huanghai/aiData
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

'commit'

Browse Source
This commit is contained in:
HuangHai
2026-01-15 15:41:30 +08:00
parent 5172dbbc49
commit ba46eb15f6
20 changed files with 335 additions and 182 deletions
Download Patch File Download Diff File Expand all files Collapse all files

57
Apps/AiTeJiYiChong/Crawler.py
View File

@@ -238,20 +238,49 @@ async def get_station_list(d, service, max_stations_count=MAX_STATIONS_COUNT):
price_screenshots = []
price_screenshots.append(after_click_path)
# 滑动处理
last_price_md5 = after_click_md5
for scroll_idx in range(2):
d.swipe(w // 2, int(h * 0.7), w // 2, int(h * 0.3), duration=0.4)
await asyncio.sleep(1)
scroll_path = take_screenshot(d, f"price_scroll_{scroll_idx}_{station_name}", save_dir=TEMP_IMAGE_DIR)
curr_price_md5 = Kit.get_image_content_md5(
scroll_path,
top_ratio=SAFE_EXCLUDE_RATIO,
bottom_ratio=BOTTOM_SAFE_EXCLUDE_RATIO
)
if curr_price_md5 == last_price_md5: break
price_screenshots.append(scroll_path)
last_price_md5 = curr_price_md5
# 1. 向下滚动到底 (根据用户反馈：只有不断向下滚动，才能看到00点的)
logger.info("正在向下滚动价格列表到底部...")
max_scroll_down = 6
for i in range(max_scroll_down):
before_scroll_path = take_screenshot(d, f"scroll_dn_{i}", save_dir=TEMP_IMAGE_DIR)
before_scroll_md5 = Kit.get_image_content_md5(before_scroll_path)
d.swipe_ext("up", scale=0.8) # 向下滚动 = 手势向上
await asyncio.sleep(1.2)
after_scroll_path = take_screenshot(d, f"scroll_dn_after_{i}", save_dir=TEMP_IMAGE_DIR)
after_scroll_md5 = Kit.get_image_content_md5(after_scroll_path)
# 清理临时截图
if os.path.exists(before_scroll_path): os.remove(before_scroll_path)
if os.path.exists(after_scroll_path): os.remove(after_scroll_path)
if before_scroll_md5 == after_scroll_md5:
logger.info(f"价格列表已到达底部 (滚动次数: {i})")
break
# 2. 向上滚动并逐页截图 (从底向上抓取)
logger.info("正在向上滚动价格列表并逐页截图...")
max_scroll_up = 8
for p_idx in range(1, max_scroll_up + 1):
# 截图当前页
p_shot = take_screenshot(d, f"price_scroll_{p_idx}_{station_name}", save_dir=TEMP_IMAGE_DIR)
# 检查是否还能向上滚动
before_up_md5 = Kit.get_image_content_md5(p_shot)
d.swipe_ext("down", scale=0.85) # 向上滚动 = 手势向下
await asyncio.sleep(1.5)
# 检查是否还有新内容
check_up_path = take_screenshot(d, f"check_up_{p_idx}", save_dir=TEMP_IMAGE_DIR)
after_up_md5 = Kit.get_image_content_md5(check_up_path)
if os.path.exists(check_up_path): os.remove(check_up_path)
price_screenshots.append(p_shot)
if before_up_md5 == after_up_md5:
logger.info(f"价格列表已到达顶部 (共抓取页数: {p_idx})")
break
# 后台处理价格图片
logger.info(f"已启动后台分析价格页: {station_name}，共 {len(price_screenshots)} 张图")