From 91ddb3672e514fa9824609ff047d7cab0c65631a Mon Sep 17 00:00:00 2001
From: Gustaf Rydholm <gustaf.rydholm@gmail.com>
Date: Tue, 9 Apr 2024 00:14:00 +0200
Subject: Refactor

---
 rag/retriever/document.py | 57 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 57 insertions(+)
 create mode 100644 rag/retriever/document.py

(limited to 'rag/retriever/document.py')

diff --git a/rag/retriever/document.py b/rag/retriever/document.py
new file mode 100644
index 0000000..54ac451
--- /dev/null
+++ b/rag/retriever/document.py
@@ -0,0 +1,57 @@
+import hashlib
+import os
+
+from langchain_community.document_loaders.blob_loaders import Blob
+import psycopg
+from loguru import logger as log
+
+TABLES = """
+CREATE TABLE IF NOT EXISTS document (
+        hash text PRIMARY KEY)
+"""
+
+
+class DocumentDB:
+    def __init__(self) -> None:
+        self.conn = psycopg.connect(
+            f"dbname={os.environ['DOCUMENT_DB_NAME']} user={os.environ['DOCUMENT_DB_USER']}"
+        )
+        self.__configure()
+
+    def close(self):
+        self.conn.close()
+
+    def __configure(self):
+        log.debug("Creating documents table if it does not exist...")
+        with self.conn.cursor() as cur:
+            cur.execute(TABLES)
+            self.conn.commit()
+
+    def __hash(self, blob: Blob) -> str:
+        log.debug("Hashing document...")
+        return hashlib.sha256(blob.as_bytes()).hexdigest()
+
+    def add(self, blob: Blob) -> bool:
+        with self.conn.cursor() as cur:
+            hash = self.__hash(blob)
+            cur.execute(
+                """
+                        SELECT * FROM document 
+                        WHERE
+                        hash = %s
+                        """,
+                (hash,),
+            )
+            exist = cur.fetchone()
+            if exist is None:
+                log.debug("Inserting document hash into documents db...")
+                cur.execute(
+                    """
+                            INSERT INTO document 
+                            (hash) VALUES 
+                            (%s)
+                            """,
+                    (hash,),
+                )
+            self.conn.commit()
+        return exist is None
-- 
cgit v1.2.3-70-g09d2