pyxllib 0.3.82__tar.gz → 0.3.83__tar.gz

{pyxllib-0.3.82/pyxllib.egg-info → pyxllib-0.3.83}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: pyxllib
-Version: 0.3.82
+Version: 0.3.83
 Summary: 厦门理工模式识别团队通用python代码工具库
 Home-page: https://github.com/XLPRUtils/pyxllib
 Author: code4101

{pyxllib-0.3.82 → pyxllib-0.3.83}/pyxllib/algo/stat.py

@@ -251,9 +251,9 @@ def xlpivot(df, index=None, columns=None, values=None):
     :return: 数据透视表的表格
 
     使用示例：
-    def func(items):
+    def func(items):  # 输入匹配的多行数据
         x = items.iloc[0]
-        return f'{x["precision"]:.0f}，{x["recall"]:.0f}，{x["hmean"]:.2f}，{x["fps"]}'
+        return f'{x["precision"]:.0f}，{x["recall"]:.0f}，{x["hmean"]:.2f}，{x["fps"]}'  # 返回显示的值
     >> df2 = xlpivot(df, ['model_type'], ['dataset', 'total_frame'], {'precision，recall，hmean，fps': func})
     """
 

{pyxllib-0.3.82 → pyxllib-0.3.83}/pyxllib/ext/kq5034lib.py

@@ -45,6 +45,7 @@ class 网课考勤:
         self.开课日期 = '2022-01-08'
         self.视频返款 = [20, 15, 10, 5, 0, 0]  # 直播(当堂)/第1天（当天）/第2天/第3天/第4天/第5天，完成观看的依次返款额。
         self.打卡返款 = {5: 100, 10: 150, 15: 200}  # 打卡满5/10/15次的返款额
+        self.课程链接 = []
         self._init(today)
 
         self.driver = None  # 浏览器脚本
@@ -220,7 +221,13 @@ class 网课考勤:
             if m := re.match(r'(\d{4}\-\d{2}\-\d{2}).+?课.*?(\d+).+?直播观看详情', f.stem):
                 stat_day, 课次 = date.fromisoformat(m.group(1)), int(m.group(2))
                 skiprows = 1
-            elif m := re.search(r'(?:[\d一-十四]+)届.+?(\d+).+?直播用户列表.+?(\d{4}\-\d{2}\-\d{2})', f.stem):  # 新版格式
+            elif m := re.search(r'届(?:念住|觉观).+?(\d+).+?直播用户列表.+?(\d{4}\-\d{2}\-\d{2})', f.stem):
+                stat_day, 课次 = date.fromisoformat(m.group(2)), int(m.group(1))
+                skiprows = 0
+            elif m := re.search(r'第(\d+)堂.+?届(?:念住|觉观).+?直播用户列表.+?(\d{4}\-\d{2}\-\d{2})', f.stem):
+                stat_day, 课次 = date.fromisoformat(m.group(2)), int(m.group(1))
+                skiprows = 0
+            elif m := re.search(r'本体音艺网课-(\d+).+?直播用户列表.+?(\d{4}\-\d{2}\-\d{2})', f.stem):
                 stat_day, 课次 = date.fromisoformat(m.group(2)), int(m.group(1))
                 skiprows = 0
             else:

{pyxllib-0.3.82 → pyxllib-0.3.83}/pyxllib/text/jiebalib.py

@@ -121,6 +121,7 @@ class TextClassifier:
 
     def get_text_tf(self, text, *,
                     function_word_weight=0.2,
+                    normalize=True,
                     ingore_words=(' ', '\t', '\n'),
                     add_flag=False):
         """ 这里可以定制提取text关键词的算法
@@ -146,12 +147,13 @@ class TextClassifier:
                     ct[word] += 1
 
         # 2 归一化一些词
-        ct2 = Counter()
-        for k, v in ct.items():
-            # 如果需要对一些词汇做归一化，也可以这里设置
-            k = re.sub(r'\d', '0', k)  # 把数字都换成0
-            ct2[k] += v
-        ct = ct2
+        if normalize:
+            ct2 = Counter()
+            for k, v in ct.items():
+                # 如果需要对一些词汇做归一化，也可以这里设置
+                k = re.sub(r'\d', '0', k)  # 把数字都换成0
+                ct2[k] += v
+            ct = ct2
 
         # 3 过滤掉一些词
         if ingore_words:
@@ -161,14 +163,14 @@ class TextClassifier:
 
         return ct
 
-    def compute_tfidf(self, outfile=None, sheet_name='tf-idf'):
+    def compute_tfidf(self, outfile=None, sheet_name='tf-idf', normalize=False, function_word_weight=0.2, add_flag=False):
         """ 重算tfidf表 """
         from math import log10
 
         # 1 统计频数和出现该词的文章数
         d = dict()
         for text in tqdm(self.texts, '词频统计'):
-            ct = self.get_text_tf(text)
+            ct = self.get_text_tf(text, normalize=normalize, function_word_weight=function_word_weight, add_flag=add_flag)
             for k, v in ct.items():
                 if k not in d:
                     d[k] = [0, 0]

{pyxllib-0.3.82 → pyxllib-0.3.83}/pyxllib/text/jscode.py

@@ -564,6 +564,7 @@ class AirScriptCodeFixer:
         code_text = re.sub(r'(\w+)\.(Row|Column)\s*\+\s*\1\.\2s\.Count\s*-\s*1', r'\1.\2End', code_text)
         code_text = re.sub(r'\bvar\b', 'let', code_text)
         code_text = code_text.replace('Sheets.Item(', 'Sheets(')
+        code_text = re.sub(r'Application.Enum.\w+.(\w+)', r'\1', code_text)
 
         # 2 代码美化
         opts = jsbeautifier.default_options()
@@ -576,7 +577,6 @@ class AirScriptCodeFixer:
     def simplify_code2(cls, code_text, indent=4):
         """ 有些规则可能在标注数据中想留着，但训练的时候想删除，则可以调用这个进一步级别的简化 """
         _, code_text = cls.simplify_code(code_text, indent)
-        code_text = code_text.replace('Application.Enum.XlLineStyle.xlDouble', 'xlDouble')
         return code_text
 
     @classmethod

{pyxllib-0.3.82 → pyxllib-0.3.83/pyxllib.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: pyxllib
-Version: 0.3.82
+Version: 0.3.83
 Summary: 厦门理工模式识别团队通用python代码工具库
 Home-page: https://github.com/XLPRUtils/pyxllib
 Author: code4101

{pyxllib-0.3.82 → pyxllib-0.3.83}/setup.py

@@ -30,7 +30,7 @@ _dir = Path(__file__).parent
 
 setup(
     name='pyxllib',  # pip 安装时用的名字
-    version='0.3.82',  # 当前版本，每次更新上传到pypi都需要修改; 第4位版本号一般是修紧急bug
+    version='0.3.83',  # 当前版本，每次更新上传到pypi都需要修改; 第4位版本号一般是修紧急bug
     author='code4101',
     author_email='877362867@qq.com',
     url='https://github.com/XLPRUtils/pyxllib',

{pyxllib-0.3.82 → pyxllib-0.3.83}/pyxllib.egg-info/requires.txt

@@ -22,11 +22,11 @@ pandas>=0.23.4
 
 [xlai]
 pillow
-xlcocotools
-visdom
 opencv-python
-opsdroid-get-image-size
+visdom
 pynvml
+opsdroid-get-image-size
+xlcocotools
 
 [xlcv]
 opencv-python