I've encountered this issue a few times while developing an app that uses PyPartPicker, but it has usually resolved itself when I fixed other problems. But now I'm in a bit of a stalemate.
I've just implemented proxy rotation and asyncio functions as per the documentation suggests, and it seems to be working fine. I moved the proxy debugging to the main function to test the proxies, but had to move it back to the response retreiver since it just pooped itself with the debugging in the main function for some reason.
Anyway, it goes through about 10 parts and then just hangs on
2025-01-20 19:15:56,761 - INFO - Browser listening on:
ws://127.0.0.1:59948/devtools/browser/f358eb04-9c23-4952-b9c6-a92e91b1fe9b
I know it's a localhost loopback ip, but I just can't see why it hangs. It should be retrying to connect if the connection is unsuccessful!
I've scoured pretty much all of StackOverflow and the RequestsHTML documentation, but I can't find anything on this issue at all. At first I thought it had something to do with rate limiting, but now I'm having second thoughts. I left it sit for about 45 miniutes, and eventually it closed the connection.
Then I though that perhaps it was something to do with the the chromium processes not terminating after completing the scrape, so I added some logic that closes it manually. But that didn't do anything. And I don't think it the custom response retriever either.
Anyways, I don't know if this is a library wide thing or just a issue with my program, but since nothing else seems to have any answers I thought i'd give it a shot here. Let me know if anyone has any ideas :)
Program below:
import pypartpicker
import time
import random
from supabase import create_client, Client
import logging
from requests.exceptions import HTTPError
import requests_html
import asyncio
from itertools import cycle
from contextlib import closing
proxy_list = [
"socks5://**CENSORED**:**CENSORED**@198.23.239.134:6540",
"socks5://**CENSORED**:**CENSORED**@207.244.217.165:6712",
"socks5://**CENSORED**:**CENSORED**@107.172.163.27:6543",
"socks5://**CENSORED**:**CENSORED**@64.137.42.112:5157",
"socks5://**CENSORED**:**CENSORED**@173.211.0.148:6641",
"socks5://**CENSORED**:**CENSORED**@161.123.152.115:6360",
"socks5://**CENSORED**:**CENSORED**[email protected]:6754",
"socks5://**CENSORED**:**CENSORED**@154.36.110.199:6853",
"socks5://**CENSORED**:**CENSORED**@173.0.9.70:5653",
"socks5://**CENSORED**:**CENSORED**@173.0.9.209:5792",
]
proxy_cycle = cycle(proxy_list)
session = requests_html.HTMLSession()
# session.browser.args = ["--no-sandbox", "--disable-setuid-sandbox", "--disable-dev-shm-usage"]
session.browser.timeout = 5
# Setup logging
logging.basicConfig(level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s")
# Initialize Supabase client
url = "https://*CENSORED*.supabase.co"
key = "*CENSORED*"
try:
supabase: Client = create_client(url, key)
logging.info("Connected to Supabase successfully.")
except Exception as e:
logging.error(f"Failed to connect to Supabase: {e}")
exit(1)
# Supported Product Types
CPU_PRODUCTS = ["Intel Core i3", "Intel Core i5", "Intel Core i7", "Intel Core i9", "Intel Xeon", "AMD Ryzen 3", "AMD Ryzen 5", "AMD Ryzen 7", "AMD Ryzen 9", "AMD Threadripper", "AMD Athlon 3000G", "AMD Athlon 200GE", "Intel Pentium G6400", "Intel Pentium 5", "Intel Pentium 6", "Intel Pentium 7", "Intel Celeron G4", "Intel Celeron G5", "Intel Celeron G6"]
def response_retriever(url):
retries = 5
backoff_time = 2 # Initial backoff in seconds
proxy_api_url = "/latest/ip"
for _ in range(retries):
proxy = next(proxy_cycle) # Rotate proxy
try:
logging.info(f"Attempting to use proxy {proxy} for {url}")
# Verify outgoing IP using the external service
proxy_check_response = session.get(
proxy_api_url, proxies={"http": proxy, "https": proxy}, timeout=10
)
if proxy_check_response.status_code != 200:
logging.warning(f"Failed to verify proxy {proxy}: Status {proxy_check_response.status_code}")
continue
try:
proxy_data = proxy_check_response.json()
outgoing_ip = proxy_data.get("ip")
if not outgoing_ip:
logging.warning(f"Proxy {proxy} did not return a valid IP.")
continue
except Exception as e:
logging.error(f"Error parsing proxy verification response: {e}")
continue
with closing(session.get(url, proxies={"http": proxy, "https": proxy})) as response:
# Verify proxy works by making the actual request
response = session.get(url, proxies={"http": proxy, "https": proxy})
if response.status_code == 200:
logging.info(f"Successfully retrieved response from {url} using proxy {outgoing_ip}")
return response
elif response.status_code == 429:
logging.warning(f"Rate limited (429) for {url}. Retrying in {backoff_time}s...")
time.sleep(backoff_time)
backoff_time *= random.uniform(1.5, 2.0) # Exponential backoff
else:
logging.warning(f"Unexpected status code {response.status_code} for {url}")
return None
except requests_html.RequestException as e:
logging.error(f"Proxy {proxy} failed for {url}: {e}")
time.sleep(backoff_time)
backoff_time *= random.uniform(1.5, 2.0) # Exponential backoff
logging.error(f"Failed to retrieve a valid response after {retries} attempts for {url}")
return None
# Initialize PyPartPicker Client with the custom response retriever
pcpp = pypartpicker.Client(response_retriever=response_retriever)
async def fetch_top_parts():
async with pypartpicker.AsyncClient() as pcpp:
# Counters for debugging
error_count = 0
warning_count = 0
skipped_parts = []
# Iterate through each product type and fetch all results
for product in CPU_PRODUCTS:
page = 1
while True:
try:
logging.info(f"Fetching {product} parts on page {page}...")
result = await pcpp.get_part_search(product, page=page, region="au")
if result and result.parts:
for part_summary in result.parts:
if part_summary and part_summary.url:
while True:
try:
proxy = next(proxy_cycle)
logging.info(f"Using proxy {proxy} for {part_summary.url}")
part = await pcpp.get_part(part_summary.url)
if part:
# Validate and prepare data for insertion
in_stock_vendors = [
vendor for vendor in part.vendors if vendor.in_stock
] if part.vendors else []
in_stock_vendors.sort(key=lambda v: v.price.total if v.price else float('inf'))
cheapest_vendor = in_stock_vendors[0] if in_stock_vendors else None
data = {
"part_type": "processor",
"name": part.name if part.name else None,
"total_price": cheapest_vendor.price.total if cheapest_vendor and cheapest_vendor.price else None,
"base_price": cheapest_vendor.price.base if cheapest_vendor and cheapest_vendor.price else None,
"discounts": cheapest_vendor.price.discounts if cheapest_vendor and cheapest_vendor.price else None,
"shipping_price": cheapest_vendor.price.shipping if cheapest_vendor and cheapest_vendor.price else None,
"tax_price": cheapest_vendor.price.tax if cheapest_vendor and cheapest_vendor.price else None,
"vendor_store": getattr(cheapest_vendor, "name", "N/A") if cheapest_vendor else None,
"store_product_url": getattr(cheapest_vendor, "buy_url", "N/A") if cheapest_vendor else None,
"vendor_logo_url": getattr(cheapest_vendor, "logo_url", "N/A") if cheapest_vendor else None,
"in_stock": bool(in_stock_vendors) and cheapest_vendor is not None,
"product_url": getattr(part, "url", "N/A"),
"image_urls": part.image_urls if part.image_urls else None,
"manufacturer": part.specs.get("Manufacturer", None) if part.specs else None,
"part_number": part.specs.get("Part #", None) if part.specs else None,
"series": part.specs.get("Series", None) if part.specs else None,
"microarchitecture": part.specs.get("Microarchitecture", None) if part.specs else None,
"core_family": part.specs.get("Core Family", None) if part.specs else None,
"socket": part.specs.get("Socket", None) if part.specs else None,
"core_count": part.specs.get("Core Count", None) if part.specs else None,
"thread_count": part.specs.get("Thread Count", None) if part.specs else None,
"performance_core_clock": part.specs.get("Performance Core Clock", None) if part.specs else None,
"performance_core_boost_clock": part.specs.get("Performance Core Boost Clock", None) if part.specs else None,
"l2_cache": part.specs.get("L2 Cache", None) if part.specs else None,
"l3_cache": part.specs.get("L3 Cache", None) if part.specs else None,
"tdp": part.specs.get("TDP", None) if part.specs else None,
"integrated_graphics": part.specs.get("Integrated Graphics", None) if part.specs else None,
"maximum_supported_memory": part.specs.get("Maximum Supported Memory", None) if part.specs else None,
"ecc_support": part.specs.get("ECC Support", None) if part.specs else None,
"includes_cooler": part.specs.get("Includes Cooler", None) if part.specs else None,
"packaging": part.specs.get("Packaging", None) if part.specs else None,
"lithography": part.specs.get("Lithography", None) if part.specs else None,
"simultaneous_multithreading": part.specs.get("Simultaneous Multithreading", None) if part.specs else None,
"rating_average": getattr(part.rating, "average", None) if part.rating else None,
}
try:
response = supabase.table("cpus").insert([data]).execute()
logging.info(f"Inserted {data['name']} into database.")
except Exception as e:
logging.error(f"Failed to insert {data['name']} into database: {e}")
else:
warning_count += 1
logging.warning("Part details could not be fetched.")
break # Exit the retry loop if successful
except AttributeError as e:
if "'NoneType' object has no attribute 'text'" in str(e):
input("Verify link and press Enter to continue...")
else:
raise e
except Exception as e:
error_count += 1
logging.error(f"Error fetching part details: {e}")
break
else:
logging.info(f"No more results for {product} on page {page}.")
break # Exit loop if no more results
page += 1
await asyncio.sleep(4) # Prevent hitting rate limits
except HTTPError as e:
error_count += 1
logging.error(f"HTTP error occurred: {e}")
await asyncio.sleep(10) # Wait before retrying
except Exception as e:
error_count += 1
logging.error(f"Unexpected error: {e}")
await asyncio.sleep(10) # Short wait before retrying
continue
# Final Debug Summary
logging.info("\nDebug Summary:")
logging.info(f"Total Errors: {error_count}")
logging.info(f"Total Warnings: {warning_count}")
logging.info(f"Total Skipped Parts: {len(skipped_parts)}")
if skipped_parts:
for name, part_number in skipped_parts:
logging.info(f"Skipped Part: {name} | Part Number: {part_number}")
# Run the main function
asyncio.run(fetch_top_parts())