.

app.py

@@ -1,143 +0,0 @@
-import os
-import uuid
-import logging
-import requests
-import gradio as gr
-
-from langdetect import detect
-from multilingual_pdf2text.pdf2text import PDF2Text
-from multilingual_pdf2text.models.document_model.document import Document
-
-import pdfplumber
-import docx2txt
-import shutil
-import subprocess
-
-print("🔍 Checking if Tesseract OCR is installed...")
-
-tess_path = shutil.which("tesseract")
-if tess_path:
-    print(f"✅ Tesseract found at: {tess_path}")
-    try:
-        version = subprocess.check_output(["tesseract", "--version"]).decode()
-        print(f"🧠 Tesseract version:\n{version}")
-    except Exception as e:
-        print(f"⚠️ Tesseract found but failed to run: {e}")
-else:
-    print("❌ Tesseract not found. Arabic PDF extraction will likely fail.")
-
-
-# --- Config ---
-FLOWISE_URL = "https://siraflowappl.happywave-0e1819cd.uaenorth.azurecontainerapps.io/api/v1/prediction/5f2aa074-ae2e-4b06-a4a2-6f8dccbe59bb"
-COUNTRIES = ["USA", "Germany", "UAE", "Turkey", "Jordan"]
-LANGUAGES = ["English", "Arabic"]
-
-logging.basicConfig(level=logging.INFO)
-
-# --- File Text Extraction ---
-def extract_text_from_file(file_path: str) -> str:
-    ext = os.path.splitext(file_path)[1].lower()
-
-    try:
-        if ext == ".pdf":
-            with pdfplumber.open(file_path) as pdf:
-                extracted = "\n".join(
-                    page.extract_text() for page in pdf.pages if page.extract_text()
-                )
-
-            sample = extracted[:1000].strip()
-            lang = detect(sample) if sample else "unknown"
-
-            if lang == "ar" or not sample:
-                arabic_doc = Document(document_path=file_path, language="ara")
-                pdf2text = PDF2Text(document=arabic_doc)
-                extracted = pdf2text.extract()
-
-                if isinstance(extracted, list):
-                    if isinstance(extracted[0], dict):
-                        return "\n".join(item.get("text", "") for item in extracted)
-                    elif isinstance(extracted[0], str):
-                        return "\n".join(extracted)
-                    else:
-                        return "[Error: Unexpected Arabic extractor format.]"
-                elif isinstance(extracted, str):
-                    return extracted
-                else:
-                    return "[Error: Unrecognized Arabic extraction output.]"
-
-            return extracted or "[Warning: No readable text found in PDF.]"
-
-        elif ext == ".docx":
-            return docx2txt.process(file_path)
-
-        elif ext == ".txt":
-            with open(file_path, "r", encoding="utf-8", errors="ignore") as f:
-                return f.read()
-
-        else:
-            return "[Unsupported file format.]"
-
-    except Exception as e:
-        logging.exception("Failed to extract text.")
-        return f"[Error reading file: {e}]"
-
-# --- Flowise Query ---
-def send_to_flowise(file, country: str, language: str) -> str:
-    if not file or not country or not language:
-        return "❗ Please upload a contract, select a country, and choose a response language."
-
-    try:
-        contract_text = extract_text_from_file(file.name).strip()
-        if not contract_text:
-            return "⚠️ No content extracted from the uploaded file."
-
-        prompt = f"""Please audit the following employment contract in light of {country} labor law.
-Only refer to the provided law document (do not use external knowledge).
-Is the contract compliant? Provide reasoning and highlight any issues, and suggest improvements.
-Be precise and don't miss any details in the contract.
-Please respond in {language}.
-<CONTRACT START>
-{contract_text}
-<CONTRACT END>"""
-
-        payload = {
-            "question": prompt,
-            "chatId": str(uuid.uuid4())
-        }
-
-        res = requests.post(FLOWISE_URL, json=payload, timeout=300)
-        logging.info("🔍 Flowise response: %s", res.text)
-
-        if res.status_code != 200:
-            return f"❌ Flowise error {res.status_code}: {res.text}"
-
-        return res.json().get("text", "⚠️ Flowise returned no text.")
-
-    except Exception as e:
-        logging.exception("Flowise request failed.")
-        return f"❌ Connection error: {e}"
-
-# --- Gradio UI ---
-with gr.Blocks(title="Contract Auditor") as demo:
-    gr.Markdown("## 📑 Contract Auditor")
-
-    with gr.Row():
-        country_dropdown = gr.Dropdown(choices=COUNTRIES, label="🌍 Select Country", interactive=True)
-        language_dropdown = gr.Dropdown(choices=LANGUAGES, label="🗣️ Select Response Language", interactive=True)
-
-    file_input = gr.File(
-        label="📎 Upload a contract file",
-        file_types=[".pdf", ".docx", ".txt"],
-        type="filepath"
-    )
-
-    submit_btn = gr.Button("📤 Audit Contract")
-    output = gr.Textbox(label="🧠 Audit Result", lines=12)
-
-    submit_btn.click(
-        fn=send_to_flowise,
-        inputs=[file_input, country_dropdown, language_dropdown],
-        outputs=output
-    )
-
-demo.launch(share=True)