Spaces:
Sleeping
Sleeping
File size: 12,909 Bytes
ec2ef5e | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 | # Technical Implementation Details
## Key Code Changes
### 1. Listing Scraper - Thread-Based Page Extraction Timeout
**Before:**
```python
def _get_products_on_page(self) -> List[Dict[str, Any]]:
page = self.page
products: List[Dict[str, Any]] = []
rows = page.locator("div.row").all() # Could hang here
for idx, row in enumerate(rows):
# ... extraction code ...
return products
```
**After:**
```python
def _get_products_on_page(self) -> List[Dict[str, Any]]:
page = self.page
products: List[Dict[str, Any]] = []
extraction_timeout = self.page_timeout - 5 # Leave 5s buffer
result_container = {"products": [], "error": None, "done": False}
def extract_products():
try:
rows = page.locator("div.row").all() # Runs in separate thread
logger.debug(f"Found {len(rows)} potential product rows")
for idx, row in enumerate(rows):
# ... extraction code ...
except Exception as e:
result_container["error"] = str(e)
finally:
result_container["done"] = True
# Run extraction in thread with timeout
thread = threading.Thread(target=extract_products, daemon=True)
thread.start()
thread.join(timeout=extraction_timeout) # Max 25 seconds
if result_container["done"]:
if result_container["error"]:
logger.error(f"Product extraction error: {result_container['error']}")
products = result_container["products"]
else:
logger.warning(f"Product extraction timed out after {extraction_timeout}s")
try:
self.session.screenshot("extraction_timeout")
except Exception:
pass
logger.info(f"Extracted {len(products)} qualifying product(s) from page")
return products
```
**Why:** Locator operations can hang indefinitely if the page is still loading or has JavaScript errors. Threading with timeout ensures we detect and recover from hangs.
---
### 2. Pagination Loop - Comprehensive Timeout System
**Before:**
```python
def iter_qualifying_products(self, studio_url: str):
current_url = studio_url
page_num = 1
visited_urls = set()
max_pages = 10000 # Unlimited effectively
while current_url and page_num <= max_pages:
if current_url in visited_urls:
logger.warning("Pagination loop detected. Stopping scan.")
break
visited_urls.add(current_url)
products = self._get_products_on_page()
# ... rest of loop ...
```
**After:**
```python
def iter_qualifying_products(self, studio_url: str):
current_url = studio_url
page_num = 1
visited_urls = set()
consecutive_failures = 0
max_consecutive_failures = 3
pages_without_products = 0
max_empty_pages = 5
self.scan_start_time = time.time()
self.last_product_found_time = self.scan_start_time
logger.info(f"Starting listing scan with min_price=${self.min_price:.2f}")
logger.info(f"Limits: max_pages={self.max_pages}, page_timeout={self.page_timeout}s, total_timeout={self.total_timeout}s")
while current_url and page_num <= self.max_pages:
# CHECK 1: User stop signal
if self.stop_event and self.stop_event.is_set():
logger.info("Listing scan stopped by user")
break
# CHECK 2: Total time exceeded
elapsed_time = time.time() - self.scan_start_time
if elapsed_time > self.total_timeout:
logger.warning(f"Total scan timeout exceeded ({elapsed_time:.0f}s > {self.total_timeout}s). Stopping scan.")
break
# CHECK 3: Pagination loop
if current_url in visited_urls:
logger.warning("Pagination loop detected. Stopping scan.")
break
# CHECK 4: Too many empty pages
if pages_without_products >= max_empty_pages:
logger.warning(f"Too many empty pages ({pages_without_products}/{max_empty_pages}). Stopping scan.")
break
# CHECK 5: No products found in extended time
if self.collected == 0 and elapsed_time > 300:
logger.warning("No products found after 5 minutes. Stopping scan.")
break
visited_urls.add(current_url)
logger.info(f"Scanning listing page {page_num}: {current_url}")
# ... process page ...
# CHECK 6: Page load failures
if not self._load_page_with_retry(current_url, page_num, max_retries=3):
consecutive_failures += 1
if consecutive_failures >= max_consecutive_failures:
logger.error(f"Too many consecutive page failures ({consecutive_failures}). Stopping scan.")
break
products = self._get_products_on_page()
if not products:
pages_without_products += 1
else:
pages_without_products = 0
self.last_product_found_time = time.time()
# ... yield products ...
current_url = next_url
page_num += 1
```
**Why:** Multiple checks create a safety net. If one check fails, others catch the problem. Total timeout prevents infinite loops on any individual check failure.
---
### 3. Page Loading - Thread-Protected Navigation
**Before:**
```python
def _load_page_with_retry(self, url: str, page_num: int, max_retries: int = 3) -> bool:
for attempt in range(1, max_retries + 1):
try:
logger.info(f"Page load attempt {attempt}/{max_retries} for page {page_num}: {url}")
if self.session.goto(url): # Could hang here
logger.info(f"✓ Successfully loaded page {page_num} on attempt {attempt}")
try:
self.session.cleanup_page()
except Exception as e:
logger.debug(f"Cleanup warning: {e}")
return True
else:
logger.warning(f"✗ Page load failed for page {page_num}, attempt {attempt}")
except Exception as e:
logger.warning(f"✗ Exception during page load: {e}")
if attempt < max_retries:
delay = retry_delays[attempt - 1]
logger.info(f"Waiting {delay}s before retry...")
time.sleep(delay)
return False
```
**After:**
```python
def _load_page_with_retry(self, url: str, page_num: int, max_retries: int = 3) -> bool:
retry_delays = [2, 5, 10]
for attempt in range(1, max_retries + 1):
try:
logger.info(f"Page load attempt {attempt}/{max_retries} for page {page_num}: {url}")
# Load page with timeout thread
load_result = {"success": False, "done": False}
def load_thread():
try:
load_result["success"] = self.session.goto(url)
except Exception as e:
logger.warning(f"Navigation exception: {e}")
finally:
load_result["done"] = True
thread = threading.Thread(target=load_thread, daemon=True)
thread.start()
thread.join(timeout=self.page_timeout) # Max 30 seconds
if load_result["done"] and load_result["success"]:
logger.info(f"✓ Successfully loaded page {page_num} on attempt {attempt}")
try:
self.session.cleanup_page()
import gc
gc.collect() # Force garbage collection
except Exception as e:
logger.debug(f"Cleanup warning: {e}")
return True
else:
if not load_result["done"]:
logger.warning(f"✗ Page load timed out after {self.page_timeout}s for page {page_num}, attempt {attempt}")
else:
logger.warning(f"✗ Page load failed for page {page_num}, attempt {attempt}")
except Exception as e:
logger.warning(f"✗ Exception during page load: {e}")
if attempt < max_retries:
delay = retry_delays[attempt - 1]
logger.info(f"Waiting {delay}s before retry...")
time.sleep(delay)
logger.error(f"Failed to load page {page_num} after {max_retries} retries: {url}")
return False
```
**Why:** Threading ensures Playwright navigation doesn't hang. If network is slow or server doesn't respond, we detect it and move on after timeout instead of waiting forever.
---
### 4. Product Scraper - Timeout-Protected Field Extraction
**Before:**
```python
def scrape_product(self, url: str) -> Dict[str, Any]:
# ...
result["title"] = self._get_title() # Could hang
result["price"] = self._get_price() # Could hang
result["upc"] = (self._meta_value("UPC") or "").strip() # Could hang
result["category"] = self._get_category() # Could hang
# ...
```
**After:**
```python
def scrape_product(self, url: str) -> Dict[str, Any]:
# ...
result["title"] = self._get_title_safe() # 8s timeout
result["price"] = self._get_price_safe() # 8s timeout
result["upc"] = (self._meta_value_safe("UPC") or "").strip() # 8s timeout
result["category"] = self._get_category_safe() # 8s timeout
# ...
def _get_title_safe(self, timeout: int = 8) -> str:
"""Extract title with timeout protection."""
result = {"value": "", "done": False}
def extract():
try:
result["value"] = self._get_title()
except Exception as e:
logger.debug(f"Title extraction error: {e}")
finally:
result["done"] = True
thread = threading.Thread(target=extract, daemon=True)
thread.start()
thread.join(timeout=timeout)
if not result["done"]:
logger.warning("Title extraction timed out")
return result["value"]
```
**Why:** Individual fields might hang on slow servers. Timeout per field prevents one slow field from blocking entire scan.
---
### 5. Browser Session - Enhanced Cleanup
**Before:**
```python
def cleanup_page(self):
try:
if self.page:
self.page.evaluate("() => { localStorage.clear(); sessionStorage.clear(); }")
logger.debug("Cleared page storage")
except Exception as e:
logger.debug(f"Page cleanup warning: {e}")
```
**After:**
```python
def cleanup_page(self):
"""Cleanup page resources to prevent memory accumulation during long pagination runs."""
try:
if self.page:
self.page.evaluate("""
() => {
localStorage.clear();
sessionStorage.clear();
// Purge cache
if (window.caches) {
caches.keys().then(names => {
names.forEach(name => caches.delete(name));
});
}
}
""")
logger.debug("Cleared page storage and cache")
except Exception as e:
logger.debug(f"Page cleanup warning: {e}")
```
**Why:** On memory-constrained live servers, every bit helps. Clearing caches and storage prevents memory accumulation across pages.
---
## Configuration Examples
### Hugging Face Space (Limited Resources)
```python
ListingScraper(
session=session,
min_price=self.min_price,
max_pages=50, # Conservative
page_timeout=15, # Shorter timeouts
total_timeout=900, # 15 minutes max
)
```
### Local Development (Full Resources)
```python
ListingScraper(
session=session,
min_price=self.min_price,
max_pages=100, # Default
page_timeout=30, # Default
total_timeout=1800, # 30 minutes
)
```
### High-Traffic Sites (Slow Servers)
```python
ListingScraper(
session=session,
min_price=self.min_price,
max_pages=100,
page_timeout=45, # Longer for slow servers
total_timeout=2400, # 40 minutes
)
```
---
## Troubleshooting
### If scan still hangs:
1. Lower `page_timeout` to detect hangs faster
2. Lower `max_pages` limit
3. Lower `total_timeout` to terminate early
4. Check network connectivity
5. Look at screenshot files in `logs/` directory for visual debugging
### If scan terminates too early:
1. Increase `page_timeout`
2. Increase `total_timeout`
3. Check logs for "Page load timed out" messages
4. Consider increasing `max_pages` if legitimate pages are being skipped
### Memory issues on Hugging Face:
1. Use conservative settings (see above)
2. Ensure `cleanup_page()` is called (happens automatically)
3. Monitor with smaller `max_pages` values
4. Consider running multiple smaller scans instead of one large scan
|