abstract-webtools 0.1.5.91__tar.gz → 0.1.5.93__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- {abstract_webtools-0.1.5.91/src/abstract_webtools.egg-info → abstract_webtools-0.1.5.93}/PKG-INFO +1 -1
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/setup.py +1 -1
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/domainManager.py +49 -1
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/seleniumManager.py +4 -1
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93/src/abstract_webtools.egg-info}/PKG-INFO +1 -1
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/LICENSE +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/README.md +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/pyproject.toml +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/setup.cfg +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/__init__.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/abstract_webtools.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/big_user_agent_list.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/main.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/__init__.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/cipherManager.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/crawlManager.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/dynamicRateLimiter.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/linkManager.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/mySocketClient.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/networkManager.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/requestManager.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/soupManager.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/sslManager.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/tlsAdapter.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/urlManager.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/userAgentManager.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/videoDownloader.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/soup_gui.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/url_grabber.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/url_grabber_new.py +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools.egg-info/SOURCES.txt +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools.egg-info/dependency_links.txt +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools.egg-info/requires.txt +0 -0
- {abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools.egg-info/top_level.txt +0 -0
{abstract_webtools-0.1.5.91/src/abstract_webtools.egg-info → abstract_webtools-0.1.5.93}/PKG-INFO
RENAMED
@@ -1,6 +1,6 @@
|
|
1
1
|
Metadata-Version: 2.1
|
2
2
|
Name: abstract_webtools
|
3
|
-
Version: 0.1.5.
|
3
|
+
Version: 0.1.5.93
|
4
4
|
Summary: Abstract Web Tools is a Python package that provides various utility functions for web scraping tasks. It is built on top of popular libraries such as `requests`, `BeautifulSoup`, and `urllib3` to simplify the process of fetching and parsing web content.
|
5
5
|
Home-page: https://github.com/AbstractEndeavors/abstract_essentials/tree/main/abstract_webtools
|
6
6
|
Author: putkoff
|
@@ -4,7 +4,7 @@ with open("README.md", "r", encoding="utf-8") as fh:
|
|
4
4
|
long_description = fh.read()
|
5
5
|
setuptools.setup(
|
6
6
|
name='abstract_webtools',
|
7
|
-
version='0.1.5.
|
7
|
+
version='0.1.5.93',
|
8
8
|
author='putkoff',
|
9
9
|
author_email='partners@abstractendeavors.com',
|
10
10
|
description='Abstract Web Tools is a Python package that provides various utility functions for web scraping tasks. It is built on top of popular libraries such as `requests`, `BeautifulSoup`, and `urllib3` to simplify the process of fetching and parsing web content.',
|
@@ -1,4 +1,37 @@
|
|
1
|
-
from
|
1
|
+
from abstract_webtools import *
|
2
|
+
import os
|
3
|
+
from abstract_webtools import *
|
4
|
+
#from .urlManager import *
|
5
|
+
from urllib.parse import urlparse
|
6
|
+
from abstract_utilities import *
|
7
|
+
from selenium import webdriver
|
8
|
+
from selenium.webdriver.chrome.options import Options
|
9
|
+
import logging
|
10
|
+
import urllib3
|
11
|
+
|
12
|
+
# Suppress urllib3 warnings and debug logs
|
13
|
+
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
|
14
|
+
logging.getLogger("urllib3").setLevel(logging.WARNING)
|
15
|
+
|
16
|
+
# Suppress Selenium logs
|
17
|
+
logging.getLogger("selenium").setLevel(logging.WARNING)
|
18
|
+
|
19
|
+
import os
|
20
|
+
from selenium import webdriver
|
21
|
+
from selenium.webdriver.chrome.options import Options
|
22
|
+
|
23
|
+
# Setup Chrome options
|
24
|
+
chrome_options = Options()
|
25
|
+
chrome_options.binary_location = "/home/profiles/solcatcher/.cache/selenium/chrome/linux64/130.0.6723.58/chrome"
|
26
|
+
chrome_options.add_argument("--headless") # Run in headless mode
|
27
|
+
chrome_options.add_argument("--no-sandbox")
|
28
|
+
chrome_options.add_argument("--disable-dev-shm-usage")
|
29
|
+
chrome_options.add_argument("--disable-gpu")
|
30
|
+
chrome_options.add_argument("--disable-software-rasterizer")
|
31
|
+
chrome_options.add_argument("--disable-extensions")
|
32
|
+
chrome_options.add_argument("--remote-debugging-port=9222")
|
33
|
+
|
34
|
+
|
2
35
|
class domainManager(metaclass=SingletonMeta):
|
3
36
|
def __init__(self, url):
|
4
37
|
if not hasattr(self, 'initialized'): # Prevent reinitialization
|
@@ -50,3 +83,18 @@ class domainManager(metaclass=SingletonMeta):
|
|
50
83
|
self.drivers[url] = driver
|
51
84
|
driver.get(url)
|
52
85
|
return self.drivers[url]
|
86
|
+
def get_selenium_source(url):
|
87
|
+
url_mgr = urlManager(url)
|
88
|
+
if url_mgr.url:
|
89
|
+
url = str(url_mgr.url)
|
90
|
+
manager = domainManager(url)
|
91
|
+
driver = manager.get_driver(url)
|
92
|
+
try:
|
93
|
+
# Get page source
|
94
|
+
page_source = driver.page_source
|
95
|
+
return page_source
|
96
|
+
finally:
|
97
|
+
# Don't quit the driver unless you're done with all interactions
|
98
|
+
pass
|
99
|
+
driver = get_selenium_source('http://solpump.io/')
|
100
|
+
input(driver)
|
@@ -39,7 +39,7 @@ class SingletonMeta(type):
|
|
39
39
|
cls._instances[cls] = instance
|
40
40
|
return cls._instances[cls]
|
41
41
|
|
42
|
-
class
|
42
|
+
class seleniumManager(metaclass=SingletonMeta):
|
43
43
|
def __init__(self, url):
|
44
44
|
if not hasattr(self, 'initialized'): # Prevent reinitialization
|
45
45
|
self.initialized = True
|
@@ -77,6 +77,8 @@ class domainManager(metaclass=SingletonMeta):
|
|
77
77
|
|
78
78
|
def get_driver(self, url):
|
79
79
|
if url and url not in self.drivers:
|
80
|
+
chrome_options = Options()
|
81
|
+
chrome_options.add_argument("--headless")
|
80
82
|
driver = webdriver.Chrome(options=chrome_options)
|
81
83
|
self.drivers[url] = driver
|
82
84
|
driver.get(url)
|
@@ -96,3 +98,4 @@ def get_selenium_source(url):
|
|
96
98
|
finally:
|
97
99
|
# Don't quit the driver unless you're done with all interactions
|
98
100
|
pass
|
101
|
+
|
{abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93/src/abstract_webtools.egg-info}/PKG-INFO
RENAMED
@@ -1,6 +1,6 @@
|
|
1
1
|
Metadata-Version: 2.1
|
2
2
|
Name: abstract_webtools
|
3
|
-
Version: 0.1.5.
|
3
|
+
Version: 0.1.5.93
|
4
4
|
Summary: Abstract Web Tools is a Python package that provides various utility functions for web scraping tasks. It is built on top of popular libraries such as `requests`, `BeautifulSoup`, and `urllib3` to simplify the process of fetching and parsing web content.
|
5
5
|
Home-page: https://github.com/AbstractEndeavors/abstract_essentials/tree/main/abstract_webtools
|
6
6
|
Author: putkoff
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
{abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/abstract_webtools.py
RENAMED
File without changes
|
File without changes
|
File without changes
|
{abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/managers/__init__.py
RENAMED
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
{abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/url_grabber.py
RENAMED
File without changes
|
{abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools/url_grabber_new.py
RENAMED
File without changes
|
{abstract_webtools-0.1.5.91 → abstract_webtools-0.1.5.93}/src/abstract_webtools.egg-info/SOURCES.txt
RENAMED
File without changes
|
File without changes
|
File without changes
|
File without changes
|