From 776a127b8db01cd4338f4db2a84ea567a65bff9f Mon Sep 17 00:00:00 2001
From: baoshiwei <baoshiwei@shlanbao.cn>
Date: 星期二, 15 四月 2025 13:52:04 +0800
Subject: [PATCH] rjuq
---
pachong.py | 36 +++++++++++++++++++++++++-----------
1 files changed, 25 insertions(+), 11 deletions(-)
diff --git a/pachong.py b/pachong.py
index 4841de4..d82db8f 100644
--- a/pachong.py
+++ b/pachong.py
@@ -1,13 +1,20 @@
+import concurrent.futures
import os
import time
import requests
import re
+
def imgdata_set(save_path, word, epoch):
+ if not os.path.exists(save_path):
+ os.makedirs(save_path)
+ else:
+ return 0
q = 0 # 鍋滄鐖彇鍥剧墖鏉′欢
a = 0 # 鍥剧墖鍚嶇О
while(True):
time.sleep(1)
+ print("寮�濮嬬埇鍙栧浘鐗�")
url = "https://image.baidu.com/search/flip?tn=baiduimage&ie=utf-8&word={}&pn={}&ct=&ic=0&lm=-1&width=0&height=0".format(word, q)
# word=闇�瑕佹悳绱㈢殑鍚嶅瓧
headers = {
@@ -16,21 +23,28 @@
response = requests.get(url, headers=headers) # 鍙戦�佽姹傝幏鍙栧搷搴�
html = response.text # 鑾峰彇鍝嶅簲鐨凥TML鍐呭
urls = re.findall('"objURL":"(.*?)"', html) # 浣跨敤姝ｅ垯琛ㄨ揪寮忔彁鍙栧浘鐗嘦RL
- for url in urls:
- try:
- print(a) # 鍥剧墖鐨勫悕瀛�
- response = requests.get(url, headers=headers) # 鍙戦�佽姹傝幏鍙栧浘鐗囧搷搴擡:\yaocai\juhua
- image = response.content # 鑾峰彇鍥剧墖鍐呭
- with open(os.path.join(save_path, "{}.jpg".format(a)), 'wb') as f: # 灏嗗浘鐗囧唴瀹逛繚瀛樺埌鎸囧畾璺緞
- f.write(image)
- a = a + 1
- except Exception as e:
- pass
- continue
+ print(len(urls))
+ # 浣跨敤concurrent.futures瀹炵幇骞跺彂涓嬭浇
+ with concurrent.futures.ThreadPoolExecutor( max_workers=10) as executor:
+ # 鎻愪氦鎵�鏈変笅杞戒换鍔″苟鏀堕泦future瀵硅薄
+ futures = [executor.submit(download_image, index, headers,save_path,url ) for index,url in enumerate(urls)]
q = q + 20
if (q / 20) >= int(epoch):
break
+
+def download_image(a, headers, save_path, url):
+ try:
+ print(a) # 鍥剧墖鐨勫悕瀛�
+ response = requests.get(url, headers=headers, timeout=10) # 鍙戦�佽姹傝幏鍙栧浘鐗囧搷搴�
+ # 濡傛灉娌℃湁涓�鐩村搷搴旀�庝箞澶勭悊
+ image = response.content # 鑾峰彇鍥剧墖鍐呭
+ with open(os.path.join(save_path, "{}.jpg".format(a)), 'wb') as f: # 灏嗗浘鐗囧唴瀹逛繚瀛樺埌鎸囧畾璺緞
+ f.write(image)
+ except Exception as e:
+ pass
+
+
if __name__ == "__main__":
save_path = input('浣犳兂淇濆瓨鐨勮矾寰勶細') # 璇㈤棶鐢ㄦ埛淇濆瓨璺緞
word = input('浣犳兂瑕佷笅杞戒粈涔堝浘鐗囷紵璇疯緭鍏�:') # 璇㈤棶鐢ㄦ埛鎼滅储鍏抽敭璇�
--
Gitblit v1.9.3