add document references

Modules/LlmDriver/app/Functions/SearchAndSummarize.php

@@ -3,12 +3,11 @@
 namespace LlmLaraHub\LlmDriver\Functions;
 
 use App\Domains\Messages\RoleEnum;
-use App\Models\DocumentChunk;
-use App\Models\Message;
-use Illuminate\Database\Eloquent\Collection;
 use Illuminate\Support\Facades\Log;
 use Laravel\Pennant\Feature;
 use LlmLaraHub\LlmDriver\HasDrivers;
+use LlmLaraHub\LlmDriver\Helpers\CreateReferencesTrait;
+use LlmLaraHub\LlmDriver\Helpers\DistanceQueryTrait;
 use LlmLaraHub\LlmDriver\LlmDriverFacade;
 use LlmLaraHub\LlmDriver\Requests\MessageInDto;
 use LlmLaraHub\LlmDriver\Responses\CompletionResponse;
@@ -17,6 +16,8 @@
 
 class SearchAndSummarize extends FunctionContract
 {
+    use CreateReferencesTrait, DistanceQueryTrait;
+
     protected string $name = 'search_and_summarize';
 
     protected string $description = 'Used to embed users prompt, search database and return summarized results.';
@@ -29,6 +30,8 @@ public function handle(
         HasDrivers $model,
         FunctionCallDto $functionCallDto): FunctionResponse
     {
+        Log::info('[LaraChain] Using Function: SearchAndSummarize');
+
         /**
          * @TODO
          *
@@ -50,22 +53,11 @@ public function handle(
 
         $embeddingSize = get_embedding_size($model->getEmbeddingDriver());
 
-        /**
-         * @TODO
-         * Track the document page for referehce
-         *
-         * @see https://github.com/orgs/LlmLaraHub/projects/1?pane=issue&itemId=60394288
-         */
-        $results = DocumentChunk::query()
-            ->join('documents', 'documents.id', '=', 'document_chunks.document_id')
-            ->selectRaw(
-                "document_chunks.{$embeddingSize} <-> ? as distance, document_chunks.content, document_chunks.{$embeddingSize} as embedding, document_chunks.id as id, document_chunks.summary as summary, document_chunks.document_id as document_id",
-                [$embedding->embedding]
-            )
-            ->where('documents.collection_id', $model->getChatable()->id)
-            ->limit(10)
-            ->orderByRaw('distance')
-            ->get();
+        $documentChunkResults = $this->distance(
+            $embeddingSize,
+            $model->getChatable()->id,
+            $embedding->embedding
+        );
 
         $content = [];
 
@@ -74,7 +66,7 @@ public function handle(
          * Yes this is a lot like the SearchAndSummarizeChatRepo
          * But just getting a sense of things
          */
-        foreach ($results as $result) {
+        foreach ($documentChunkResults as $result) {
             $contentString = remove_ascii($result->content);
             if (Feature::active('reduce_text')) {
                 $result = reduce_text_size($contentString);
@@ -107,10 +99,7 @@ public function handle(
 
         $message = $model->getChat()->addInput($response->content, RoleEnum::Assistant);
 
-        /**
-         * We want to trigger the job to build up document reference history
-         */
-        $this->saveDocumentReference($message, $results);
+        $this->saveDocumentReference($message, $documentChunkResults);
 
         return FunctionResponse::from(
             [
@@ -120,20 +109,6 @@ public function handle(
         );
     }
 
-    protected function saveDocumentReference(
-        Message $model,
-        Collection $documentChunks
-    ): void {
-        //add each one to a batch job or do the work here.
-        foreach ($documentChunks as $documentChunk) {
-            $model->message_document_references()->create([
-                'document_chunk_id' => $documentChunk->id,
-                'distance' => $documentChunk->distance,
-                'reference' => $documentChunk->summary,
-            ]);
-        }
-    }
-
     /**
      * @return PropertyDto[]
      */

Modules/LlmDriver/app/Helpers/CreateReferencesTrait.php

@@ -0,0 +1,22 @@
+<?php
+
+namespace LlmLaraHub\LlmDriver\Helpers;
+
+use App\Models\Message;
+use Illuminate\Database\Eloquent\Collection;
+
+trait CreateReferencesTrait
+{
+    protected function saveDocumentReference(
+        Message $model,
+        Collection $documentChunks
+    ): void {
+        //add each one to a batch job or do the work here.
+        foreach ($documentChunks as $documentChunk) {
+            $model->message_document_references()->create([
+                'document_chunk_id' => $documentChunk->id,
+                'distance' => $documentChunk->distance,
+            ]);
+        }
+    }
+}

Modules/LlmDriver/app/Helpers/DistanceQueryTrait.php

@@ -0,0 +1,37 @@
+<?php
+
+namespace LlmLaraHub\LlmDriver\Helpers;
+
+use App\Models\DocumentChunk;
+use Illuminate\Database\Eloquent\Collection;
+use Pgvector\Laravel\Vector;
+
+trait DistanceQueryTrait
+{
+    /**
+     * @TODO
+     * Track the document page for referehce
+     *
+     * @see https://github.com/orgs/LlmLaraHub/projects/1?pane=issue&itemId=60394288
+     */
+    protected function distance(
+        string $embeddingSize,
+        int $collectionId,
+        Vector $embedding
+    ): Collection {
+
+        $documentChunkResults = DocumentChunk::query()
+            ->join('documents', 'documents.id', '=', 'document_chunks.document_id')
+            ->selectRaw(
+                "document_chunks.{$embeddingSize} <-> ? as distance, document_chunks.content as content, document_chunks.{$embeddingSize} as embedding, document_chunks.id as id, document_chunks.summary as summary, document_chunks.document_id as document_id",
+                [$embedding]
+            )
+            /** @phpstan-ignore-next-line */
+            ->where('documents.collection_id', $collectionId)
+            ->limit(10)
+            ->orderByRaw('distance')
+            ->get();
+
+        return $documentChunkResults;
+    }
+}

app/Domains/Messages/SearchAndSummarizeChatRepo.php

@@ -6,12 +6,16 @@
 use App\Models\DocumentChunk;
 use Illuminate\Support\Facades\Log;
 use Laravel\Pennant\Feature;
+use LlmLaraHub\LlmDriver\Helpers\CreateReferencesTrait;
+use LlmLaraHub\LlmDriver\Helpers\DistanceQueryTrait;
 use LlmLaraHub\LlmDriver\LlmDriverFacade;
 use LlmLaraHub\LlmDriver\Responses\CompletionResponse;
 use LlmLaraHub\LlmDriver\Responses\EmbeddingsResponseDto;
 
 class SearchAndSummarizeChatRepo
 {
+    use CreateReferencesTrait, DistanceQueryTrait;
+
     public function search(Chat $chat, string $input): string
     {
         /**
@@ -29,26 +33,16 @@ public function search(Chat $chat, string $input): string
 
         $embeddingSize = get_embedding_size($chat->chatable->getEmbeddingDriver());
 
-        /**
-         * @TODO
-         * Track the document page for referehce
-         *
-         * @see https://github.com/orgs/LlmLaraHub/projects/1?pane=issue&itemId=60394288
-         */
-        $results = DocumentChunk::query()
-            ->join('documents', 'documents.id', '=', 'document_chunks.document_id')
-            ->selectRaw(
-                "document_chunks.{$embeddingSize} <-> ? as distance, document_chunks.content, document_chunks.{$embeddingSize} as embedding, document_chunks.id as id",
-                [$embedding->embedding]
-            )
-            ->where('documents.collection_id', $chat->chatable->id)
-            ->limit(10)
-            ->orderByRaw('distance')
-            ->get();
-
+        $documentChunkResults = $this->distance(
+            $embeddingSize,
+            /** @phpstan-ignore-next-line */
+            $chat->getChatable()->id,
+            $embedding->embedding
+        );
         $content = [];
 
-        foreach ($results as $result) {
+        /** @var DocumentChunk $result */
+        foreach ($documentChunkResults as $result) {
             $contentString = remove_ascii($result->content);
             if (Feature::active('reduce_text')) {
                 $result = reduce_text_size($contentString);
@@ -68,13 +62,17 @@ public function search(Chat $chat, string $input): string
         );
 
         $latestMessagesArray = $chat->getChatResponse();
+
         Log::info('[LaraChain] Getting the Summary');
+
         /** @var CompletionResponse $response */
         $response = LlmDriverFacade::driver(
             $chat->chatable->getDriver()
         )->chat($latestMessagesArray);
 
-        $chat->addInput($response->content, RoleEnum::Assistant);
+        $message = $chat->addInput($response->content, RoleEnum::Assistant);
+
+        $this->saveDocumentReference($message, $documentChunkResults);
 
         return $response->content;
     }

app/Http/Resources/MessageDocumentReferenceResource.php

@@ -0,0 +1,26 @@
+<?php
+
+namespace App\Http\Resources;
+
+use Illuminate\Http\Request;
+use Illuminate\Http\Resources\Json\JsonResource;
+
+class MessageDocumentReferenceResource extends JsonResource
+{
+    /**
+     * Transform the resource into an array.
+     *
+     * @return array<string, mixed>
+     */
+    public function toArray(Request $request): array
+    {
+        return [
+            'id' => $this->id,
+            'document_name' => $this->document_chunk->document->file_path,
+            'page' => $this->document_chunk->sort_order,
+            'distance' => round($this->distance, 2),
+            'summary' => str($this->document_chunk->summary)->markdown(),
+            'taggings' => TagResource::collection($this->document_chunk->tags),
+        ];
+    }
+}

app/Http/Resources/MessageResource.php

@@ -22,6 +22,8 @@ public function toArray(Request $request): array
             'body' => $this->body,
             'body_markdown' => str($this->body)->markdown(),
             'diff_for_humans' => $this->created_at->diffForHumans(),
+            'message_document_references' => MessageDocumentReferenceResource::collection(
+                $this->message_document_references()->orderBy('distance', 'asc')->limit(5)->get()),
         ];
     }
 }

app/Http/Resources/TagResource.php

@@ -14,6 +14,9 @@ class TagResource extends JsonResource
      */
     public function toArray(Request $request): array
     {
-        return parent::toArray($request);
+        return [
+            'id' => $this->id,
+            'name' => str($this->name)->headline()->toString(),
+        ];
     }
 }

database/factories/MessageDocumentReferenceFactory.php

@@ -21,7 +21,6 @@ public function definition(): array
         return [
             'message_id' => Message::factory(),
             'document_chunk_id' => DocumentChunk::factory(),
-            'reference' => fake()->text(),
             'distance' => 11.321851037400464,
         ];
     }

database/migrations/2024_04_27_170436_create_chat_document_references_table.php

@@ -17,7 +17,6 @@ public function up(): void
             $table->id();
             $table->foreignIdFor(Message::class);
             $table->foreignIdFor(DocumentChunk::class);
-            $table->string('reference')->nullable();
             $table->decimal('distance', 18, 15)->nullable();
             $table->timestamps();
         });