Pdf Ingestion pipeline completed

2024-08-05 22:14:19 +01:00
parent b0c3eb8032
commit c34de21971
15 changed files with 318 additions and 90 deletions
@@ -0,0 +1,25 @@
+from data_ingestion.utils import search, load_embedded_data
+import sys, os
+
+#  Add the root directory to sys.path
+sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
+from loggings.logging_config import logger
+
+
+
+# loading the embedded data
+embed_db = load_embedded_data()
+
+
+
+if __name__ == "__main__":
+    logger.info("Receiving the search query")
+    query = input("Enter the search query: ")
+    logger.info(f"Searching for {query}")
+    page_content, all, pages = search(embed_db, query)
+    logger.info("Search completed")
+    logger.info(f"Page content: {page_content}")
+    print(f"Page content: {page_content}")
+    print(f"Pages: {pages}")
+    print(f"All: {all}")
+    print("Search completed")