Dmitry Beresnev
commited on
Commit
·
5fbc88e
1
Parent(s):
2da1d42
fix news UI
Browse files- app/components/news.py +29 -115
- app/services/twitter_news_playwright.py +14 -14
app/components/news.py
CHANGED
|
@@ -373,120 +373,34 @@ def display_breaking_news_banner(df: pd.DataFrame):
|
|
| 373 |
hours = time_diff.seconds // 3600
|
| 374 |
time_ago = f"{hours}h ago" if hours < 24 else f"{time_diff.days}d ago"
|
| 375 |
|
| 376 |
-
# TradingView-style breaking news banner
|
| 377 |
-
banner_html = f"""
|
| 378 |
-
|
| 379 |
-
|
| 380 |
-
|
| 381 |
-
|
| 382 |
-
|
| 383 |
-
|
| 384 |
-
|
| 385 |
-
|
| 386 |
-
|
| 387 |
-
|
| 388 |
-
|
| 389 |
-
|
| 390 |
-
|
| 391 |
-
|
| 392 |
-
|
| 393 |
-
|
| 394 |
-
|
| 395 |
-
|
| 396 |
-
|
| 397 |
-
|
| 398 |
-
|
| 399 |
-
|
| 400 |
-
|
| 401 |
-
|
| 402 |
-
|
| 403 |
-
|
| 404 |
-
|
| 405 |
-
right: 0;
|
| 406 |
-
bottom: 0;
|
| 407 |
-
background: repeating-linear-gradient(
|
| 408 |
-
45deg,
|
| 409 |
-
transparent,
|
| 410 |
-
transparent 10px,
|
| 411 |
-
rgba(255, 255, 255, 0.03) 10px,
|
| 412 |
-
rgba(255, 255, 255, 0.03) 20px
|
| 413 |
-
);
|
| 414 |
-
pointer-events: none;
|
| 415 |
-
"></div>
|
| 416 |
-
|
| 417 |
-
<!-- Content -->
|
| 418 |
-
<div style="position: relative; z-index: 1;">
|
| 419 |
-
<div style="display: flex; align-items: center; gap: 16px; margin-bottom: 12px;">
|
| 420 |
-
<!-- Animated icon -->
|
| 421 |
-
<div style="
|
| 422 |
-
font-size: 32px;
|
| 423 |
-
animation: pulse-glow 1s ease-in-out infinite;
|
| 424 |
-
filter: drop-shadow(0 2px 8px rgba(0, 0, 0, 0.3));
|
| 425 |
-
">🚨</div>
|
| 426 |
-
|
| 427 |
-
<!-- Header -->
|
| 428 |
-
<div style="flex: 1;">
|
| 429 |
-
<div style="
|
| 430 |
-
color: white;
|
| 431 |
-
font-size: 14px;
|
| 432 |
-
font-weight: 700;
|
| 433 |
-
letter-spacing: 1.5px;
|
| 434 |
-
text-transform: uppercase;
|
| 435 |
-
margin-bottom: 4px;
|
| 436 |
-
font-family: -apple-system, BlinkMacSystemFont, 'Trebuchet MS', Roboto, Ubuntu, sans-serif;
|
| 437 |
-
text-shadow: 0 2px 4px rgba(0, 0, 0, 0.2);
|
| 438 |
-
">⚡ Breaking News</div>
|
| 439 |
-
<div style="
|
| 440 |
-
color: rgba(255, 255, 255, 0.9);
|
| 441 |
-
font-size: 11px;
|
| 442 |
-
display: flex;
|
| 443 |
-
align-items: center;
|
| 444 |
-
gap: 8px;
|
| 445 |
-
">
|
| 446 |
-
<span style="
|
| 447 |
-
background: rgba(255, 255, 255, 0.2);
|
| 448 |
-
padding: 2px 8px;
|
| 449 |
-
border-radius: 4px;
|
| 450 |
-
font-weight: 600;
|
| 451 |
-
">{source}</span>
|
| 452 |
-
<span style="opacity: 0.8;">•</span>
|
| 453 |
-
<span style="opacity: 0.8;">{time_ago}</span>
|
| 454 |
-
</div>
|
| 455 |
-
</div>
|
| 456 |
-
|
| 457 |
-
<!-- Read button -->
|
| 458 |
-
<a href="{url}" target="_blank" style="
|
| 459 |
-
background: white;
|
| 460 |
-
color: #F23645;
|
| 461 |
-
padding: 10px 20px;
|
| 462 |
-
border-radius: 6px;
|
| 463 |
-
font-size: 13px;
|
| 464 |
-
font-weight: 700;
|
| 465 |
-
text-decoration: none;
|
| 466 |
-
display: inline-flex;
|
| 467 |
-
align-items: center;
|
| 468 |
-
gap: 6px;
|
| 469 |
-
transition: all 0.2s ease;
|
| 470 |
-
box-shadow: 0 2px 8px rgba(0, 0, 0, 0.2);
|
| 471 |
-
" onmouseover="this.style.transform='translateY(-2px)'; this.style.boxShadow='0 4px 12px rgba(0, 0, 0, 0.3)';"
|
| 472 |
-
onmouseout="this.style.transform='translateY(0)'; this.style.boxShadow='0 2px 8px rgba(0, 0, 0, 0.2)';">
|
| 473 |
-
READ NOW →
|
| 474 |
-
</a>
|
| 475 |
-
</div>
|
| 476 |
-
|
| 477 |
-
<!-- News summary -->
|
| 478 |
-
<div style="
|
| 479 |
-
color: white;
|
| 480 |
-
font-size: 16px;
|
| 481 |
-
font-weight: 500;
|
| 482 |
-
line-height: 1.5;
|
| 483 |
-
margin-left: 48px;
|
| 484 |
-
font-family: -apple-system, BlinkMacSystemFont, 'Trebuchet MS', Roboto, Ubuntu, sans-serif;
|
| 485 |
-
text-shadow: 0 1px 2px rgba(0, 0, 0, 0.2);
|
| 486 |
-
animation: slide-in 0.5s ease-out;
|
| 487 |
-
">{summary}</div>
|
| 488 |
-
</div>
|
| 489 |
-
</div>
|
| 490 |
-
"""
|
| 491 |
|
| 492 |
st.markdown(banner_html, unsafe_allow_html=True)
|
|
|
|
| 373 |
hours = time_diff.seconds // 3600
|
| 374 |
time_ago = f"{hours}h ago" if hours < 24 else f"{time_diff.days}d ago"
|
| 375 |
|
| 376 |
+
# TradingView-style breaking news banner (no leading whitespace)
|
| 377 |
+
banner_html = f"""<style>
|
| 378 |
+
@keyframes pulse-glow {{
|
| 379 |
+
0%, 100% {{ box-shadow: 0 0 20px rgba(242, 54, 69, 0.6); }}
|
| 380 |
+
50% {{ box-shadow: 0 0 30px rgba(242, 54, 69, 0.9); }}
|
| 381 |
+
}}
|
| 382 |
+
@keyframes slide-in {{
|
| 383 |
+
from {{ transform: translateX(-10px); opacity: 0; }}
|
| 384 |
+
to {{ transform: translateX(0); opacity: 1; }}
|
| 385 |
+
}}
|
| 386 |
+
</style>
|
| 387 |
+
<div style="background: linear-gradient(135deg, #F23645 0%, #C91B28 100%); border: 2px solid #FF6B78; border-radius: 12px; padding: 20px 24px; margin-bottom: 24px; animation: pulse-glow 2s ease-in-out infinite; position: relative; overflow: hidden;">
|
| 388 |
+
<div style="position: absolute; top: 0; left: 0; right: 0; bottom: 0; background: repeating-linear-gradient(45deg, transparent, transparent 10px, rgba(255, 255, 255, 0.03) 10px, rgba(255, 255, 255, 0.03) 20px); pointer-events: none;"></div>
|
| 389 |
+
<div style="position: relative; z-index: 1;">
|
| 390 |
+
<div style="display: flex; align-items: center; gap: 16px; margin-bottom: 12px;">
|
| 391 |
+
<div style="font-size: 32px; animation: pulse-glow 1s ease-in-out infinite; filter: drop-shadow(0 2px 8px rgba(0, 0, 0, 0.3));">🚨</div>
|
| 392 |
+
<div style="flex: 1;">
|
| 393 |
+
<div style="color: white; font-size: 14px; font-weight: 700; letter-spacing: 1.5px; text-transform: uppercase; margin-bottom: 4px; font-family: -apple-system, BlinkMacSystemFont, 'Trebuchet MS', Roboto, Ubuntu, sans-serif; text-shadow: 0 2px 4px rgba(0, 0, 0, 0.2);">⚡ Breaking News</div>
|
| 394 |
+
<div style="color: rgba(255, 255, 255, 0.9); font-size: 11px; display: flex; align-items: center; gap: 8px;">
|
| 395 |
+
<span style="background: rgba(255, 255, 255, 0.2); padding: 2px 8px; border-radius: 4px; font-weight: 600;">{source}</span>
|
| 396 |
+
<span style="opacity: 0.8;">•</span>
|
| 397 |
+
<span style="opacity: 0.8;">{time_ago}</span>
|
| 398 |
+
</div>
|
| 399 |
+
</div>
|
| 400 |
+
<a href="{url}" target="_blank" style="background: white; color: #F23645; padding: 10px 20px; border-radius: 6px; font-size: 13px; font-weight: 700; text-decoration: none; display: inline-flex; align-items: center; gap: 6px; transition: all 0.2s ease; box-shadow: 0 2px 8px rgba(0, 0, 0, 0.2);" onmouseover="this.style.transform='translateY(-2px)'; this.style.boxShadow='0 4px 12px rgba(0, 0, 0, 0.3)';" onmouseout="this.style.transform='translateY(0)'; this.style.boxShadow='0 2px 8px rgba(0, 0, 0, 0.2)';">READ NOW →</a>
|
| 401 |
+
</div>
|
| 402 |
+
<div style="color: white; font-size: 16px; font-weight: 500; line-height: 1.5; margin-left: 48px; font-family: -apple-system, BlinkMacSystemFont, 'Trebuchet MS', Roboto, Ubuntu, sans-serif; text-shadow: 0 1px 2px rgba(0, 0, 0, 0.2); animation: slide-in 0.5s ease-out;">{summary}</div>
|
| 403 |
+
</div>
|
| 404 |
+
</div>"""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 405 |
|
| 406 |
st.markdown(banner_html, unsafe_allow_html=True)
|
app/services/twitter_news_playwright.py
CHANGED
|
@@ -208,8 +208,8 @@ class TwitterFinanceMonitor:
|
|
| 208 |
logger.warning("Chromium not found in standard paths")
|
| 209 |
return '/usr/bin/chromium' # Fallback
|
| 210 |
|
| 211 |
-
def _scrape_twitter_profile(self, source_name: str, source_info: Dict, timeout: int =
|
| 212 |
-
"""Scrape tweets from a single Twitter profile using Playwright
|
| 213 |
if not PLAYWRIGHT_AVAILABLE:
|
| 214 |
logger.warning("Playwright not available")
|
| 215 |
return []
|
|
@@ -234,7 +234,7 @@ class TwitterFinanceMonitor:
|
|
| 234 |
)
|
| 235 |
page = context.new_page()
|
| 236 |
|
| 237 |
-
# Block images, fonts, css, and videos for
|
| 238 |
def route_intercept(route):
|
| 239 |
if route.request.resource_type in ["image", "media", "font", "stylesheet", "video"]:
|
| 240 |
route.abort()
|
|
@@ -243,19 +243,19 @@ class TwitterFinanceMonitor:
|
|
| 243 |
|
| 244 |
page.route("**/*", route_intercept)
|
| 245 |
|
| 246 |
-
# Navigate to profile with
|
| 247 |
logger.info(f"Scraping {source_name}...")
|
| 248 |
-
page.goto(source_info['url'], timeout=timeout * 1000, wait_until="domcontentloaded")
|
| 249 |
|
| 250 |
-
# Wait for tweets to load with
|
| 251 |
try:
|
| 252 |
-
page.wait_for_selector("article", timeout=
|
| 253 |
except PlaywrightTimeoutError:
|
| 254 |
logger.warning(f"Timeout waiting for tweets from {source_name}")
|
| 255 |
browser.close()
|
| 256 |
return []
|
| 257 |
|
| 258 |
-
# Extract tweet texts (limit to 15
|
| 259 |
tweet_elements = page.locator("article div[data-testid='tweetText']").all()
|
| 260 |
|
| 261 |
news_items = []
|
|
@@ -331,19 +331,19 @@ class TwitterFinanceMonitor:
|
|
| 331 |
reverse=True
|
| 332 |
)
|
| 333 |
|
| 334 |
-
# Scrape sources in parallel with
|
| 335 |
-
# 8 workers = 19 sources in 3 batches (~
|
| 336 |
with ThreadPoolExecutor(max_workers=8) as executor:
|
| 337 |
futures = []
|
| 338 |
for name, info in sorted_sources:
|
| 339 |
-
#
|
| 340 |
-
future = executor.submit(_self._scrape_twitter_profile, name, info, timeout=
|
| 341 |
futures.append((future, name))
|
| 342 |
|
| 343 |
for future, source_name in futures:
|
| 344 |
try:
|
| 345 |
-
# Wait max
|
| 346 |
-
news_items = future.result(timeout=
|
| 347 |
|
| 348 |
# Deduplicate based on text similarity
|
| 349 |
unique_items = []
|
|
|
|
| 208 |
logger.warning("Chromium not found in standard paths")
|
| 209 |
return '/usr/bin/chromium' # Fallback
|
| 210 |
|
| 211 |
+
def _scrape_twitter_profile(self, source_name: str, source_info: Dict, timeout: int = 30) -> List[Dict]:
|
| 212 |
+
"""Scrape tweets from a single Twitter profile using Playwright"""
|
| 213 |
if not PLAYWRIGHT_AVAILABLE:
|
| 214 |
logger.warning("Playwright not available")
|
| 215 |
return []
|
|
|
|
| 234 |
)
|
| 235 |
page = context.new_page()
|
| 236 |
|
| 237 |
+
# Block images, fonts, css, and videos for speed
|
| 238 |
def route_intercept(route):
|
| 239 |
if route.request.resource_type in ["image", "media", "font", "stylesheet", "video"]:
|
| 240 |
route.abort()
|
|
|
|
| 243 |
|
| 244 |
page.route("**/*", route_intercept)
|
| 245 |
|
| 246 |
+
# Navigate to profile with increased timeout
|
| 247 |
logger.info(f"Scraping {source_name}...")
|
| 248 |
+
page.goto(source_info['url'], timeout=timeout * 1000, wait_until="domcontentloaded")
|
| 249 |
|
| 250 |
+
# Wait for tweets to load with increased timeout
|
| 251 |
try:
|
| 252 |
+
page.wait_for_selector("article", timeout=15000) # Increased to 15 seconds
|
| 253 |
except PlaywrightTimeoutError:
|
| 254 |
logger.warning(f"Timeout waiting for tweets from {source_name}")
|
| 255 |
browser.close()
|
| 256 |
return []
|
| 257 |
|
| 258 |
+
# Extract tweet texts (limit to 15)
|
| 259 |
tweet_elements = page.locator("article div[data-testid='tweetText']").all()
|
| 260 |
|
| 261 |
news_items = []
|
|
|
|
| 331 |
reverse=True
|
| 332 |
)
|
| 333 |
|
| 334 |
+
# Scrape sources in parallel with moderate concurrency
|
| 335 |
+
# 8 workers = 19 sources in 3 batches (~60-90 seconds total)
|
| 336 |
with ThreadPoolExecutor(max_workers=8) as executor:
|
| 337 |
futures = []
|
| 338 |
for name, info in sorted_sources:
|
| 339 |
+
# Increased timeout for better success rate
|
| 340 |
+
future = executor.submit(_self._scrape_twitter_profile, name, info, timeout=30)
|
| 341 |
futures.append((future, name))
|
| 342 |
|
| 343 |
for future, source_name in futures:
|
| 344 |
try:
|
| 345 |
+
# Wait max 35 seconds per source (increased for reliability)
|
| 346 |
+
news_items = future.result(timeout=35)
|
| 347 |
|
| 348 |
# Deduplicate based on text similarity
|
| 349 |
unique_items = []
|