totalPackages}`)
+      console.log(`📄 Total pages: ${totalPages}`)
+      
+      // Parse first page packages
+      const firstPagePackages = this.parsePackagesFromHtml(firstPageHtml)
+      allPackages.push(...firstPagePackages)
+      
+      if (onProgress) {
+        onProgress(allPackages.length, totalPackages, firstPagePackages[0]?.name || 'N/A')
+      }
+      
+      // Fetch remaining pages
+      for (currentPage = 2; currentPage <= totalPages; currentPage++) {
+        const pageUrl = `${this.baseUrl}/?page=${currentPage}`
+        console.log(`📄 Fetching page ${currentPage}/${totalPages}...`)
+        
+        try {
+          const response = await fetch(pageUrl)
+          if (!response.ok) {
+            console.error(`❌ Failed to fetch page ${currentPage}: ${response.status}`)
+            continue
+          }
+          
+          const html = await response.text()
+          const packages = this.parsePackagesFromHtml(html)
+          
+          allPackages.push(...packages)
+          
+          if (onProgress && packages.length > 0) {
+            onProgress(allPackages.length, totalPackages, packages[0].name)
+          }
+          
+          // Rate limiting: wait 100ms between requests
+          await new Promise(resolve => setTimeout(resolve, 100))
+          
+        } catch (error) {
+          console.error(`❌ Error fetching page ${currentPage}:`, error)
+        }
+      }
+      
+      console.log(`✅ Fetch completed! Total packages: ${allPackages.length}`)
+      return allPackages
+      
+    } catch (error) {
+      console.error('❌ Error fetching Arch packages:', error)
+      throw error
+    }
+  }
+  
+  /**
+   * Parse packages from HTML using regex
+   */
+  private parsePackagesFromHtml(html: string): ArchPackage[] {
+    const packages: ArchPackage[] = []
+    
+    // Match all table rows
+    const rowRegex = /<tr>\s*<td>([^<]+)<\/td>\s*<td>([^<]+)<\/td>\s*<td><a[^>]*>([^<]+)<\/a><\/td>\s*<td>([^<]+)<\/td>\s*<td[^>]*>([^<]*)<\/td>\s*<td>([^<]*)<\/td>/gi
+    
+    let match
+    while ((match = rowRegex.exec(html)) !== null) {
+      try {
+        const architecture = match[1].trim()
+        const repository = match[2].trim()
+        const name = match[3].trim()
+        const version = match[4].trim()
+        const description = match[5].trim()
+        const lastUpdated = match[6].trim()
+        
+        if (name && version) {
+          packages.push({
+            name,
+            version,
+            description: description || 'No description available',
+            architecture,
+            repository: repository.toLowerCase(),
+            lastUpdated
+          })
+        }
+      } catch (error) {
+        console.error('Error parsing package row:', error)
+      }
+    }
+    
+    return packages
+  }
+  
+  /**
+   * Store packages in database
+   */
+  async storePackages(
+    packages: ArchPackage[],
+    onProgress?: (current: number, total: number) => void
+  ): Promise<void> {
+    const { query } = await import('@/lib/database/config')
+    
+    console.log(`💾 Storing ${packages.length} Arch packages in database...`)
+    
+    const batchSize = 100
+    let stored = 0
+    let updated = 0
+    let skipped = 0
+    
+    for (let i = 0; i < packages.length; i += batchSize) {
+      const batch = packages.slice(i, i + batchSize)
+      
+      for (const pkg of batch) {
+        try {
+          // Check if package already exists
+          const existingResult = await query(
+            'SELECT id, version FROM packages WHERE name = $1 AND platform_id = $2',
+            [pkg.name, 'arch']
+          )
+          
+          const repository = (pkg.repository === 'core' || pkg.repository === 'extra') 
+            ? 'official' 
+            : 'third-party'
+          
+          if (existingResult.rows.length === 0) {
+            // Insert new package (id will be auto-generated by SERIAL)
+            await query(
+              `INSERT INTO packages (id, name, description, version, platform_id, type, repository, popularity_score, is_active)
+               VALUES (gen_random_uuid(), $1, $2, $3, $4, $5, $6, $7, $8)`,
+              [pkg.name, pkg.description, pkg.version, 'arch', 'cli', repository, 0, true]
+            )
+            stored++
+          } else if (existingResult.rows[0].version !== pkg.version) {
+            // Update if version changed
+            await query(
+              `UPDATE packages 
+               SET version = $1, description = $2, repository = $3, updated_at = NOW()
+               WHERE id = $4`,
+              [pkg.version, pkg.description, repository, existingResult.rows[0].id]
+            )
+            updated++
+          } else {
+            skipped++
+          }
+          
+          if (onProgress && (stored + updated + skipped) % 100 === 0) {
+            onProgress(stored + updated + skipped, packages.length)
+          }
+          
+        } catch (error) {
+          console.error(`Error storing package ${pkg.name}:`, error)
+        }
+      }
+    }
+    
+    console.log(`✅ Stored ${stored} new packages, updated ${updated}, skipped ${skipped}`)
+  }
+}