Brancheneinstufung2/devtools/push_mh_to_notion.py

import sys
import os
import csv
import requests
import json
import logging
from collections import defaultdict

# Setup Logger
logging.basicConfig(level=logging.INFO, format='%(levelname)s: %(message)s')
logger = logging.getLogger("NotionPusher")

# Config
CSV_PATH = "./docs/miller_heiman_augmented.csv"
NOTION_TOKEN_FILE = "/app/notion_token.txt"
INDUSTRIES_DB_ID = "2ec88f4285448014ab38ea664b4c2b81"

def load_notion_token():
    try:
        with open(NOTION_TOKEN_FILE, "r") as f:
            return f.read().strip()
    except:
        logger.error("Token file not found.")
        sys.exit(1)

def add_strategy_property(token):
    """Try to add the 'Strategy Briefing' property to the database schema."""
    url = f"https://api.notion.com/v1/databases/{INDUSTRIES_DB_ID}"
    headers = {
        "Authorization": f"Bearer {token}",
        "Notion-Version": "2022-06-28",
        "Content-Type": "application/json"
    }
    
    # First, get current schema to check if it exists
    resp = requests.get(url, headers=headers)
    if resp.status_code != 200:
        logger.error(f"Failed to fetch DB schema: {resp.text}")
        return False
        
    current_properties = resp.json().get("properties", {})
    if "Strategy Briefing" in current_properties:
        logger.info("Property 'Strategy Briefing' already exists.")
        return True
        
    logger.info("Property 'Strategy Briefing' missing. Attempting to create...")
    
    payload = {
        "properties": {
            "Strategy Briefing": {
                "rich_text": {}
            }
        }
    }
    
    update_resp = requests.patch(url, headers=headers, json=payload)
    if update_resp.status_code == 200:
        logger.info("Successfully added 'Strategy Briefing' property.")
        return True
    else:
        logger.error(f"Failed to add property: {update_resp.text}")
        return False

def get_pages_map(token):
    """Returns a map of Vertical Name -> Page ID"""
    url = f"https://api.notion.com/v1/databases/{INDUSTRIES_DB_ID}/query"
    headers = {
        "Authorization": f"Bearer {token}",
        "Notion-Version": "2022-06-28",
        "Content-Type": "application/json"
    }
    
    mapping = {}
    has_more = True
    next_cursor = None
    
    while has_more:
        payload = {}
        if next_cursor:
            payload["start_cursor"] = next_cursor
        
        resp = requests.post(url, headers=headers, json=payload)
        if resp.status_code != 200:
            break
            
        data = resp.json()
        for page in data.get("results", []):
            props = page.get("properties", {})
            name_parts = props.get("Vertical", {}).get("title", [])
            if name_parts:
                name = "".join([t.get("plain_text", "") for t in name_parts])
                mapping[name] = page["id"]
                
        has_more = data.get("has_more", False)
        next_cursor = data.get("next_cursor")
        
    return mapping

def update_page(token, page_id, pains, gains, strategy):
    url = f"https://api.notion.com/v1/pages/{page_id}"
    headers = {
        "Authorization": f"Bearer {token}",
        "Notion-Version": "2022-06-28",
        "Content-Type": "application/json"
    }
    
    # Construct Payload
    props = {}
    
    # Only update if content exists (safety)
    if pains:
        props["Pains"] = {"rich_text": [{"text": {"content": pains[:2000]}}]} # Limit to 2000 chars to be safe
    if gains:
        props["Gains"] = {"rich_text": [{"text": {"content": gains[:2000]}}]}
    if strategy:
        props["Strategy Briefing"] = {"rich_text": [{"text": {"content": strategy[:2000]}}]}
        
    if not props:
        return
        
    payload = {"properties": props}
    resp = requests.patch(url, headers=headers, json=payload)
    
    if resp.status_code != 200:
        logger.error(f"Failed to update page {page_id}: {resp.text}")
    else:
        logger.info(f"Updated page {page_id}")

def main():
    token = load_notion_token()
    
    # 1. Ensure Schema
    if not add_strategy_property(token):
        logger.warning("Could not add 'Strategy Briefing' column. Please add it manually in Notion as a 'Text' property.")
        # We continue, maybe it failed because of permissions but column exists?
        # Actually if it failed, the update later will fail too if key is missing.
        # But let's try.
        
    # 2. Map Notion Verticals
    logger.info("Mapping existing Notion pages...")
    notion_map = get_pages_map(token)
    logger.info(f"Found {len(notion_map)} verticals in Notion.")
    
    # 3. Read CSV
    logger.info(f"Reading CSV: {CSV_PATH}")
    csv_data = {}
    
    with open(CSV_PATH, "r", encoding="utf-8-sig") as f:
        reader = csv.DictReader(f, delimiter=";")
        for row in reader:
            vertical = row.get("Vertical")
            if not vertical:
                continue
            
            # Aggregate Strategy
            parts = []
            if row.get("MH Coach Hypothesis"):
                parts.append(f"🧠 MH Coach: {row.get('MH Coach Hypothesis')}")
            if row.get("MH Early Red Flags"):
                parts.append(f"🚩 Red Flags: {row.get('MH Early Red Flags')}")
            if row.get("MH Adjustments / Ergänzungen"):
                parts.append(f"🔧 Adjustments: {row.get('MH Adjustments / Ergänzungen')}")
            
            strategy = "\n\n".join(parts)
            
            # Store (last row wins if duplicates, but they should be identical for vertical data)
            csv_data[vertical] = {
                "pains": row.get("Pains (clean)", ""),
                "gains": row.get("Gains (clean)", ""),
                "strategy": strategy
            }
            
    # 4. Push Updates
    logger.info("Starting Batch Update...")
    count = 0
    for vertical, data in csv_data.items():
        if vertical in notion_map:
            page_id = notion_map[vertical]
            logger.info(f"Updating '{vertical}'...")
            update_page(token, page_id, data["pains"], data["gains"], data["strategy"])
            count += 1
        else:
            logger.warning(f"Skipping '{vertical}' (Not found in Notion)")
            
    logger.info(f"Finished. Updated {count} verticals.")

if __name__ == "__main__":
    main()