Upload 2 files

app.py

@@ -0,0 +1,107 @@
+import os
+import streamlit as st
+from paddleocr import PaddleOCR
+from langchain_groq import ChatGroq
+from langchain.output_parsers import PydanticOutputParser
+from langchain_core.prompts import PromptTemplate
+from pydantic import BaseModel, Field
+import fitz  
+import json
+from PIL import Image
+ocr = PaddleOCR(use_angle_cls=True, lang='es')
+
+st.set_page_config(layout="wide")
+
+class CarInfoEntity(BaseModel):
+    dealer_name: str = Field(description="Nombre del concesionario o empresa.")
+    dealer_address: str = Field(description="Dirección física del concesionario.")
+    tax_id: str = Field(description="Número de identificación fiscal del concesionario.")
+    contact_phone: str = Field(description="Número de teléfono principal para contactar con el concesionario.")
+    contact_fax: str = Field(description="Número de fax del concesionario.")
+    contact_email: str = Field(description="Dirección de correo electrónico para consultas.")
+    website_url: str = Field(description="Sitio web oficial del concesionario.")
+    operating_hours: str = Field(description="Horario habitual de atención del concesionario.")
+    saturday_hours: str = Field(description="Horario de atención específico para los sábados.")
+    order_date: str = Field(description="Fecha en que se realizó el pedido.")
+    order_number: str = Field(description="Identificador único del pedido.")
+    sales_rep: str = Field(description="Nombre del vendedor que maneja la transacción.")
+    customer_full_name: str = Field(description="Nombre completo del comprador.")
+    customer_address: str = Field(description="Dirección del comprador.")
+    customer_city: str = Field(description="Ciudad donde reside el comprador.")
+    customer_postal_code: str = Field(description="Código postal de la dirección del comprador.")
+    customer_province: str = Field(description="Provincia donde se encuentra el comprador.")
+    customer_id: str = Field(description="Número de identificación del comprador (NIF).")
+    customer_phone: str = Field(description="Número de teléfono del comprador.")
+    vehicle_description: str = Field(description="Descripción del vehículo que se está comprando, incluyendo marca, modelo y año.")
+    vehicle_color: str = Field(description="Color del vehículo.")
+    vehicle_price: str = Field(description="Precio total del vehículo, incluyendo impuestos.")
+
+model = ChatGroq(
+    model="llama-3.1-70b-versatile",
+    temperature=0,
+    max_tokens=None,
+    timeout=None,
+    max_retries=2,
+    api_key='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'
+)
+
+os.environ['GROQ_API_KEY'] = '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'
+
+entity = ['dealer_name', 'dealer_address', 'tax_id', 'contact_phone', 'contact_fax', 'contact_email', 'website_url', 
+          'operating_hours', 'saturday_hours', 'order_date', 'order_number', 'sales_rep', 
+          'customer_full_name', 'customer_address', 'customer_city', 'customer_postal_code', 
+          'customer_province', 'customer_id','customer_phone', 'vehicle_description','vehicle_color','vehicle_price']
+
+# Streamlit App
+st.title("Vehicle Information Extractor")
+st.write("Upload a PDF file to extract vehicle information.")
+
+uploaded_file = st.file_uploader("Choose a PDF file", type="pdf")
+
+if uploaded_file is not None:
+    with open("temp.pdf", "wb") as f:
+        f.write(uploaded_file.read())
+
+    col1, col2 = st.columns(2)
+
+    with col1:
+        doc = fitz.open("temp.pdf")
+        st.write("Uploaded PDF:")
+        for page_num in range(len(doc)):
+            page = doc.load_page(page_num)
+            pix = page.get_pixmap()
+            img = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
+            st.image(img, caption=f"Page {page_num+1}", use_column_width=True)
+
+    content = ocr.ocr("temp.pdf")
+
+    extracted_text = []
+    for page in content:
+        for result in page:
+            text = result[1][0]
+            extracted_text.append(text)
+
+    all_text = " ".join(extracted_text)
+
+    prompt_text = """Task: Analyze the {all_text} and find out given entity value:{entity} from the {all_text}:
+
+    Output Format: A table with the entity and value. First column contains the {entity} and second column contains the value fetched from the {all_text}.
+                   
+    Do not include any additional explanations or unnecessary details. 
+    {format_instructions}"""
+
+    parser = PydanticOutputParser(pydantic_object=CarInfoEntity)
+
+    prompt = PromptTemplate(
+        template=prompt_text,
+        input_variables=["all_text", "entity"],
+        partial_variables={"format_instructions": parser.get_format_instructions()},
+    )
+
+    chain = prompt | model | parser
+
+    output = chain.invoke({"all_text": all_text, "entity": entity})
+
+    with col2:
+        st.write("Extracted Vehicle Information (Table):")
+        st.table(output)

requirements.txt

@@ -0,0 +1,7 @@
+paddleocr==2.8.1
+langchain==0.3.3
+paddlepaddle==2.6.2
+langchain_groq==0.2.0
+PyMuPDF==1.24.11
+pillow==10.4.0
+streamlit==1.39.0