redis migration

Signed-off-by: ale <ale@manalejandro.com>

app/api/search/route.ts

@@ -1,152 +1,52 @@
 import { NextRequest, NextResponse } from 'next/server';
-import { esClient, INDEX_NAME, initializeIndex } from '@/lib/elasticsearch';
+import { storeHashDocument, findByPlaintext, findByHash, initializeRedis } from '@/lib/redis';
 import { generateHashes, detectHashType } from '@/lib/hash';
 
-interface HashDocument {
-  plaintext: string;
-  md5: string;
-  sha1: string;
-  sha256: string;
-  sha512: string;
-  created_at?: string;
-}
-
-// Maximum allowed query length
-const MAX_QUERY_LENGTH = 1000;
-
-// Characters that could be used in NoSQL/Elasticsearch injection attacks
-const DANGEROUS_PATTERNS = [
-  /[{}\[\]]/g,           // JSON structure characters
-  /\$[a-zA-Z]/g,         // MongoDB-style operators
-  /\\u[0-9a-fA-F]{4}/g,  // Unicode escapes
-  /<script/gi,           // XSS attempts
-  /javascript:/gi,       // XSS attempts
-];
-
-/**
- * Sanitize input to prevent NoSQL injection attacks
- * For hash lookups, we only need alphanumeric characters and $
- * For plaintext, we allow more characters but sanitize dangerous patterns
- */
-function sanitizeInput(input: string): string {
-  // Trim and take first word only
-  let sanitized = input.trim().split(/\s+/)[0] || '';
-  
-  // Limit length
-  if (sanitized.length > MAX_QUERY_LENGTH) {
-    sanitized = sanitized.substring(0, MAX_QUERY_LENGTH);
-  }
-  
-  // Remove null bytes
-  sanitized = sanitized.replace(/\0/g, '');
-  
-  // Check for dangerous patterns
-  for (const pattern of DANGEROUS_PATTERNS) {
-    sanitized = sanitized.replace(pattern, '');
-  }
-  
-  return sanitized;
-}
-
-/**
- * Validate that the input is safe for use in Elasticsearch queries
- */
-function isValidInput(input: string): boolean {
-  // Check for empty input
-  if (!input || input.length === 0) {
-    return false;
-  }
-  
-  // Check for excessively long input
-  if (input.length > MAX_QUERY_LENGTH) {
-    return false;
-  }
-  
-  // Check for control characters (except normal whitespace)
-  if (/[\x00-\x08\x0B\x0C\x0E-\x1F\x7F]/.test(input)) {
-    return false;
-  }
-  
-  return true;
-}
-
 export async function POST(request: NextRequest) {
   try {
-    const body = await request.json();
-    
-    // Validate request body structure
-    if (!body || typeof body !== 'object') {
-      return NextResponse.json(
-        { error: 'Invalid request body' },
-        { status: 400 }
-      );
-    }
+    const { query } = await request.json();
 
-    const { query } = body;
-
-    // Validate query type
     if (!query || typeof query !== 'string') {
       return NextResponse.json(
-        { error: 'Query parameter is required and must be a string' },
+        { error: 'Query parameter is required' },
         { status: 400 }
       );
     }
 
-    // Validate input before processing
-    if (!isValidInput(query)) {
-      return NextResponse.json(
-        { error: 'Invalid query: contains forbidden characters or is too long' },
-        { status: 400 }
-      );
-    }
+    // Ensure Redis is connected
+    await initializeRedis();
 
-    // Sanitize input
-    const cleanQuery = sanitizeInput(query);
+    const cleanQuery = query.trim().split(/\s+/)[0];
     
     if (!cleanQuery) {
       return NextResponse.json(
-        { error: 'Invalid query: only whitespace or invalid characters provided' },
+        { error: 'Invalid query: only whitespace provided' },
         { status: 400 }
       );
     }
 
-    // Ensure index exists
-    await initializeIndex();
-
     const cleanQueryLower = cleanQuery.toLowerCase();
     const hashType = detectHashType(cleanQueryLower);
 
     if (hashType) {
-      // Query is a hash - search for it in Elasticsearch
-      const searchResponse = await esClient.search<HashDocument>({
-        index: INDEX_NAME,
-        query: {
-          term: {
-            [hashType]: cleanQueryLower
-          }
-        }
-      });
+      // Query is a hash - search for it in Redis
+      const doc = await findByHash(hashType, cleanQueryLower);
 
-      const hits = searchResponse.hits.hits;
-
-      if (hits.length > 0) {
+      if (doc) {
         // Found matching plaintext
         return NextResponse.json({
           found: true,
           hashType,
           hash: cleanQuery,
-          results: hits.map((hit) => {
-            const source = hit._source!;
-            return {
-              plaintext: source.plaintext,
-              hashes: {
-                md5: source.md5,
-                sha1: source.sha1,
-                sha256: source.sha256,
-                sha512: source.sha512,
-              }
-            };
-          })
+          results: [{
+            plaintext: doc.plaintext,
+            hashes: {
+              md5: doc.md5,
+              sha1: doc.sha1,
+              sha256: doc.sha256,
+              sha512: doc.sha512,
+            }
+          }]
         });
       } else {
         // Hash not found in database
@@ -159,20 +59,13 @@ export async function POST(request: NextRequest) {
       }
     } else {
       // Query is plaintext - check if it already exists first
-      const existsResponse = await esClient.search<HashDocument>({
-        index: INDEX_NAME,
-        query: {
-          term: {
-            'plaintext.keyword': cleanQuery
-          }
-        }
-      });
+      const existingDoc = await findByPlaintext(cleanQuery);
 
       let hashes;
+      let wasGenerated = false;
 
-      if (existsResponse.hits.hits.length > 0) {
+      if (existingDoc) {
         // Plaintext found, retrieve existing hashes
-        const existingDoc = existsResponse.hits.hits[0]._source!;
         hashes = {
           md5: existingDoc.md5,
           sha1: existingDoc.sha1,
@@ -180,44 +73,22 @@ export async function POST(request: NextRequest) {
           sha512: existingDoc.sha512,
         };
       } else {
-        // Plaintext not found, generate hashes and check if any hash already exists
-        hashes = generateHashes(cleanQuery);
+        // Plaintext not found, generate and store hashes
+        hashes = await generateHashes(cleanQuery);
         
-        const hashExistsResponse = await esClient.search<HashDocument>({
-          index: INDEX_NAME,
-          query: {
-            bool: {
-              should: [
-                { term: { md5: hashes.md5 } },
-                { term: { sha1: hashes.sha1 } },
-                { term: { sha256: hashes.sha256 } },
-                { term: { sha512: hashes.sha512 } },
-              ],
-              minimum_should_match: 1
-            }
-          }
+        await storeHashDocument({
+          ...hashes,
+          created_at: new Date().toISOString()
         });
-
-        if (hashExistsResponse.hits.hits.length === 0) {
-          // No duplicates found, insert new document
-          await esClient.index({
-            index: INDEX_NAME,
-            document: {
-              ...hashes,
-              created_at: new Date().toISOString()
-            }
-          });
-
-          // Refresh index to make the document searchable immediately
-          await esClient.indices.refresh({ index: INDEX_NAME });
-        }
+        
+        wasGenerated = true;
       }
 
       return NextResponse.json({
         found: true,
         isPlaintext: true,
         plaintext: cleanQuery,
-        wasGenerated: existsResponse.hits.hits.length === 0,
+        wasGenerated,
         hashes: {
           md5: hashes.md5,
           sha1: hashes.sha1,