add test ocr

client/cli.py

@@ -1,6 +1,8 @@
 import ctypes
 import configparser
 from crawler.__main__ import main
+from crawler.imgtool import ocr
+from PIL import Image
 import os
 
 def has_admin():
@@ -17,6 +19,17 @@ def has_admin():
 
     parser = configparser.ConfigParser()
     parser.read('config.ini', encoding="UTF-8")
+
+    test_ocr = parser.get('basic', 'test_ocr', fallback='False')
+    test_ocr = test_ocr.lower() == 'true'
+    if test_ocr:
+        if not os.path.isfile("test.png"):
+            ctypes.windll.user32.MessageBoxW(0, "当前目录下没有test.png", "测试文件不存在", 0)
+            exit(-1)
+        text = ocr(Image.open("test.png"))
+        ctypes.windll.user32.MessageBoxW(0, text, "识别结果", 0)
+        exit(0)
+
     lock_input = parser.get('basic', 'lock_input', fallback='False')
     print("lock input {}".format(lock_input))
     is_admin = has_admin()

client/crawler/wechatautomator.py

@@ -395,7 +395,7 @@ def process_page(self, account_name, items, lastpage_clicked_titles, states,
     def extract_read_count(self, fn=None):
         self.browser_page_down(30)
         # 初步定位
-        for i in range(20):
+        for i in range(30):
             img_array = imgtool.snap_shot(self.browser.rectangle())
             if fn:
                 bottom = imgtool.locate_content_bottom(img_array, fn+"_coarse_"+str(i))
@@ -547,7 +547,7 @@ def locate_user(self, user, retry=5):
     articles = []
     #states = [{'url': 'https://mp.weixin.qq.com/s/e7EJ2URIvuEXkLweRMNKLg'}]
     states = []
-    result = automator.crawl_gongzhonghao("新智元", articles, max_pages=5, states=states,
+    result = automator.crawl_gongzhonghao("黄梅芦花庵", articles, max_pages=5, states=states,
                                           detail=[], latest_date="2021-05-04", crawl_counter=True,
                                           debug_ocr=True)
     print(result)