cr

Added aws texract extraction python file to the document loader module .
This will help in extracting images text with aws texract service.

langchain/document_loaders/aws_textract.py

@@ -0,0 +1,58 @@
+from io import BytesIO
+from typing import List
+
+import boto3
+from PIL import Image
+
+from langchain.docstore.document import Document
+
+
+class AwsTextractExtraction:
+    def __init__(
+        self,
+        aws_region_name: str,
+        aws_secret_key: str,
+        aws_access_key: str,
+        aws_session_token: str,
+        file_path: str,
+    ):
+        self.aws_region_name = aws_region_name
+        self.aws_secret_key = aws_secret_key
+        self.aws_access_key = aws_access_key
+        self.aws_session_token = aws_session_token
+        self.file_path = file_path
+        try:
+            import boto3  # noqa: F401
+
+        except ImportError:
+            raise ValueError(
+                "Could not import aws boto3 package. "
+                "Please install it with `pip install boto3`."
+            )
+
+    def get_text_from_image(self) -> List[Document]:
+        output = []
+        page_no = 0
+
+        textract_client = boto3.client(
+            "textract",
+            self.aws_region_name,
+            self.aws_access_key,
+            self.aws_access_key,
+            self.aws_session_token,
+        )
+        pil_image_obj = Image.open(self.file_path)
+        buf = BytesIO()
+        pil_image_obj.save(buf, format="PNG")
+        image_bytes = buf.getvalue()
+
+        response = textract_client.detect_document_text(Document={"Bytes": image_bytes})
+        detected_txt = ""
+        for item in response["Blocks"]:
+            if item["BlockType"] == "LINE":
+                detected_txt += item["Text"] + "\n"
+
+        metadata = {"source": self.file_path, "page": page_no}
+
+        output.append(Document(page_content=detected_txt, metadata=metadata))
+        return output