Ofure/Web_scraping_project
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Refactor RedisManager methods for improved error handling and logging; streamline job validation process by ensuring all compulsory fields are checked before processing.

Browse Source
This commit is contained in:
Ofure Ikheloa 2025-12-15 10:34:41 +01:00
parent 87c67265f8
commit e2e1bc442e
1 changed files with 49 additions and 88 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

65
scraper.py
View File

@ -17,26 +17,20 @@ from scraping_engine import FingerprintScrapingEngine
from dotenv import load_dotenv from dotenv import load_dotenv
from ssl_connection import create_ssl_connection_parameters # Import from ssl.py from ssl_connection import create_ssl_connection_parameters # Import from ssl.py
import redis import redis
load_dotenv() load_dotenv()
# Configure logging # Configure logging
logging.basicConfig(level=logging.INFO, logging.basicConfig(level=logging.INFO,
format='%(asctime)s - %(levelname)s - %(message)s') format='%(asctime)s - %(levelname)s - %(message)s')
logger = logging.getLogger(__name__) logger = logging.getLogger(__name__)
# Environment variables # Environment variables
RABBITMQ_HOST = os.getenv("RABBITMQ_HOST") RABBITMQ_HOST = os.getenv("RABBITMQ_HOST")
RABBITMQ_PORT = int(os.getenv("RABBITMQ_PORT", "5671")) RABBITMQ_PORT = int(os.getenv("RABBITMQ_PORT", "5671"))
RABBITMQ_SSL_ENABLED = os.getenv("RABBITMQ_SSL_ENABLED", "false").lower() == "true" RABBITMQ_SSL_ENABLED = os.getenv("RABBITMQ_SSL_ENABLED", "false").lower() == "true"
# Redis configuration # Redis configuration
REDIS_HOST = os.getenv('REDIS_HOST') REDIS_HOST = os.getenv('REDIS_HOST')
REDIS_PORT = int(os.getenv('REDIS_PORT', '6380')) REDIS_PORT = int(os.getenv('REDIS_PORT', '6380'))
REDIS_PASSWORD = os.getenv('REDIS_PASSWORD') REDIS_PASSWORD = os.getenv('REDIS_PASSWORD')
REDIS_SSL_ENABLED = os.getenv('REDIS_SSL_ENABLED', 'true').lower() == 'true' REDIS_SSL_ENABLED = os.getenv('REDIS_SSL_ENABLED', 'true').lower() == 'true'
class RedisManager: class RedisManager:
"""Manages Redis connection and operations for job tracking and caching.""" """Manages Redis connection and operations for job tracking and caching."""
@ -145,14 +139,12 @@ class MultiPlatformJobScraper:
await self.close_browser() await self.close_browser()
except: except:
await self.close_browser() await self.close_browser()
if self.browser is None: if self.browser is None:
try: try:
profile = self.engine._select_profile() profile = self.engine._select_profile()
renderer = random.choice(self.engine.common_renderers[self.engine.os]) renderer = random.choice(self.engine.common_renderers[self.engine.os])
vendor = random.choice(self.engine.common_vendors) vendor = random.choice(self.engine.common_vendors)
spoof_script = self.engine._get_spoof_script(renderer, vendor) spoof_script = self.engine._get_spoof_script(renderer, vendor)
self.pw = await async_playwright().start() self.pw = await async_playwright().start()
self.browser = await self.pw.chromium.launch( self.browser = await self.pw.chromium.launch(
headless=True, headless=True,
@ -178,7 +170,6 @@ class MultiPlatformJobScraper:
logger.warning("Browser appears dead. Reinitializing...") logger.warning("Browser appears dead. Reinitializing...")
await self.close_browser() await self.close_browser()
await self.init_browser() await self.init_browser()
profile = self.engine._select_profile() profile = self.engine._select_profile()
context = await AsyncNewContext(self.browser, fingerprint=profile) context = await AsyncNewContext(self.browser, fingerprint=profile)
await context.add_init_script(f""" await context.add_init_script(f"""
@ -274,35 +265,29 @@ class MultiPlatformJobScraper:
if platform == "unknown": if platform == "unknown":
logger.info(f"⏭️ Skipping unsupported platform: {job_url}") logger.info(f"⏭️ Skipping unsupported platform: {job_url}")
return True return True
job_id = job_url.strip("/").split("/")[-1] job_id = job_url.strip("/").split("/")[-1]
if await self._is_job_seen(job_id): if await self._is_job_seen(job_id):
logger.info(f"⏭️ Skipping already processed job: {job_id}") logger.info(f"⏭️ Skipping already processed job: {job_id}")
return True return True
cached_result = await self._get_cached_llm_result(job_url) cached_result = await self._get_cached_llm_result(job_url)
if cached_result: if cached_result:
logger.info(f"📦 Using cached LLM result for: {job_url}") logger.info(f"📦 Using cached LLM result for: {job_url}")
await self.llm_agent.save_job_data(cached_result, company_name) await self.llm_agent.save_job_data(cached_result, company_name)
await self._mark_job_seen(job_id) await self._mark_job_seen(job_id)
return True return True
context = None context = None
page = None page = None
start_time = time.time() start_time = time.time()
try: try:
context = await self.create_fresh_context() context = await self.create_fresh_context()
page = await context.new_page() page = await context.new_page()
timeout_ms = self.engine.optimization_params.get("request_timeout", 120000) timeout_ms = self.engine.optimization_params.get("request_timeout", 120000)
temp_fetcher = StealthyFetcher(self.engine, self.browser, context) temp_fetcher = StealthyFetcher(self.engine, self.browser, context)
fetch_timeout = 60000 if platform == "lever" else timeout_ms fetch_timeout = 60000 if platform == "lever" else timeout_ms
job_page = await asyncio.wait_for( job_page = await asyncio.wait_for(
temp_fetcher.fetch_url(job_url, wait_for_selector="h1", timeout=fetch_timeout), temp_fetcher.fetch_url(job_url, wait_for_selector="h1", timeout=fetch_timeout),
timeout=fetch_timeout / 1000.0 timeout=fetch_timeout / 1000.0
) )
# Check if job still exists (minimal content validation) # Check if job still exists (minimal content validation)
page_content = await job_page.content() page_content = await job_page.content()
if len(page_content.strip()) < 500: # Arbitrary threshold for "page exists" if len(page_content.strip()) < 500: # Arbitrary threshold for "page exists"
@ -310,7 +295,6 @@ class MultiPlatformJobScraper:
await self._add_job_to_redis_cache(job_url, job_id, "job_not_found") await self._add_job_to_redis_cache(job_url, job_id, "job_not_found")
self.engine.report_outcome("job_not_found", url=job_url) self.engine.report_outcome("job_not_found", url=job_url)
return False return False
if platform == "ashby": if platform == "ashby":
try: try:
await job_page.wait_for_selector("div[class*='job-posting'], article, main", timeout=60000) await job_page.wait_for_selector("div[class*='job-posting'], article, main", timeout=60000)
@ -324,7 +308,6 @@ class MultiPlatformJobScraper:
await job_page.wait_for_selector("div.job-desc, section", timeout=60000) await job_page.wait_for_selector("div.job-desc, section", timeout=60000)
except Exception: except Exception:
pass pass
# Extract page content for initial validation # Extract page content for initial validation
page_content = await self._extract_page_content_for_llm(job_page) page_content = await self._extract_page_content_for_llm(job_page)
@ -346,7 +329,6 @@ class MultiPlatformJobScraper:
await self._add_job_to_redis_cache(job_url, job_id, "job_not_found") await self._add_job_to_redis_cache(job_url, job_id, "job_not_found")
self.engine.report_outcome("job_not_found", url=job_url) self.engine.report_outcome("job_not_found", url=job_url)
return False return False
# 🔑 APPLY TYPE LOGIC # 🔑 APPLY TYPE LOGIC
if platform in ["ashby", "lever", "greenhouse"]: if platform in ["ashby", "lever", "greenhouse"]:
apply_type = 'AI' # Always AI for these platforms apply_type = 'AI' # Always AI for these platforms
@ -368,11 +350,9 @@ class MultiPlatformJobScraper:
apply_type = 'signup' apply_type = 'signup'
else: else:
apply_type = 'signup' apply_type = 'signup'
final_url = job_url final_url = job_url
# Hardcode posted_date to Dec 1st 2025 # Hardcode posted_date to Dec 1st 2025
posted_date = "12/01/25" posted_date = "12/01/25"
raw_data = { raw_data = {
"page_content": page_content, "page_content": page_content,
"url": final_url, "url": final_url,
@ -380,34 +360,32 @@ class MultiPlatformJobScraper:
"search_keywords": company_name, "search_keywords": company_name,
"posted_date": posted_date "posted_date": posted_date
} }
llm_timeout = max(60, self.engine.feedback.get("avg_response_time", 10) * 2) llm_timeout = max(60, self.engine.feedback.get("avg_response_time", 10) * 2)
refined_data = await asyncio.wait_for( refined_data = await asyncio.wait_for(
self.llm_agent.refine_job_data(raw_data, self.user_request), self.llm_agent.refine_job_data(raw_data, self.user_request),
timeout=llm_timeout timeout=llm_timeout
) )
success = False success = False
if refined_data and refined_data.get("title", "N/A") != "N/A": if refined_data:
# Define all compulsory fields that must be present and valid # Define compulsory fields that must be present and valid
compulsory_fields = ['company_name', 'job_id', 'url', 'title', 'description'] compulsory_fields = ['title', 'company_name', 'description', 'job_id', 'url']
# Validate all compulsory fields # Check if ALL compulsory fields are present and valid BEFORE any processing
missing_fields = [] missing_fields = []
for field in compulsory_fields: for field in compulsory_fields:
field_value = refined_data.get(field, "").strip() field_value = refined_data.get(field, "")
if not field_value or field_value in ["N/A", "Unknown", ""]: if not field_value or str(field_value).strip() in ["", "N/A", "Unknown", "Not provided", "Not available", "Company", "Job"]:
missing_fields.append(field) missing_fields.append(field)
# If any compulsory field or description is missing, discard the job # If any compulsory field is missing, discard the job immediately
if missing_fields: if missing_fields:
logger.error(f"❌ Job discarded - missing compulsory fields {missing_fields}: {final_url}") logger.error(f"❌ Job discarded - missing compulsory fields: {', '.join(missing_fields)} : {final_url}")
await self._add_job_to_redis_cache(final_url, job_id, "job_not_found") error_type = "missing_compulsory_fields"
self.engine.report_outcome("job_not_found", url=final_url) await self._add_job_to_redis_cache(final_url, job_id, error_type)
self.engine.report_outcome(error_type, url=final_url)
return False return False
# If we get here, all compulsory fields are present and valid # If we get here, all compulsory fields are valid - now add additional metadata
# Update with additional metadata
refined_data.update({ refined_data.update({
'apply_type': apply_type, 'apply_type': apply_type,
'scraped_at': datetime.now().isoformat(), 'scraped_at': datetime.now().isoformat(),
@ -417,6 +395,7 @@ class MultiPlatformJobScraper:
'platform': platform 'platform': platform
}) })
# Save to database and markdown
await self.llm_agent.save_job_data(refined_data, company_name) await self.llm_agent.save_job_data(refined_data, company_name)
await self._cache_llm_result(job_url, refined_data) await self._cache_llm_result(job_url, refined_data)
await self._mark_job_seen(job_id) await self._mark_job_seen(job_id)
@ -429,9 +408,7 @@ class MultiPlatformJobScraper:
logger.warning(f"🟡 LLM failed to refine: {final_url}") logger.warning(f"🟡 LLM failed to refine: {final_url}")
await self._add_job_to_redis_cache(final_url, job_id, "llm_failure") await self._add_job_to_redis_cache(final_url, job_id, "llm_failure")
self.engine.report_outcome("llm_failure", url=final_url) self.engine.report_outcome("llm_failure", url=final_url)
return success return success
except asyncio.TimeoutError: except asyncio.TimeoutError:
logger.error(f"⏰ Timeout processing job ({platform}): {job_url}") logger.error(f"⏰ Timeout processing job ({platform}): {job_url}")
await self._add_job_to_redis_cache(job_url, job_id, "timeout") await self._add_job_to_redis_cache(job_url, job_id, "timeout")
@ -468,7 +445,6 @@ class MultiPlatformJobScraper:
await context.close() await context.close()
except Exception: except Exception:
pass pass
# Global metrics # Global metrics
METRICS = { METRICS = {
"processed": 0, "processed": 0,
@ -477,18 +453,14 @@ METRICS = {
"skipped": 0, "skipped": 0,
"start_time": time.time() "start_time": time.time()
} }
async def process_message_async(scraper: MultiPlatformJobScraper, ch, method, properties, body): async def process_message_async(scraper: MultiPlatformJobScraper, ch, method, properties, body):
try: try:
job_data = json.loads(body) job_data = json.loads(body)
job_link = job_data['job_link'] job_link = job_data['job_link']
company_name = job_data['company_name'] company_name = job_data['company_name']
message_id = properties.message_id or f"msg_{int(time.time()*1000)}" message_id = properties.message_id or f"msg_{int(time.time()*1000)}"
logger.info(f"📥 Processing job: {job_link} (ID: {message_id})") logger.info(f"📥 Processing job: {job_link} (ID: {message_id})")
success = await scraper.scrape_job(job_link, company_name, message_id) success = await scraper.scrape_job(job_link, company_name, message_id)
METRICS["processed"] += 1 METRICS["processed"] += 1
if success: if success:
METRICS["success"] += 1 METRICS["success"] += 1
@ -502,14 +474,10 @@ async def process_message_async(scraper: MultiPlatformJobScraper, ch, method, pr
METRICS["failed"] += 1 METRICS["failed"] += 1
finally: finally:
ch.basic_ack(delivery_tag=method.delivery_tag) ch.basic_ack(delivery_tag=method.delivery_tag)
def callback_wrapper(scraper: MultiPlatformJobScraper): def callback_wrapper(scraper: MultiPlatformJobScraper):
def callback(ch, method, properties, body): def callback(ch, method, properties, body):
asyncio.run(process_message_async(scraper, ch, method, properties, body)) asyncio.run(process_message_async(scraper, ch, method, properties, body))
return callback return callback
def start_consumer(): def start_consumer():
engine = FingerprintScrapingEngine( engine = FingerprintScrapingEngine(
seed="multiplatform_scraper", seed="multiplatform_scraper",
@ -517,7 +485,6 @@ def start_consumer():
num_variations=10 num_variations=10
) )
scraper = MultiPlatformJobScraper(engine) scraper = MultiPlatformJobScraper(engine)
connection = None connection = None
for attempt in range(5): for attempt in range(5):
try: try:
@ -527,22 +494,18 @@ def start_consumer():
logger.info(f"Connecting to RabbitMQ over SSL at {RABBITMQ_HOST}:{RABBITMQ_PORT}") logger.info(f"Connecting to RabbitMQ over SSL at {RABBITMQ_HOST}:{RABBITMQ_PORT}")
else: else:
logger.info(f"Connecting to RabbitMQ at {RABBITMQ_HOST}:{RABBITMQ_PORT}") logger.info(f"Connecting to RabbitMQ at {RABBITMQ_HOST}:{RABBITMQ_PORT}")
connection = pika.BlockingConnection(parameters) connection = pika.BlockingConnection(parameters)
break break
except Exception as e: except Exception as e:
logger.error(f"RabbitMQ connection attempt {attempt + 1} failed: {e}") logger.error(f"RabbitMQ connection attempt {attempt + 1} failed: {e}")
time.sleep(2 ** attempt) time.sleep(2 ** attempt)
if not connection: if not connection:
logger.error("Failed to connect to RabbitMQ after retries") logger.error("Failed to connect to RabbitMQ after retries")
return return
channel = connection.channel() channel = connection.channel()
channel.queue_declare(queue='job_queue', durable=True) channel.queue_declare(queue='job_queue', durable=True)
channel.basic_qos(prefetch_count=1) channel.basic_qos(prefetch_count=1)
channel.basic_consume(queue='job_queue', on_message_callback=callback_wrapper(scraper)) channel.basic_consume(queue='job_queue', on_message_callback=callback_wrapper(scraper))
logger.info('Waiting for messages (Ashby, Lever, Greenhouse). To exit press CTRL+C') logger.info('Waiting for messages (Ashby, Lever, Greenhouse). To exit press CTRL+C')
try: try:
channel.start_consuming() channel.start_consuming()
@ -551,7 +514,5 @@ def start_consumer():
channel.stop_consuming() channel.stop_consuming()
connection.close() connection.close()
asyncio.run(scraper.close_browser()) asyncio.run(scraper.close_browser())
if __name__ == "__main__": if __name__ == "__main__":
start_consumer() start_consumer()