app.py

import requests
import os
import json
from fastapi import FastAPI, HTTPException
from fastapi.responses import FileResponse

app = FastAPI()

ALL_COOKIES = {}
print("loadaaadingggg")

REQUIRED_COOKIES = [
    'ka_sessionid',
    '__Host-KAGGLEID',
    'CLIENT-TOKEN',
    'XSRF-TOKEN',
    'CSRF-TOKEN',
    'build-hash'
]

for key, value in os.environ.items():
    if key.startswith("KAGGLE_COOKIES_") and key.rsplit('_', 1)[-1].isdigit():
        account_id = key.rsplit('_', 1)[-1]
        account_cookies = {}
        netscape_file_content = value

        for line in netscape_file_content.splitlines():
            if not line.strip() or line.startswith('#'):
                continue

            parts = line.split('\t')
            if len(parts) == 7:
                cookie_name = parts[5]
                cookie_value = parts[6]
                if cookie_name in REQUIRED_COOKIES:
                    account_cookies[cookie_name] = cookie_value

        if all(c in account_cookies for c in REQUIRED_COOKIES):
            ALL_COOKIES[account_id] = account_cookies
            print(f"negro caca '{account_id}' passed.")
        else:
            missing = [c for c in REQUIRED_COOKIES if c not in account_cookies]
            print(f"negro '{account_id}' poo poo'd. needs: {missing}")

if not ALL_COOKIES:
    print("retardation alert 1")

@app.post("/get-file-list")
async def get_file_list(request_body: dict):
    try:
        account_id = str(request_body.get("accountId"))
        resource_type = request_body.get("resourceType")
        metadata = request_body.get("metadata")
        cookies = ALL_COOKIES.get(account_id)
        if not cookies: raise HTTPException(status_code=404, detail=f"ass '{account_id}' not found")

        api_url = "https://www.kaggle.com/api/i/datasets.databundles.DatabundleService/GetDatabundleExternal"

        if resource_type == 'dataset':
            body = {"verificationInfo": {"databundleVersionId": metadata["versionId"], "datasetId": metadata["id"]}}
        else: 
            body = {"verificationInfo": {"databundleVersionId": metadata["versionId"], "modelInstanceVersionId": metadata["instanceVersionId"]}}

        headers = {
            "accept": "application/json", "content-type": "application/json",
            "origin": "https://www.kaggle.com", "referer": "https://www.kaggle.com/",
            "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/123.0.0.0 Safari/537.36",
            "x-kaggle-build-version": cookies["build-hash"], "x-xsrf-token": cookies["XSRF-TOKEN"]
        }

        response = requests.post(api_url, headers=headers, cookies=cookies, json=body)
        response.raise_for_status()
        data = response.json()

        files = data.get('dataSource', {}).get('databundleVersion', {}).get('filesetInfo', {}).get('files', {}).get('children', [])
        file_list = [{"name": f['name'], "firestorePath": f['path']} for f in files if f['name'].lower().endswith(('.mp4', '.mkv', '.mov', '.avi'))]

        return { "files": file_list } 
    except requests.exceptions.HTTPError as err:
        error_detail = f"{str(err)} - Response: {err.response.text}"
        raise HTTPException(status_code=err.response.status_code, detail=error_detail)
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))

@app.post("/get-download-url")
async def get_download_url(request_body: dict):
    try:
        account_id = str(request_body.get("accountId"))
        resource_type = request_body.get("resourceType")
        metadata = request_body.get("metadata")
        file_info = request_body.get("file")
        cookies = ALL_COOKIES.get(account_id)
        if not cookies: raise HTTPException(status_code=404, detail=f"Account ID '{account_id}' not found or cookies not set.")

        api_url = "https://www.kaggle.com/api/i/datasets.DatasetService/GetDataViewExternal"

        if resource_type == 'dataset':
            body = {"verificationInfo": {"datasetId": metadata["id"], "databundleVersionId": metadata["versionId"]}, "firestorePath": file_info["firestorePath"]}
        else: 
            body = {"verificationInfo": {"modelInstanceVersionId": metadata["instanceVersionId"], "databundleVersionId": metadata["versionId"]}, "firestorePath": file_info["firestorePath"]}

        headers = {
            "accept": "application/json", "content-type": "application/json",
            "origin": "https://www.kaggle.com", "referer": "https://www.kaggle.com/",
            "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/123.0.0.0 Safari/537.36",
            "x-kaggle-build-version": cookies["build-hash"], "x-xsrf-token": cookies["XSRF-TOKEN"]
        }

        response = requests.post(api_url, headers=headers, cookies=cookies, json=body)
        response.raise_for_status()
        data = response.json()
        return {"direct_url": data['dataView']['dataUrl']['url']}
    except requests.exceptions.HTTPError as err:
        error_detail = f"{str(err)} - Response: {err.response.text}"
        raise HTTPException(status_code=err.response.status_code, detail=error_detail)
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))

@app.get("/")
async def read_index():
    return FileResponse('index.html')