Update app.py

app.py

@@ -2,106 +2,59 @@ import gradio as gr
 import os
 import json
 import requests
-import fitz  # PyMuPDF for PDFs
+from Pdf_utils import extract_text_from_pdf
+from ocr_utils import extract_text_from_image
+from PromptBuilder import build_prompt
 
 # ----------------------------
-# SECRETS FROM ENVIRONMENT
+# SECRETS
 # ----------------------------
 GROQ_API_KEY = os.environ.get("GROQ_API_KEY")
-OCR_API_KEY = os.environ.get("OCR_API_KEY")  # optional for image OCR
+OCR_API_KEY = os.environ.get("OCR_API_KEY")  # optional
 
 if not GROQ_API_KEY:
     raise ValueError("GROQ_API_KEY not found in environment variables.")
 
 # ----------------------------
-# OCR FUNCTION
-# ----------------------------
-def extract_text_from_image(file_path):
-    if not OCR_API_KEY:
-        return ""
-    url = "https://api.ocr.space/parse/image"
-    with open(file_path, "rb") as f:
-        r = requests.post(
-            url,
-            files={"file": f},
-            data={"apikey": OCR_API_KEY, "language": "eng"}
-        )
-    try:
-        return r.json()['ParsedResults'][0]['ParsedText']
-    except:
-        return ""
-
-# ----------------------------
-# PDF FUNCTION
-# ----------------------------
-def extract_text_from_pdf(file_path):
-    text = ""
-    try:
-        doc = fitz.open(file_path)
-        for page in doc:
-            text += page.get_text()
-    except:
-        text = ""
-    return text
-
-# ----------------------------
-# PROMPT BUILDER
-# ----------------------------
-def build_prompt(user_text, file_text, link_info, mentions, hashtags, custom_prompt, checked_suggestions):
-    prompt = "You are a professional LinkedIn content creator.\n\n"
-    prompt += f"Main Content:\n{user_text}\n{file_text}\n\n"
-    
-    if link_info:
-        prompt += f"Reference Link: {link_info.get('url')} ({link_info.get('description')})\n"
-    
-    if mentions:
-        for m in mentions:
-            prompt += f"Mention {m.get('name')} ({m.get('role')}): {m.get('purpose')} | Link: {m.get('linkedin_url')}\n"
-    
-    if hashtags:
-        prompt += f"Include hashtags: {', '.join(hashtags)}\n"
-    
-    if custom_prompt:
-        prompt += f"User Instructions: {custom_prompt}\n"
-    
-    if checked_suggestions:
-        prompt += f"Include suggestions: {', '.join(checked_suggestions)}\n"
-    
-    prompt += "\nWrite a LinkedIn-ready post in human-like, engaging, professional style."
-    
-    return prompt
-
-# ----------------------------
-# GENERATE POST FUNCTION (Groq GPT)
+# GENERATE POST FUNCTION
 # ----------------------------
 def generate_post(file, user_text, link_info, mentions, hashtags, custom_prompt, checked_suggestions):
     file_text = ""
+    # Extract text from uploaded file
     if file:
-        if file.name.endswith(".pdf"):
-            file_text = extract_text_from_pdf(file.name)
-        else:
-            file_text = extract_text_from_image(file.name)
-    
+        try:
+            if file.name.lower().endswith(".pdf"):
+                file_text = extract_text_from_pdf(file.name)
+            else:
+                file_text = extract_text_from_image(file.name, OCR_API_KEY)
+        except Exception as e:
+            return f"File processing error: {str(e)}"
+
     prompt = build_prompt(user_text, file_text, link_info, mentions, hashtags, custom_prompt, checked_suggestions)
-    
-    url = "https://api.groq.com/v1/generate"  # replace with actual Groq GPT endpoint
+
+    # ----------------------------
+    # Groq GPT API call
+    # ----------------------------
+    url = "https://api.groq.com/v1/models/groq-gpt-large/infer"  # replace with actual model-id
     headers = {
         "Authorization": f"Bearer {GROQ_API_KEY}",
         "Content-Type": "application/json"
     }
     data = {
-        "model": "groq-gpt-large",   # example, replace with available model
         "prompt": prompt,
         "max_tokens": 400
     }
-    
-    r = requests.post(url, headers=headers, json=data)
+
     try:
+        r = requests.post(url, headers=headers, json=data, timeout=60)
+        if r.status_code != 200:
+            return f"API error {r.status_code}: {r.text}"
         res_json = r.json()
-        if "choices" in res_json and len(res_json["choices"]) > 0:
-            return res_json["choices"][0].get("text", "")
-        elif "generated_text" in res_json:
+        # Safe parsing
+        if "generated_text" in res_json:
             return res_json["generated_text"]
+        elif "choices" in res_json and len(res_json["choices"]) > 0:
+            return res_json["choices"][0].get("text", "")
         else:
             return f"Error: Unexpected response format: {res_json}"
     except Exception as e:
@@ -112,7 +65,7 @@ def generate_post(file, user_text, link_info, mentions, hashtags, custom_prompt,
 # ----------------------------
 with gr.Blocks() as demo:
     gr.Markdown("## LinkedIn Post Generator (Groq GPT + Gradio)")
-    
+
     file_input = gr.File(label="Upload PDF/Image", type="filepath")
     user_text = gr.Textbox(label="Main Text", lines=3)
     link_input = gr.Textbox(label="Reference Link URL")
@@ -130,9 +83,9 @@ with gr.Blocks() as demo:
         "Make it professional"
     ])
     mentions_input = gr.Textbox(label="Mentions JSON (name, role, purpose, linkedin_url)")
-    
+
     output_text = gr.Textbox(label="Generated LinkedIn Post", lines=10)
-    
+
     def wrap(file, user_text, link_input, link_desc, hashtags, custom_prompt, suggestions, mentions):
         link_info = {"url": link_input, "description": link_desc} if link_input else None
         hashtags_list = [h.strip() for h in hashtags.split(",")] if hashtags else []
@@ -141,7 +94,7 @@ with gr.Blocks() as demo:
         except:
             mentions_list = []
         return generate_post(file, user_text, link_info, mentions_list, hashtags_list, custom_prompt, suggestions)
-    
+
     gr.Button("Generate Post").click(
         wrap,
         inputs=[file_input, user_text, link_input, link_desc, hashtags_input, custom_prompt_input, suggestions_input, mentions_input],