pdf reader added

app.py

@@ -10,6 +10,7 @@ from io import BytesIO
 import re
 from pathlib import Path
 import openai
+import pdfplumber
 
 ## utilty functions
 def is_image_extension(filename: str) -> bool:
@@ -21,25 +22,28 @@ def load_file(path: str) -> list | dict:
     """Based on the file extension, load the file into a suitable object."""
     
     image = None
-    excel = None
-    csv = None
     text = None
     ext = Path(path).suffix.lower() # same as os.path.splitext(filename)[1].lower()
 
     if ext.endswith(".png") or ext.endswith(".jpg") or ext.endswith(".jpeg"):
         image = Image.open(path).convert("RGB")            # pillow object
     elif ext.endswith(".xlsx") or ext.endswith(".xls"):
-        excel = pd.read_excel(path)                        # DataFrame
+        text = pd.read_excel(path)                        # DataFrame
     elif ext.endswith(".csv"):
-        csv = pd.read_csv(path)                          # DataFrame
+        text = pd.read_csv(path)                          # DataFrame
+    elif ext.endswith(".pdf"):
+        with pdfplumber.open(path) as pdf:
+            text = "\n".join(page.extract_text() for page in pdf.pages if page.extract_text())
     elif ext.endswith(".py") or ext.endswith(".txt"):
         with open(path, 'r') as f:
             text = f.read()                               # plain text str
               
     if image is not None:
         return [image]                 
+    elif ext.endswith(".mp3") or ext.endswith(".wav"):
+        return {"raw document text": text, "audio path": path} 
     else:
-        return {"excel": excel, "csv": csv, "raw text": text, "audio path": path} 
+        return {"raw document text": text, "file path": path}
 
 
 ## tools definition
@@ -68,8 +72,8 @@ def download_images(image_urls: str) -> list:
             print(f"Failed to download from {url}: {e}")
     return images
 
-@tool 
-def transcribe_audio(audio_path: str) -> str: # since they gave us OpenAI API credits, we can keep using it
+@tool # since they gave us OpenAI API credits, we can keep using it
+def transcribe_audio(audio_path: str) -> str:
     """
     Transcribe audio file using OpenAI Whisper API.
     Args: