blackteay/DrissionPage
1
0
Fork 0
mirror of https://gitee.com/g1879/DrissionPage.git synced 2024-12-10 04:00:23 +08:00
Code Issues Projects Releases Wiki Activity

改进css_path；修复sr中可能获取错元素问题，未完成

Browse Source
This commit is contained in:
g1879 2023-11-23 18:05:57 +08:00
parent 8699bc82d3
commit 75f05062fb
8 changed files with 233 additions and 56 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
DrissionPage/_base/browser.py
View File

@ -6,9 +6,11 @@
from time import sleep, perf_counter
from .chromium_driver import BrowserDriver, ChromiumDriver
from .._commons.tools import stop_process_on_port
from .._commons.tools import stop_process_on_port, raise_error
from .._units.download_manager import DownloadManager
__ERROR__ = 'error'
class Browser(object):
BROWSERS = {}
@ -88,7 +90,8 @@ class Browser(object):
:param cmd_args: 参数
:return: 执行的结果
"""
return self._driver.run(cmd, **cmd_args)
r = self._driver.run(cmd, **cmd_args)
return r if __ERROR__ not in r else raise_error(r)
@property
def driver(self):

154
DrissionPage/_commons/locator.py
View File

@ -13,17 +13,18 @@ def is_loc(text):
'text^', 'text$', 'xpath:', 'xpath=', 'x:', 'x=', 'css:', 'css=', 'c:', 'c='))
def get_loc(loc, translate_css=False):
def get_loc(loc, translate_css=False, css_mode=False):
"""接收本库定位语法或selenium定位元组，转换为标准定位元组，可翻译css selector为xpath
:param loc: 本库定位语法或selenium定位元组
:param translate_css: 是否翻译css selector为xpath
:param css_mode: 是否尽量用css selector方式
:return: DrissionPage定位元组
"""
if isinstance(loc, tuple):
loc = translate_loc(loc)
loc = translate_css_loc(loc) if css_mode else translate_loc(loc)
elif isinstance(loc, str):
loc = str_to_loc(loc)
loc = str_to_css_loc(loc) if css_mode else str_to_loc(loc)
else:
raise TypeError('loc参数只能是tuple或str。')
@ -127,6 +128,100 @@ def str_to_loc(loc):
return loc_by, loc_str
def str_to_css_loc(loc):
"""处理元素查找语句
:param loc: 查找语法字符串
:return: 匹配符元组
"""
return str_to_loc(loc)
loc_by = 'css selector'
if loc.startswith('.'):
if loc.startswith(('.=', '.:', '.^', '.$')):
loc = loc.replace('.', '@class', 1)
else:
loc = loc.replace('.', '@class=', 1)
elif loc.startswith('#'):
if loc.startswith(('#=', '#:', '#^', '#$')):
loc = loc.replace('#', '@id', 1)
else:
loc = loc.replace('#', '@id=', 1)
elif loc.startswith(('t:', 't=')):
loc = f'tag:{loc[2:]}'
elif loc.startswith(('tx:', 'tx=', 'tx^', 'tx$')):
loc = f'text{loc[2:]}'
# ------------------------------------------------------------------
# 多属性查找
if loc.startswith('@@') and loc != '@@':
loc_str = _make_multi_xpath_str('*', loc)
elif loc.startswith('@|') and loc != '@|':
loc_str = _make_multi_xpath_str('*', loc, False)
# 单属性查找
elif loc.startswith('@') and loc != '@':
loc_str = _make_single_xpath_str('*', loc)
# 根据tag name查找
elif loc.startswith(('tag:', 'tag=')) and loc not in ('tag:', 'tag='):
at_ind = loc.find('@')
if at_ind == -1:
loc_str = loc[4:]
else:
if loc[at_ind:].startswith('@@'):
loc_str = _make_multi_xpath_str(loc[4:at_ind], loc[at_ind:])
elif loc[at_ind:].startswith('@|'):
loc_str = _make_multi_xpath_str(loc[4:at_ind], loc[at_ind:], False)
else:
loc_str = _make_single_xpath_str(loc[4:at_ind], loc[at_ind:])
# 根据文本查找
elif loc.startswith('text='):
loc_by = 'xpath'
loc_str = f'//*[text()={_make_search_str(loc[5:])}]'
elif loc.startswith('text:') and loc != 'text:':
loc_by = 'xpath'
loc_str = f'//*/text()[contains(., {_make_search_str(loc[5:])})]/..'
elif loc.startswith('text^') and loc != 'text^':
loc_by = 'xpath'
loc_str = f'//*/text()[starts-with(., {_make_search_str(loc[5:])})]/..'
elif loc.startswith('text$') and loc != 'text$':
loc_by = 'xpath'
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_by = 'xpath'
loc_str = loc[6:]
elif loc.startswith(('x:', 'x=')) and loc not in ('x:', 'x='):
loc_by = 'xpath'
loc_str = loc[2:]
# 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:
loc_by = 'xpath'
loc_str = f'//*/text()[contains(., {_make_search_str(loc)})]/..'
else:
loc_str = '*'
return loc_by, loc_str
def _make_single_xpath_str(tag: str, text: str) -> str:
"""生成xpath语句
:param tag: 标签名
@ -298,3 +393,56 @@ def translate_loc(loc):
raise ValueError('无法识别的定位符。')
return loc_by, loc_str
def translate_css_loc(loc):
"""把By类型的loc元组转换为css selector或xpath类型的
:param loc: By类型的loc元组
:return: css selector或xpath类型的loc元组
"""
if len(loc) != 2:
raise ValueError('定位符长度必须为2。')
loc_by = By.CSS_SELECTOR
loc_0 = loc[0].lower()
if loc_0 == By.XPATH:
loc_by = By.XPATH
loc_str = loc[1]
elif loc_0 == By.CSS_SELECTOR:
loc_by = loc_0
loc_str = loc[1]
elif loc_0 == By.ID:
loc_str = f'#{css_trans(loc[1])}'
elif loc_0 == By.CLASS_NAME:
loc_str = f'.{css_trans(loc[1])}'
elif loc_0 == By.PARTIAL_LINK_TEXT:
loc_by = By.XPATH
loc_str = f'//a[text()="{css_trans(loc[1])}"]'
elif loc_0 == By.NAME:
loc_str = f'*[@name={css_trans(loc[1])}]'
elif loc_0 == By.TAG_NAME:
loc_str = loc[1]
elif loc_0 == By.PARTIAL_LINK_TEXT:
loc_by = By.XPATH
loc_str = f'//a[contains(text(),"{loc[1]}")]'
else:
raise ValueError('无法识别的定位符。')
if loc_by == By.CSS_SELECTOR:
pass
return loc_by, loc_str
def css_trans(txt):
c = ('!', '"', '#', '$', '%', '&', '\'', '(', ')', '*', '+', ',', '-', '.', '/', ':', ';', '<', '=', '>', '?', '@',
'[', '\\', ']', '^', '`', ',', '{', '|', '}', '~', ' ')
return ''.join([fr'\{i}' if i in c else i for i in txt])

10
DrissionPage/_commons/locator.pyi
View File

@ -6,13 +6,19 @@
from typing import Union
def is_loc(text:str) -> bool: ...
def is_loc(text: str) -> bool: ...
def get_loc(loc: Union[tuple, str], translate_css: bool = False) -> tuple: ...
def get_loc(loc: Union[tuple, str], translate_css: bool = False, css_mode: bool = False) -> tuple: ...
def str_to_loc(loc: str) -> tuple: ...
def translate_loc(loc: tuple) -> tuple: ...
def translate_css_loc(loc: tuple) -> tuple: ...
def css_trans(txt: str) -> str: ...

29
DrissionPage/_commons/tools.py
View File

@ -3,8 +3,8 @@
@Author : g1879
@Contact : g1879@qq.com
"""
from platform import system
from pathlib import Path
from platform import system
from re import search, sub
from shutil import rmtree
from time import perf_counter, sleep
@ -12,6 +12,7 @@ from time import perf_counter, sleep
from psutil import process_iter, AccessDenied, NoSuchProcess, ZombieProcess
from .._configs.options_manage import OptionsManager
from ..errors import ContextLostError, ElementLostError, CDPError, PageClosedError, NoRectError, AlertExistsError
def get_usable_path(path, is_file=True, parents=True):
@ -250,3 +251,29 @@ def configs_to_here(save_name=None):
om = OptionsManager('default')
save_name = f'{save_name}.ini' if save_name is not None else 'dp_configs.ini'
om.save(save_name)
def raise_error(r):
"""抛出error对应报错
:param r: 包含error的dict
:return: None
"""
error = r['error']
if error in ('Cannot find context with specified id', 'Inspected target navigated or closed'):
raise ContextLostError
elif error in ('Could not find node with given id', 'Could not find object with given id',
'No node with given id found', 'Node with given id does not belong to the document',
'No node found for given backend id'):
raise ElementLostError
elif error == 'tab closed':
raise PageClosedError
elif error == 'timeout':
raise TimeoutError
elif error == 'alert exists.':
raise AlertExistsError
elif error in ('Node does not have a layout object', 'Could not compute box model.'):
raise NoRectError
elif r['type'] == 'call_method_error':
raise CDPError(f'\n错误：{r["error"]}\nmethod：{r["method"]}\nargs：{r["args"]}')
else:
raise RuntimeError(r)

3
DrissionPage/_commons/tools.pyi
View File

@ -42,3 +42,6 @@ def stop_process_on_port(port: Union[int, str]) -> None: ...
def configs_to_here(file_name: Union[Path, str] = None) -> None: ...
def raise_error(r: dict) -> None: ...

51
DrissionPage/_elements/chromium_element.py
View File

@ -10,9 +10,9 @@ from time import perf_counter, sleep
from .none_element import NoneElement
from .session_element import make_session_ele
from .._base.base import DrissionElement, BaseElement
from .._commons.settings import Settings
from .._commons.keys import keys_to_typing, keyDescriptionForString, keyDefinitions
from .._commons.locator import get_loc
from .._commons.settings import Settings
from .._commons.tools import get_usable_path
from .._commons.web import make_absolute_link, get_ele_txt, format_html, is_js_func, offset_scroll
from .._units.clicker import Clicker
@ -711,7 +711,7 @@ class ChromiumElement(DrissionElement):
elif mode == 'css':
txt1 = ''
txt3 = ''
txt4 = '''path = '>' + ":nth-child(" + nth + ")" + path;'''
txt4 = '''path = '>' + el.tagName + ":nth-child(" + nth + ")" + path;'''
txt5 = '''return path.substr(1);'''
else:
@ -736,7 +736,7 @@ class ChromiumElement(DrissionElement):
return e(this);}
'''
t = self.run_js(js)
return f':root{t}' if mode == 'css' else t
return f'{t}' if mode == 'css' else t
def _set_file_input(self, files):
"""对上传控件写入路径
@ -1022,31 +1022,42 @@ class ChromiumShadowRoot(BaseElement):
:param raise_err: 找不到元素是是否抛出异常，为None时根据全局设置
:return: ChromiumElement对象或其组成的列表
"""
loc = get_loc(loc_or_str)
loc = get_loc(loc_or_str, css_mode=False)
if loc[0] == 'css selector' and str(loc[1]).startswith(':root'):
loc = loc[0], loc[1][5:]
result = None
timeout = timeout if timeout is not None else self.page.timeout
end_time = perf_counter() + timeout
eles = make_session_ele(self.html).eles(loc)
while not eles and perf_counter() <= end_time:
eles = make_session_ele(self.html).eles(loc)
while not result and perf_counter() <= end_time:
if loc[0] == 'css selector':
if single:
nod_id = self.page.run_cdp('DOM.querySelector', nodeId=self._node_id, selector=loc[1])['nodeId']
result = make_chromium_ele(self.page, node_id=nod_id) if nod_id else NoneElement()
if not eles:
return NoneElement() if single else eles
else:
nod_ids = self.page.run_cdp('DOM.querySelectorAll', nodeId=self._node_id, selector=loc[1])['nodeId']
result = [make_chromium_ele(self.page, node_id=n) for n in nod_ids]
css_paths = [i.css_path[47:] for i in eles]
if single:
node_id = self.page.run_cdp('DOM.querySelector', nodeId=self._node_id, selector=css_paths[0])['nodeId']
return make_chromium_ele(self.page, node_id=node_id) if node_id else NoneElement()
else:
eles = make_session_ele(self.html).eles(loc)
if not eles:
result = NoneElement() if single else eles
continue
else:
results = []
for i in css_paths:
node_id = self.page.run_cdp('DOM.querySelector', nodeId=self._node_id, selector=i)['nodeId']
if node_id:
results.append(make_chromium_ele(self.page, node_id=node_id))
return results
css = [i.css_path[61:] for i in eles]
if single:
node_id = self.page.run_cdp('DOM.querySelector', nodeId=self._node_id, selector=css[0])['nodeId']
result = make_chromium_ele(self.page, node_id=node_id) if node_id else NoneElement()
else:
result = []
for i in css:
node_id = self.page.run_cdp('DOM.querySelector', nodeId=self._node_id, selector=i)['nodeId']
if node_id:
result.append(make_chromium_ele(self.page, node_id=node_id))
return result
def _get_node_id(self, obj_id):
"""返回元素node id"""

4
DrissionPage/_elements/session_element.py
View File

@ -267,14 +267,14 @@ class SessionElement(DrissionElement):
while ele:
if mode == 'css':
brothers = len(ele.eles(f'xpath:./preceding-sibling::*'))
path_str = f'>:nth-child({brothers + 1}){path_str}'
path_str = f'>{ele.tag}:nth-child({brothers + 1}){path_str}'
else:
brothers = len(ele.eles(f'xpath:./preceding-sibling::{ele.tag}'))
path_str = f'/{ele.tag}[{brothers + 1}]{path_str}' if brothers > 0 else f'/{ele.tag}{path_str}'
ele = ele.parent()
return f':root{path_str[1:]}' if mode == 'css' else path_str
return f'{path_str[1:]}' if mode == 'css' else path_str
def make_session_ele(html_or_ele, loc=None, single=True):

31
DrissionPage/_pages/chromium_base.py
View File

@ -10,9 +10,9 @@ from threading import Thread
from time import perf_counter, sleep
from .._base.base import BasePage
from .._commons.settings import Settings
from .._commons.locator import get_loc, is_loc
from .._commons.tools import get_usable_path
from .._commons.settings import Settings
from .._commons.tools import get_usable_path, raise_error
from .._commons.web import location_in_viewport
from .._elements.chromium_element import ChromiumElement, run_js, make_chromium_ele
from .._elements.none_element import NoneElement
@ -25,8 +25,8 @@ from .._units.scroller import PageScroller
from .._units.setter import ChromiumBaseSetter
from .._units.states import PageStates
from .._units.waiter import BaseWaiter
from ..errors import (ContextLostError, ElementLostError, CDPError, PageClosedError, NoRectError, AlertExistsError,
GetDocumentError, ElementNotFoundError)
from ..errors import (ContextLostError, ElementLostError, CDPError, PageClosedError, GetDocumentError,
ElementNotFoundError)
__ERROR__ = 'error'
@ -441,28 +441,7 @@ class ChromiumBase(BasePage):
:return: 执行的结果
"""
r = self.driver.run(cmd, **cmd_args)
if __ERROR__ not in r:
return r
error = r[__ERROR__]
if error in ('Cannot find context with specified id', 'Inspected target navigated or closed'):
raise ContextLostError
elif error in ('Could not find node with given id', 'Could not find object with given id',
'No node with given id found', 'Node with given id does not belong to the document',
'No node found for given backend id'):
raise ElementLostError
elif error == 'tab closed':
raise PageClosedError
elif error == 'timeout':
raise TimeoutError
elif error == 'alert exists.':
raise AlertExistsError
elif error in ('Node does not have a layout object', 'Could not compute box model.'):
raise NoRectError
elif r['type'] == 'call_method_error':
raise CDPError(f'\n错误：{r["error"]}\nmethod：{r["method"]}\nargs：{r["args"]}')
else:
raise RuntimeError(r)
return r if __ERROR__ not in r else raise_error(r)
def run_cdp_loaded(self, cmd, **cmd_args):
"""执行Chrome DevTools Protocol语句，执行前等待页面加载完毕