app.py

from flask import Flask, request, jsonify
import requests
import random
import string
import time

app = Flask(__name__)

# Authorization value used in requests (update with a valid Authorization)
TOKEN = "Bearer bp_pat_vTuxol25N0ymBpYaWqtWpFfGPKt260IfT784"

# -------------------------------------------------------------------
# Helper functions
# -------------------------------------------------------------------
def generate_random_name(length=5):
    return ''.join(random.choices(string.ascii_letters, k=length))

# -------------------------------------------------------------------
# Workspace and Bot management
# -------------------------------------------------------------------
def create_workspace():
    url = "https://api.botpress.cloud/v1/admin/workspaces"
    headers = {
        "User-Agent": "Mozilla/5.0",
        "Authorization": TOKEN
    }
    payload = {"name": generate_random_name()}
    try:
        r = requests.post(url, headers=headers, json=payload, timeout=60)
        if r.status_code == 200:
            ws_id = r.json().get("id")
            if ws_id:
                print(f"Workspace created: {ws_id}")
                return ws_id
        print(f"Workspace creation failed: {r.status_code} {r.text}")
    except Exception as e:
        print(f"Error creating workspace: {e}")
    return None

def create_bot(workspace_id):
    if not workspace_id:
        print("No workspace_id provided for bot creation")
        return None
    url = "https://api.botpress.cloud/v1/admin/bots"
    headers = {
        "User-Agent": "Mozilla/5.0",
        "x-workspace-id": workspace_id,
        "Authorization": TOKEN,
        "Content-Type": "application/json"
    }
    payload = {"name": generate_random_name()}
    try:
        r = requests.post(url, headers=headers, json=payload, timeout=60)
        if r.status_code == 200:
            bot_id = r.json().get("bot", {}).get("id")
            if bot_id:
                print(f"Bot created: {bot_id} in workspace: {workspace_id}")
                # Optionally install integration (same as original)
                installed = install_bot_integration(bot_id, workspace_id)
                if installed:
                    print(f"Integration installed for bot {bot_id}")
                else:
                    print(f"Integration install failed for bot {bot_id} (continuing)")
                return bot_id
        print(f"Bot creation failed: {r.status_code} {r.text}")
    except Exception as e:
        print(f"Error creating bot: {e}")
    return None

def install_bot_integration(bot_id, workspace_id):
    if not bot_id or not workspace_id:
        print("Missing bot_id or workspace_id for integration")
        return False
    url = f"https://api.botpress.cloud/v1/admin/bots/{bot_id}"
    headers = {
        "User-Agent": "Mozilla/5.0",
        "Authorization": TOKEN,
        "Content-Type": "application/json",
        "x-bot-id": bot_id,
        "x-workspace-id": workspace_id
    }
    payload = {
        "integrations": {
            "intver_01JZ6J0NKYBXC0V6K5DBDZKKDK": {
                "enabled": True
            }
        }
    }
    try:
        r = requests.put(url, headers=headers, json=payload, timeout=60)
        if r.status_code == 200:
            return True
        print(f"Integration install failed: {r.status_code} {r.text}")
    except Exception as e:
        print(f"Error installing integration: {e}")
    return False

def delete_bot(bot_id, workspace_id):
    if not bot_id or not workspace_id:
        print("Missing bot_id or workspace_id for bot deletion")
        return False
    url = f"https://api.botpress.cloud/v1/admin/bots/{bot_id}"
    headers = {
        "User-Agent": "Mozilla/5.0",
        "x-workspace-id": workspace_id,
        "Authorization": TOKEN
    }
    try:
        r = requests.delete(url, headers=headers, timeout=60)
        if r.status_code in [200, 204]:
            print(f"Bot deleted: {bot_id}")
            return True
        print(f"Bot deletion failed: {r.status_code} {r.text}")
    except Exception as e:
        print(f"Error deleting bot: {e}")
    return False

def delete_workspace(workspace_id):
    if not workspace_id:
        print("Missing workspace_id for workspace deletion")
        return False
    url = f"https://api.botpress.cloud/v1/admin/workspaces/{workspace_id}"
    headers = {
        "User-Agent": "Mozilla/5.0",
        "Authorization": TOKEN
    }
    try:
        r = requests.delete(url, headers=headers, timeout=60)
        if r.status_code in [200, 204]:
            print(f"Workspace deleted: {workspace_id}")
            return True
        print(f"Workspace deletion failed: {r.status_code} {r.text}")
    except Exception as e:
        print(f"Error deleting workspace: {e}")
    return False

# -------------------------------------------------------------------
# Capture Screenshot (Botpress chat action)
# -------------------------------------------------------------------
def capture_screenshot(url, width, height, full_page, bot_id, workspace_id, timeout=120):
    if not bot_id or not workspace_id:
        return {"error": "Missing bot_id or workspace_id"}

    api_url = "https://api.botpress.cloud/v1/chat/actions"
    headers = {
        "User-Agent": "Mozilla/5.0",
        "x-bot-id": bot_id,
        "x-workspace-id": workspace_id,
        "Content-Type": "application/json",
        "Authorization": TOKEN
    }
    payload = {
        "type": "browser:captureScreenshot",
        "input": {
            "url": url,
            "cssToInject": "greregreg",
            "javascriptToInject": "gerregr",
            "width": width,
            "height": height,
            "fullPage": bool(full_page),
            "waitFor": 10000
        }
    }

    max_retries = 3
    for attempt in range(max_retries):
        try:
            print(f"Capture attempt {attempt+1} for {url}")
            r = requests.post(api_url, headers=headers, json=payload, timeout=timeout)
            if r.status_code == 200:
                data = r.json()
                # Expecting:
                # { "output": {"imageUrl": "...", "htmlUrl": "..."}, "meta": {"cached": false}}
                return data
            elif r.status_code in [404, 408, 502, 503, 504]:
                print(f"Transient error {r.status_code}: retrying...")
                time.sleep(2)
                continue
            elif r.status_code == 403:
                print(f"Forbidden (403): {r.text}")
                # Not rotating bot/workspace here; per-request lifecycle handles cleanup
                return {"error": "Forbidden", "status": 403, "details": safe_text(r)}
            else:
                print(f"Unexpected {r.status_code}: {r.text}")
                return {"error": "Unexpected error", "status": r.status_code, "details": safe_text(r)}
        except requests.exceptions.Timeout:
            print("Request timed out, retrying...")
            if attempt < max_retries - 1:
                continue
            return {"error": "Timeout"}
        except Exception as e:
            print(f"Exception: {e}")
            if attempt < max_retries - 1:
                time.sleep(1)
                continue
            return {"error": str(e)}

    return {"error": "Failed after retries"}

def safe_text(response):
    try:
        return response.text
    except Exception:
        return "<unavailable>"

# -------------------------------------------------------------------
# Flask Endpoint
# -------------------------------------------------------------------
@app.route("/capture", methods=["POST"])
def capture_endpoint():
    """
    JSON body example:
    {
      "urls": ["https://x.com/USTreasury", "https://example.com"],   // required, array of one or more
      "width": 10850,                                                // optional, default 1080
      "height": 1920,                                                // optional, default 1920
      "fullPage": true                                               // optional, default true
    }
    Response:
    {
      "results": [
        {
          "url": "...",
          "output": {
            "imageUrl": "...png",
            "htmlUrl": "...html"
          },
          "meta": { "cached": false }
        },
        { ... }
      ],
      "errors": [
        { "url": "...", "error": "...", "status": 403, "details": "..." }
      ]
    }
    Behavior:
    - Create workspace, then bot.
    - Run capture for all URLs with the same bot/workspace.
    - Delete bot first, then workspace, at the end of the request.
    """
    # Parse and validate body
    try:
        body = request.get_json(force=True) or {}
    except Exception:
        return jsonify({"error": "Invalid JSON"}), 400

    urls = body.get("urls")
    if not urls or not isinstance(urls, list):
        return jsonify({"error": "urls is required and must be a list"}), 400

    # Optional params
    width = body.get("width", 1080)
    height = body.get("height", 1920)
    full_page = body.get("fullPage", True)

    # Normalize numeric params
    try:
        width = int(width)
        if width <= 0:
            width = 1080
    except Exception:
        width = 1080
    try:
        height = int(height)
        if height <= 0:
            height = 1920
    except Exception:
        height = 1920
    full_page = bool(full_page)

    workspace_id = None
    bot_id = None

    # Create (Workspace then Bot)
    workspace_id = create_workspace()
    if not workspace_id:
        return jsonify({"error": "Failed to create workspace"}), 500

    bot_id = create_bot(workspace_id)
    if not bot_id:
        # Cleanup: if bot creation fails, delete workspace
        delete_workspace(workspace_id)
        return jsonify({"error": "Failed to create bot"}), 500

    results = []
    errors = []

    # Process multiple URLs using the same bot/workspace
    for url in urls:
        if not isinstance(url, str) or not url.strip():
            errors.append({"url": url, "error": "Invalid URL"})
            continue

        capture = capture_screenshot(
            url=url.strip(),
            width=width,
            height=height,
            full_page=full_page,
            bot_id=bot_id,
            workspace_id=workspace_id
        )

        if isinstance(capture, dict) and capture.get("error"):
            errors.append({"url": url, **capture})
        else:
            # Ensure consistent output structure
            output = capture.get("output", {})
            meta = capture.get("meta", {})
            results.append({
                "url": url,
                "output": {
                    "imageUrl": output.get("imageUrl"),
                    "htmlUrl": output.get("htmlUrl")
                },
                "meta": meta
            })

    # Delete (Bot first, then Workspace) regardless of success
    try:
        if bot_id:
            delete_bot(bot_id, workspace_id)
    finally:
        if workspace_id:
            delete_workspace(workspace_id)

    response_body = {"results": results}
    if errors:
        response_body["errors"] = errors

    # If all failed, return 502; else 200
    if results:
        return jsonify(response_body), 200
    else:
        return jsonify(response_body), 502

# -------------------------------------------------------------------
# Run the Flask app
# -------------------------------------------------------------------
if __name__ == "__main__":
    app.run(host="0.0.0.0", port=7860, debug=True)