checkvectordb.py

import os
from flask import Flask, request,render_template, send_from_directory

from flask import Flask, request, jsonify, render_template_string
import openai
import langchain
import os
from langchain.document_loaders import PyPDFLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.embeddings.openai import OpenAIEmbeddings
from langchain.vectorstores import Chroma
from langchain import OpenAI
from langchain.schema import Document
from langchain.chains.question_answering import load_qa_chain
from langchain.chains import RetrievalQA
from langchain.chat_models import ChatOpenAI
import re
from pdfminer.high_level import extract_pages
from pdfminer.layout import LTTextContainer, LTChar
from langchain.prompts import PromptTemplate
from langchain.chains import LLMChain

import json

import pdfplumber
import fitz  # PyMuPDF


import chromadb
from chromadb.config import Settings

app = Flask(__name__)
os.environ["OPENAI_API_KEY"] = "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"

app.config['UPLOAD_FOLDER'] = 'uploads'
openapi_key="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"
fpath="./SCHOOL ADMISSI0N TEST 2025-2026.pdf"
text_splitter=RecursiveCharacterTextSplitter(chunk_size=800,chunk_overlap=50,length_function=len,separators=["\n\n","\n"," "])
global_var = "I am global"


def loadmiyapurcourtorders(pathtosjon):
  try:
    with open(pathtosjon, "r", encoding="utf-8") as f:
     print("--processing pls wait---") 
     courtcontent = json.load(f)

     docs = []
     for content in courtcontent:

      content1 = f"""
      Court Level: {content["court_level"]}
      Case No: {content["caseno"]}
      Year: {content["year"]}
      Prayer: {content["prayer"]}
      Verdict: {content["verdict"]}
      Judgement Date: {content["verdictdate"]}
      Status: {content["status"]}
      Case Type: {content["casetype"]}
      petitioner: {content["petitioner"]}
      respondent 1: {content["Respondent 1"]}
      judge: {content["judge"]}

      """


      metadata = {
        "case no": str(content["caseno"]),
        "year": content["year"]
       

      }

      

    docs.append(Document(page_content=content1, metadata=metadata))

    embeddings = OpenAIEmbeddings(openai_api_key=openapi_key)
   # store all IPC section Articles here
    vectorstore = Chroma.from_documents(docs, embeddings, persist_directory="./MiyapurCase_db1")
  except Exception as er:
    print("--exception---",er)
    return jsonify("exception occured while processing...") 
  

  return "Miyapur court cases are loaded sucessfully"



def loadIPCsections(pathtosjon):
  
  try:
   with open(pathtosjon, "r", encoding="utf-8") as f:
    print("--processing pls wait---") 
    ipc_sections = json.load(f)

   docs = []
   for section in ipc_sections:
    content = f"Section {section['Section']}: {section['section_title']}\n{section['section_desc']}"
    metadata = {
        "chapter": str(section["chapter"]),
        "chapter_title": section["chapter_title"],
        "section": str(section["Section"]),
        "section_title": section["section_title"]
    }
    docs.append(Document(page_content=content, metadata=metadata))

   embeddings = OpenAIEmbeddings(openai_api_key=openapi_key)
  # store all IPC section Articles here
   vectorstore = Chroma.from_documents(docs, embeddings, persist_directory="./ChromaIPC_db")
  except Exception as er:
    print("--exception---",er)
    return jsonify("expcetion occured while processing...") 
  

  return "IPC sections are loaded sucessfully"

def trainpdf1(fpath1):

   print("- fpath1---",fpath1)
   try:
    filepath = os.path.join(app.config['UPLOAD_FOLDER'], fpath1.filename)
    fpath1.save(filepath)
    embeddings = OpenAIEmbeddings(openai_api_key=openapi_key)
    loader = PyPDFLoader(filepath)
    pages = loader.load()  # Returns list of Document objects

   
    court_text = "\n".join([page.page_content for page in pages])

    # Wrap as LangChain document
    doc = Document(
    page_content=court_text,
    metadata={"source": "court order"}
    )

    print("- filepath---",filepath)

    #text_splitter = RecursiveCharacterTextSplitter(
    #    chunk_size=800,
    #    chunk_overlap=200
    #)
    #documents = text_splitter.split_documents(pages)


    # store all court order documents here
    vectorstore = Chroma.from_documents([doc], embeddings, persist_directory="./ChromaCOURT_db")
    vectorstore.persist()    
    
    llm = ChatOpenAI(model="gpt-4", temperature=0)

    prompt = PromptTemplate.from_template("""
    You are a legal assistant. Given the following court order, list the top 5 relevant legal issues or areas that this case involves (e.g., property rights, public nuisance, fundamental rights, illegal construction, etc.)

    Court Order:
    {order}

    List 5 legal areas:
    """)
    
    chain = LLMChain(llm=llm, prompt=prompt)
    response = chain.run(order=court_text)
 
    IPCsearch = Chroma(persist_directory="./ChromaIPC_db", embedding_function=embeddings )

   
    areas = [area.strip("1234567890. ").strip() for area in response.split('\n') if area.strip()]
    ipc_matches = []
    ipc_results = []

    for area in areas[:5]:  # limit to top 5 areas
     results = IPCsearch.similarity_search(area, k=1)
    if results:
        ipc_matches.append((area, results[0]))
        
    # 🖨️ Print matched IPC sections
    for topic, doc in ipc_matches:
     print(f"\n📘 Legal Area: {topic}")
     print(f"🔗 IPC Section: {doc.metadata.get('section')} - {doc.metadata.get('section_title')}")
     print(f"📄 Description: {doc.page_content}")

    for topic, doc in ipc_matches:
     ipc_results.append({
        "legal_area": topic,
        "ipc_section": doc.metadata.get("section"),
        "section_title": doc.metadata.get("section_title"),
        "description": doc.page_content,
        "Orderdocumentation":response
    })


    
   except Exception as er:
    print("--exception---",er)
    return jsonify("This pdf cannot be trained") 
   

   return ipc_results

def getdata(query):
 
 embeddings = OpenAIEmbeddings(openai_api_key=openapi_key)
 os.environ['OPENAI_API_KEY'] = openapi_key

 your_case_db = Chroma(persist_directory="./ChromaIPC_db", embedding_function=embeddings)
 your_case_text = your_case_db.similarity_search("relavant IPC section for bribe", k=1)[0].page_content

 print("---your_case_text-----",your_case_text)

# Load SC case database
 supreme_db = Chroma(persist_directory="./MiyapurCase_db", embedding_function=embeddings)
 retriever = supreme_db.as_retriever(search_kwargs={"k": 5})

# Ask for relevant judgments
 llm = ChatOpenAI(model="gpt-4", temperature=1)
 qa_chain = RetrievalQA.from_chain_type(llm=llm, retriever=retriever)

 response = qa_chain.run(query)




 return response
   
 

    



        



if __name__ == '__main__':
    app.run(port=8080)