Created
July 26, 2024 14:02
-
-
Save ianrussel/60f466dd7aa91859d32f01f49d30778d to your computer and use it in GitHub Desktop.
amaozn query
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
def _request_product(asin): | |
print(f" asin {asin}") | |
retry_count = 0 | |
max_retries = 4 | |
while retry_count < max_retries: | |
try: | |
products = api.query(asin, domain='GB', rating=1, buybox=1, update=0) | |
break # If the request is successful, break out of the loop | |
except ReadTimeout: | |
retry_count += 1 | |
print(f"Request timed out for ASIN {asin}. Retrying {retry_count}/{max_retries}...") | |
if retry_count == max_retries: | |
print("Max retries reached. Moving on to the next task.") | |
return | |
except Exception as e: | |
print(f"An error occurred while querying the API: {e}") | |
return | |
# print(f" rating {products[0]['data']['RATING']}") | |
with PostgreSQLConnector() as db_connector: | |
try: | |
newpricetime = products[0]['data']['NEW_time'] | |
price = get_price(products[0]['csv'][18]) | |
stock = float(products[0]['numberOfItems']) | |
stock = max(0, round(stock)) | |
final_rating = None | |
final_review = None | |
if 'RATING' in products[0]['data']: | |
rating = products[0]['data']['RATING'] | |
final_rating = round(float(rating[-1]), 1) | |
if 'COUNT_REVIEWS' in products[0]['data']: | |
reviews = products[0]['data']['COUNT_REVIEWS'] | |
final_review = float(reviews[-1]) | |
print(f" price {price} asin {asin} stock {products[0]['numberOfItems']} stock rounded {stock}") | |
in_stock_status = 0 | |
if products[0]['availabilityAmazon'] in [0, 4]: | |
in_stock_status = 1 | |
amazon_object = db_connector.session.query(Amazon).filter(Amazon.asin == asin).one() | |
if amazon_object: | |
amazon_object.price = price | |
amazon_object.stock = stock | |
amazon_object.in_stock_status = in_stock_status | |
amazon_object.customer_reviews = final_review | |
amazon_object.total_no_of_ratings = final_rating | |
amazon_object.last_update = newpricetime[-1].strftime("%Y/%m/%d %H:%M:%S") | |
db_connector.session.commit() | |
else: | |
print("Amazon object with the given ASIN not found") | |
except NoResultFound: | |
print("Amazon object with the given ASIN not found") | |
db_connector.session.rollback() | |
except Exception as e: | |
print(f"An error occurred: {e}") | |
db_connector.session.rollback() | |
finally: | |
db_connector.session.close() | |
def run_main(): | |
token = get_token_status() | |
usable_tokens = float(token['tokensLeft']) - float(keep_tokens) | |
if float(usable_tokens) <= float(keep_tokens): | |
# raise Exception("Not enough tokens left") | |
print(f" keep tokens {float(keep_tokens)} usable tokens {float(usable_tokens)} Not enough tokens left") | |
time.sleep(60) | |
limit = math.floor(usable_tokens) | |
print(f" limit {limit} usable tokens {usable_tokens}") | |
with PostgreSQLConnector() as db_connector: | |
try: | |
asins = db_connector.session.query(Amazon.asin) \ | |
.join(Product, Product.asin == Amazon.asin) \ | |
.filter(Amazon.asin is not None, Amazon.asin != '', Amazon.is_linked == 1) \ | |
.filter(Product.status == 1, Product.master != 2, Product.is_category_hidden != 1) \ | |
.filter(or_(Product.primary_category_path.like('%gaming%'), Product.primary_category_path.like('%computing%'))).order_by( # noqa | |
case( | |
(Amazon.modified_at == None, 1), # noqa: E711 | |
(text( | |
"NOT (CAST(amazon.modified_at AS TEXT) ~ '^\\d{4}-\\d{2}-\\d{2} \\d{2}:\\d{2}:\\d{2}\\.\\d{6}\\+\\d{2}$')"), 1), # noqa | |
else_=2 | |
), | |
Amazon.modified_at.asc() | |
).limit(200) | |
# Convert result to a list of ASIN strings | |
asin_list = [asin[0] for asin in asins] | |
print(f"Total ASINs to process: {len(asin_list)})") | |
# Processing ASINs in batches | |
batch_size = 60 | |
num_batches = len(asin_list) // batch_size + (1 if len(asin_list) % batch_size > 0 else 0) | |
for i in range(num_batches): | |
start_index = i * batch_size | |
end_index = start_index + batch_size | |
batch = asin_list[start_index:end_index] | |
for asin in batch: | |
print(f"Processing ASIN: {asin}") | |
_request_product(asin) | |
# Sleep for 2 seconds after processing each batch | |
time.sleep(1) | |
# Check if it's not the last batch before sleeping for a minute | |
if i < num_batches - 1: | |
print("Waiting for 10 seconds before processing the next batch...") | |
time.sleep(10) | |
except Exception: | |
db_connector.session.rollback() | |
finally: | |
db_connector.session.close() |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment