完成视觉相对定位；定位语法增加tag()

DrissionPage/_elements/chromium_element.py

@@ -351,6 +351,21 @@ class ChromiumElement(DrissionElement):
         else:
             return NoneElement(page=self.owner, method='on()', args={'timeout': timeout})
 
+    def offset(self, offset_x, offset_y):
+        """获取相对本元素左上角左边指定偏移量位置的元素
+        :param offset_x: 横坐标偏移量，向右为正
+        :param offset_y: 纵坐标偏移量，向下为正
+        :return: 元素对象
+        """
+        x, y = self.rect.location
+        try:
+            return ChromiumElement(owner=self.owner,
+                                   backend_id=self.owner.run_cdp('DOM.getNodeForLocation', x=x + offset_x,
+                                                                 y=y + offset_y, includeUserAgentShadowDOM=True,
+                                                                 ignorePointerEventsNone=False)['backendNodeId'])
+        except CDPError:
+            return NoneElement(page=self.owner, method='offset()', args={'offset_x': offset_x, 'offset_y': offset_y})
+
     def east(self, locator=None, index=1):
         """获取元素右边某个指定元素
         :param locator: 定位符，只支持str，且不支持xpath和css方式，传入int按像素距离获取
@@ -433,7 +448,6 @@ class ChromiumElement(DrissionElement):
         value = -3 if minus else 3
         size = self.owner.rect.size
         max_len = size[0] if mode == 'east' else size[1]
-        # loc_data = {'tag': None, 'and': True, 'args': [('属性名称', '匹配内容', '匹配方式', '是否否定')]}
         loc_data = locator_to_tuple(locator) if locator else None
         curr_ele = None
         while 0 < cdp_data[variable] < max_len:
@@ -444,13 +458,12 @@ class ChromiumElement(DrissionElement):
                     continue
                 else:
                     curr_ele = bid
-                ele = self.owner.run_cdp('DOM.describeNode', backendNodeId=bid)['node']
+                ele = ChromiumElement(self.owner, backend_id=bid)
 
-                # print(ele)
+                if loc_data is None or _check_ele(ele, loc_data):
-                if loc_data is None:  # todo
                     num += 1
                     if num == index:
-                        return ChromiumElement(owner=self.owner, backend_id=bid)
+                        return ele
             except:
                 pass
 
@@ -1634,3 +1647,59 @@ class Pseudo(object):
     def after(self):
         """返回当前元素的::after伪元素内容"""
         return self._ele.style('content', 'after')
+
+
+def _check_ele(ele, loc_data):
+    """检查元素是否符合loc_data指定的要求
+    :param ele: 元素对象
+    :param loc_data: 格式： {'and': bool, 'args': ['属性名称', '匹配方式', '属性值', 是否否定]}
+    :return: bool
+    """
+    attrs = ele.attrs
+    if loc_data['and']:
+        ok = True
+        for i in loc_data['args']:
+            name, symbol, value, deny = i
+            if name == 'tag()':
+                arg = ele.tag
+                symbol = '='
+            elif name == 'text()':
+                arg = ele.raw_text
+            elif name is None:
+                arg = None
+            else:
+                arg = attrs.get(name, '')
+
+            if ((symbol == '=' and ((deny and arg == value) or (not deny and arg != value)))
+                    or (symbol == ':' and ((deny and value in arg) or (not deny and value not in arg)))
+                    or (symbol == '^' and ((deny and arg.startswith(value))
+                                           or (not deny and not arg.startswith(value))))
+                    or (symbol == '$' and ((deny and arg.endswith(value)) or (not deny and not arg.endswith(value))))
+                    or (arg is None and attrs)):
+                ok = False
+                break
+
+    else:
+        ok = False
+        for i in loc_data['args']:
+            name, value, symbol, deny = i
+            if name == 'tag()':
+                arg = ele.tag
+                symbol = '='
+            elif name == 'text()':
+                arg = ele.text
+            elif name is None:
+                arg = None
+            else:
+                arg = attrs.get(name, '')
+
+            if ((symbol == '=' and ((not deny and arg == value) or (deny and arg != value)))
+                    or (symbol == ':' and ((not deny and value in arg) or (deny and value not in arg)))
+                    or (symbol == '^' and ((not deny and arg.startswith(value))
+                                           or (deny and not arg.startswith(value))))
+                    or (symbol == '$' and ((not deny and arg.endswith(value)) or (deny and not arg.endswith(value))))
+                    or (arg is None and not attrs)):
+                ok = True
+                break
+
+    return ok

DrissionPage/_functions/locator.py

@@ -8,128 +8,79 @@
 from re import split
 from .by import By
 
-格式 = {
-    'and': True,
-    'args': ('属性名称', '属性值', '匹配方式', '是否否定')
-}
-
 
 def locator_to_tuple(loc):
+    """解析定位字符串生成dict格式数据
+    :param loc: 待处理的字符串
+    :return: 格式： {'and': bool, 'args': ['属性名称', '匹配方式', '属性值', 是否否定]}
+    """
     loc = _preprocess(loc)
-    # todo
 
     # 多属性查找
     if loc.startswith(('@@', '@|', '@!')) and loc not in ('@@', '@|', '@!'):
-        loc_str = _make_multi_xpath_str('*', loc)[1]
+        args = _get_args(loc)
 
     # 单属性查找
     elif loc.startswith('@') and loc != '@':
-        loc_str = _make_single_xpath_str('*', loc)[1]
+        arg = _get_arg(loc[1:])
+        arg.append(False)
+        args = {'and': True, 'args': [arg]}
 
     # 根据tag name查找
-    elif loc.startswith(('tag:', 'tag=')) and loc not in ('tag:', 'tag='):
+    elif loc.startswith(('tag:', 'tag=', 'tag^', 'tag$')) and loc not in ('tag:', 'tag=', 'tag^', 'tag$'):
         at_ind = loc.find('@')
         if at_ind == -1:
-            loc_str = f'//*[name()="{loc[4:]}"]'
+            args = {'and': True, 'args': [['tag()', '=', loc[4:].lower(), False]]}
-        elif loc[at_ind:].startswith(('@@', '@|', '@!')):
-            loc_str = _make_multi_xpath_str(loc[4:at_ind], loc[at_ind:])[1]
         else:
-            loc_str = _make_single_xpath_str(loc[4:at_ind], loc[at_ind:])[1]
+            args_str = loc[at_ind:]
+            if args_str.startswith(('@@', '@|', '@!')):
+                args = _get_args(args_str)
+                args['args'].append([f'tag()', '=', loc[4:at_ind].lower(), False])
+            else:  # t:div@aa=bb的格式
+                arg = _get_arg(loc[at_ind + 1:])
+                arg.append(False)
+                args = {'and': True, 'args': [['tag()', '=', loc[4:at_ind].lower(), False], arg]}
 
     # 根据文本查找
-    elif loc.startswith('text='):
+    elif loc.startswith(('text=', 'text:', 'text^', 'text$')):
-        loc_str = f'//*[text()={_make_search_str(loc[5:])}]'
+        args = {'and': True, 'args': [['text()', loc[4], loc[5:], False]]}
-    elif loc.startswith('text:') and loc != 'text:':
-        loc_str = f'//*/text()[contains(., {_make_search_str(loc[5:])})]/..'
-    elif loc.startswith('text^') and loc != 'text^':
-        loc_str = f'//*/text()[starts-with(., {_make_search_str(loc[5:])})]/..'
-    elif loc.startswith('text$') and loc != 'text$':
-        loc_str = f'//*/text()[substring(., string-length(.) - string-length({_make_search_str(loc[5:])}) +1) = ' \
-                  f'{_make_search_str(loc[5:])}]/..'
-
-    # 用xpath查找
-    elif loc.startswith(('xpath:', 'xpath=')) and loc not in ('xpath:', 'xpath='):
-        loc_str = loc[6:]
-    elif loc.startswith(('x:', 'x=')) and loc not in ('x:', 'x='):
-        loc_str = loc[2:]
-
-    # 用css selector查找
-    elif loc.startswith(('css:', 'css=')) and loc not in ('css:', 'css='):
-        loc_by = 'css selector'
-        loc_str = loc[4:]
-    elif loc.startswith(('c:', 'c=')) and loc not in ('c:', 'c='):
-        loc_by = 'css selector'
-        loc_str = loc[2:]
 
     # 根据文本模糊查找
-    elif loc:
-        loc_str = f'//*/text()[contains(., {_make_search_str(loc)})]/..'
     else:
-        loc_str = '//*'
+        args = {'and': True, 'args': [['text()', '=', loc, False]]}
 
-    return {}
+    return args
 
 
-def _get_args(tag: str = None, text: str = '') -> tuple:
+def _get_args(text: str = '') -> dict:
-    """生成多属性查找的xpath语句
+    """解析定位参数字符串生成dict格式数据
-    :param tag: 标签名
     :param text: 待处理的字符串
-    :return: xpath字符串
+    :return: 格式： {'and': bool, 'args': ['属性名称', '匹配方式', '属性值', 是否否定]}
     """
-    # todo
     arg_list = []
     args = split(r'(@!|@@|@\|)', text)[1:]
     if '@@' in args and '@|' in args:
         raise ValueError('@@和@|不能同时出现在一个定位语句中。')
-    elif '@@' in args:
+    _and = '@|' not in args
-        _and = True
-    else:  # @|
-        _and = False
 
     for k in range(0, len(args) - 1, 2):
-        r = split(r'([:=$^])', args[k + 1], maxsplit=1)
+        arg = _get_arg(args[k + 1])
-        arg_str = ''
+        if arg:
-        len_r = len(r)
+            arg.append(True if args[k] == '@!' else False)  # 是否去除某个属性
+        arg_list.append(arg)
 
-        if not r[0]:  # 不查询任何属性
+    return {'and': _and, 'args': arg_list}
-            arg_str = 'not(@*)'
 
-        else:
-            ignore = True if args[k] == '@!' else False  # 是否去除某个属性
-            if len_r != 3:  # 只有属性名没有属性内容，查询是否存在该属性
-                arg_str = 'normalize-space(text())' if r[0] in ('text()', 'tx()') else f'@{r[0]}'
 
-            elif len_r == 3:  # 属性名和内容都有
+def _get_arg(text) -> list:
-                arg = '.' if r[0] in ('text()', 'tx()') else f'@{r[0]}'
+    """解析arg=abc格式字符串，生成格式：['属性名称', '匹配方式', '属性值', 是否否定]，不是式子的返回None"""
-                symbol = r[1]
+    r = split(r'([:=$^])', text, maxsplit=1)
-                if symbol == '=':
+    if not r[0]:
-                    arg_str = f'{arg}={_make_search_str(r[2])}'
+        return [None, None, None, None]
-
+    # !=时只有属性名没有属性内容，查询是否存在该属性
-                elif symbol == ':':
+    name = r[0] if r[0] != 'tx()' else 'text()'
-                    arg_str = f'contains({arg},{_make_search_str(r[2])})'
+    name = name if name != 't()' else 'teg()'
-
+    return [name, None, None] if len(r) != 3 else [name, r[1], r[2]]
-                elif symbol == '^':
-                    arg_str = f'starts-with({arg},{_make_search_str(r[2])})'
-
-                elif symbol == '$':
-                    arg_str = f'substring({arg}, string-length({arg}) - string-length({_make_search_str(r[2])}) +1) ' \
-                              f'= {_make_search_str(r[2])}'
-
-                else:
-                    raise ValueError(f'符号不正确：{symbol}')
-
-            if arg_str and ignore:
-                arg_str = f'not({arg_str})'
-
-        if arg_str:
-            arg_list.append(arg_str)
-
-    arg_str = ' and '.join(arg_list) if _and else ' or '.join(arg_list)
-    if tag != '*':
-        condition = f' and ({arg_str})' if arg_str else ''
-        arg_str = f'name()="{tag}"{condition}'
-
-    return 'xpath', f'//*[{arg_str}]' if arg_str else f'//*'
 
 
 def is_loc(text):
@@ -183,7 +134,7 @@ def str_to_xpath_loc(loc):
         loc_str = _make_single_xpath_str('*', loc)[1]
 
     # 根据tag name查找
-    elif loc.startswith(('tag:', 'tag=')) and loc not in ('tag:', 'tag='):
+    elif loc.startswith(('tag:', 'tag=', 'tag^', 'tag$')) and loc not in ('tag:', 'tag=', 'tag^', 'tag$'):
         at_ind = loc.find('@')
         if at_ind == -1:
             loc_str = f'//*[name()="{loc[4:]}"]'
@@ -206,16 +157,11 @@ def str_to_xpath_loc(loc):
     # 用xpath查找
     elif loc.startswith(('xpath:', 'xpath=')) and loc not in ('xpath:', 'xpath='):
         loc_str = loc[6:]
-    elif loc.startswith(('x:', 'x=')) and loc not in ('x:', 'x='):
-        loc_str = loc[2:]
 
     # 用css selector查找
     elif loc.startswith(('css:', 'css=')) and loc not in ('css:', 'css='):
         loc_by = 'css selector'
         loc_str = loc[4:]
-    elif loc.startswith(('c:', 'c=')) and loc not in ('c:', 'c='):
-        loc_by = 'css selector'
-        loc_str = loc[2:]
 
     # 根据文本模糊查找
     elif loc:
@@ -243,7 +189,7 @@ def str_to_css_loc(loc):
         loc_by, loc_str = _make_single_css_str('*', loc)
 
     # 根据tag name查找
-    elif loc.startswith(('tag:', 'tag=')) and loc not in ('tag:', 'tag='):
+    elif loc.startswith(('tag:', 'tag=', 'tag^', 'tag$')) and loc not in ('tag:', 'tag=', 'tag^', 'tag$'):
         at_ind = loc.find('@')
         if at_ind == -1:
             loc_str = loc[4:]
@@ -253,14 +199,12 @@ def str_to_css_loc(loc):
             loc_by, loc_str = _make_single_css_str(loc[4:at_ind], loc[at_ind:])
 
     # 根据文本查找
-    elif loc.startswith(('text=', 'text:', 'text^', 'text$', 'xpath=', 'xpath:', 'x:', 'x=')):
+    elif loc.startswith(('text=', 'text:', 'text^', 'text$', 'xpath=', 'xpath:')):
         loc_by, loc_str = str_to_xpath_loc(loc)
 
     # 用css selector查找
     elif loc.startswith(('css:', 'css=')) and loc not in ('css:', 'css='):
         loc_str = loc[4:]
-    elif loc.startswith(('c:', 'c=')) and loc not in ('c:', 'c='):
-        loc_str = loc[2:]
 
     # 根据文本模糊查找
     elif loc:
@@ -289,6 +233,9 @@ def _make_single_xpath_str(tag: str, text: str) -> tuple:
         len_r = len(r)
         len_r0 = len(r[0])
         if len_r == 3 and len_r0 > 1:
+            if r[0] in ('@tag()', '@t()'):
+                arg_str = f'name()="{r[2].lower()}"'
+            else:
                 symbol = r[1]
                 if symbol == '=':  # 精确查找
                     arg = '.' if r[0] in ('@text()', '@tx()') else r[0]
@@ -303,12 +250,12 @@ def _make_single_xpath_str(tag: str, text: str) -> tuple:
 
                 elif symbol == '$':  # 匹配结尾
                     if r[0] in ('@text()', '@tx()'):
-                    txt_str = f'/text()[substring(., string-length(.) - string-length({_make_search_str(r[2])}) +1) ' \
+                        txt_str = (f'/text()[substring(., string-length(.) - string-length({_make_search_str(r[2])}) '
-                              f'= {_make_search_str(r[2])}]/..'
+                                   f'+1) = {_make_search_str(r[2])}]/..')
                         arg_str = ''
                     else:
-                    arg_str = f'substring({r[0]}, string-length({r[0]}) - string-length({_make_search_str(r[2])}) +1)' \
+                        arg_str = (f'substring({r[0]}, string-length({r[0]}) - string-length({_make_search_str(r[2])}) '
-                              f' = {_make_search_str(r[2])}'
+                                   f'+1) = {_make_search_str(r[2])}')
 
                 elif symbol == ':':  # 模糊查找
                     if r[0] in ('@text()', '@tx()'):
@@ -321,6 +268,9 @@ def _make_single_xpath_str(tag: str, text: str) -> tuple:
                     raise ValueError(f'符号不正确：{symbol}')
 
         elif len_r != 3 and len_r0 > 1:
+            if r[0] in ('@tag()', '@t()'):
+                arg_str = ''
+            else:
                 arg_str = 'normalize-space(text())' if r[0] in ('@text()', '@tx()') else f'{r[0]}'
 
     if arg_str:
@@ -339,10 +289,9 @@ def _make_multi_xpath_str(tag: str, text: str) -> tuple:
     args = split(r'(@!|@@|@\|)', text)[1:]
     if '@@' in args and '@|' in args:
         raise ValueError('@@和@|不能同时出现在一个定位语句中。')
-    elif '@@' in args:
+    _and = '@|' not in args
-        _and = True
+    tags = [] if tag == '*' else [f'name()="{tag}"']
-    else:  # @|
+    tags_connect = ' or '
-        _and = False
 
     for k in range(0, len(args) - 1, 2):
         r = split(r'([:=$^])', args[k + 1], maxsplit=1)
@@ -355,23 +304,39 @@ def _make_multi_xpath_str(tag: str, text: str) -> tuple:
         else:
             ignore = True if args[k] == '@!' else False  # 是否去除某个属性
             if len_r != 3:  # 只有属性名没有属性内容，查询是否存在该属性
+                if r[0] in ('tag()', 't()'):
+                    continue
                 arg_str = 'normalize-space(text())' if r[0] in ('text()', 'tx()') else f'@{r[0]}'
 
             elif len_r == 3:  # 属性名和内容都有
-                arg = '.' if r[0] in ('text()', 'tx()') else f'@{r[0]}'
+                if r[0] in ('tag()', 't()'):
+                    if ignore:
+                        tags.append(f'not(name()="{r[2]}")')
+                        tags_connect = ' and '
+                    else:
+                        tags.append(f'name()="{r[2]}"')
+                    continue
+
                 symbol = r[1]
+                if r[0] in ('text()', 'tx()'):
+                    arg = '.'
+                    txt = r[2]
+                else:
+                    arg = f'@{r[0]}'
+                    txt = r[2]
+
                 if symbol == '=':
-                    arg_str = f'{arg}={_make_search_str(r[2])}'
+                    arg_str = f'{arg}={_make_search_str(txt)}'
 
                 elif symbol == ':':
-                    arg_str = f'contains({arg},{_make_search_str(r[2])})'
+                    arg_str = f'contains({arg},{_make_search_str(txt)})'
 
                 elif symbol == '^':
-                    arg_str = f'starts-with({arg},{_make_search_str(r[2])})'
+                    arg_str = f'starts-with({arg},{_make_search_str(txt)})'
 
                 elif symbol == '$':
-                    arg_str = f'substring({arg}, string-length({arg}) - string-length({_make_search_str(r[2])}) +1) ' \
+                    arg_str = f'substring({arg}, string-length({arg}) - string-length({_make_search_str(txt)}) +1) ' \
-                              f'= {_make_search_str(r[2])}'
+                              f'= {_make_search_str(txt)}'
 
                 else:
                     raise ValueError(f'符号不正确：{symbol}')
@@ -383,9 +348,9 @@ def _make_multi_xpath_str(tag: str, text: str) -> tuple:
             arg_list.append(arg_str)
 
     arg_str = ' and '.join(arg_list) if _and else ' or '.join(arg_list)
-    if tag != '*':
+    if tags:
         condition = f' and ({arg_str})' if arg_str else ''
-        arg_str = f'name()="{tag}"{condition}'
+        arg_str = f'({tags_connect.join(tags)}){condition}'
 
     return 'xpath', f'//*[{arg_str}]' if arg_str else f'//*'
 
@@ -417,10 +382,7 @@ def _make_multi_css_str(tag: str, text: str) -> tuple:
     args = split(r'(@!|@@|@\|)', text)[1:]
     if '@@' in args and '@|' in args:
         raise ValueError('@@和@|不能同时出现在一个定位语句中。')
-    elif '@@' in args:
+    _and = '@|' not in args
-        _and = True
-    else:  # @|
-        _and = False
 
     for k in range(0, len(args) - 1, 2):
         r = split(r'([:=$^])', args[k + 1], maxsplit=1)
@@ -431,9 +393,18 @@ def _make_multi_css_str(tag: str, text: str) -> tuple:
         len_r = len(r)
         ignore = True if args[k] == '@!' else False  # 是否去除某个属性
         if len_r != 3:  # 只有属性名没有属性内容，查询是否存在该属性
+            if r[0] in ('tag()', 't()'):
+                continue
             arg_str = f'[{r[0]}]'
 
         elif len_r == 3:  # 属性名和内容都有
+            if r[0] in ('tag()', 't()'):
+                if tag == '*':
+                    tag = f':not({r[2].lower()})' if ignore else f'{r[2]}'
+                else:
+                    tag += f',:not({r[2].lower()})' if ignore else f',{r[2]}'
+                continue
+
             d = {'=': '', '^': '^', '$': '$', ':': '*'}
             arg_str = f'[{r[0]}{d[r[1]]}={css_trans(r[2])}]'
 
@@ -459,6 +430,9 @@ def _make_single_css_str(tag: str, text: str) -> tuple:
         return _make_single_xpath_str(tag, text)
 
     r = split(r'([:=$^])', text, maxsplit=1)
+    if r[0] in ('@tag()', '@t()'):
+        return 'css selector', r[2]
+
     if len(r) == 3:
         d = {'=': '', '^': '^', '$': '$', ':': '*'}
         arg_str = f'[{r[0][1:]}{d[r[1]]}={css_trans(r[2])}]'
@@ -581,4 +555,10 @@ def _preprocess(loc):
     elif loc.startswith(('tx:', 'tx=', 'tx^', 'tx$')):
         loc = f'text{loc[2:]}'
 
+    elif loc.startswith(('c:', 'c=')):
+        loc = f'css:{loc[2:]}'
+
+    elif loc.startswith(('x:', 'x=')):
+        loc = f'xpath:{loc[2:]}'
+
     return loc

DrissionPage/_functions/locator.pyi

@@ -20,6 +20,9 @@ def get_loc(loc: Union[tuple, str], translate_css: bool = False, css_mode: bool
 def str_to_xpath_loc(loc: str) -> tuple: ...
 
 
+def str_to_css_loc(loc: str) -> tuple: ...
+
+
 def translate_loc(loc: tuple) -> tuple: ...