From ce7b936f4fd1246953de51faf408b6a6bc13d27b Mon Sep 17 00:00:00 2001
From: Saifeddine ALOUI <aloui.seifeddine@gmail.com>
Date: Sun, 21 Jul 2024 22:37:12 +0200
Subject: [PATCH] fix

---
 lollms/app.py      | 6 +++++-
 lollms/internet.py | 6 +++---
 2 files changed, 8 insertions(+), 4 deletions(-)

diff --git a/lollms/app.py b/lollms/app.py
index 59c084b..eafb4b2 100644
--- a/lollms/app.py
+++ b/lollms/app.py
@@ -974,7 +974,11 @@ class LollmsApplication(LoLLMsCom):
                     
                     if len(chunks)>0:
                         for chunk in chunks:
-                            internet_search_infos.append(chunk.doc.title)
+                            internet_search_infos.append({
+                                "title":chunk.doc.title,
+                                "url":chunk.doc.path,
+                                "brief":chunk.text
+                            })
                             internet_search_results += f"{self.start_header_id_template}search result chunk{self.end_header_id_template}\nchunk_infos:{chunk.doc.path}\nchunk_title:{chunk.doc.title}\ncontent:{chunk.text}\n"
                     else:
                         internet_search_results += "The search response was empty!\nFailed to recover useful information from the search engine.\n"
diff --git a/lollms/internet.py b/lollms/internet.py
index a6b710c..2fc8b92 100644
--- a/lollms/internet.py
+++ b/lollms/internet.py
@@ -337,12 +337,12 @@ def internet_search_with_vectorization(query, chromedriver_path=None, internet_n
         from lollmsvectordb.lollms_vectorizers.word2vec_vectorizer import Word2VecVectorizer
         v = Word2VecVectorizer()
 
-    vectorizer = VectorDatabase("", v, TikTokenTokenizer())
+    vectorizer = VectorDatabase("", v, TikTokenTokenizer(), internet_vectorization_chunk_size, internet_vectorization_overlap_size)
 
     formatted_text = ""
     nb_non_empty = 0
     ise = InternetSearchEnhancer()
-    results = ise.search(query)
+    results = ise.search(query, num_results=internet_nb_search_pages)
     
     if len(results)>0:
         for i, result in enumerate(results):
@@ -351,7 +351,7 @@ def internet_search_with_vectorization(query, chromedriver_path=None, internet_n
             brief = result["snippet"]
             href = result["url"]
             if quick_search:
-                vectorizer.add_document({'url':href, 'title':title, 'brief': brief}, brief)
+                vectorizer.add_document(title, brief, href)
             else:
                 get_relevant_text_block(href, vectorizer, title)
             nb_non_empty += 1