PyWebCopy © 6

Created By : Raja Tomar License : MIT Email: [email protected]

Python websites and webpages cloning at ease. Web Scraping or Saving Complete webpages and websites with python.

Web scraping and archiving tool written in Python Archive any online website and its assets, css, js and images for offilne reading, storage or whatever reasons. It's easy with pywebcopy.

Why it's great? because it -

• respects robots.txt
• saves a webpage with css, js and images with one call
• clones a complete website with assets and links remapped in one call
• have direct apis for simplicity and ease
• subclassing for advanced usage
• custom html tags handler support
• lots of configuration for many custom needs
• provides several scraping packages in one objects for scraping under one class
  • lxml
  • requests
  • beautifulsoup4
  • pyquery
  • requests_html

Email me at [email protected] of any query :)

1.1 Installation

pywebcopy is available on PyPi and is easily installable using pip

$ pip install pywebcopy

You are ready to go. Read the tutorials below to get started.

1.1.1 First steps

You should always check if the latest pywebcopy is installed successfully.

>>> import pywebcopy
>>> pywebcopy.__version__
6.0.0

Your version may be different, now you can continue the tutorial.

1.2 Basic Usages

To save any single page, just type in python console

from pywebcopy import save_webpage

kwargs = {'project_name': 'some-fancy-name'}

save_webpage(
    url='http://example-site.com/index.html',
    project_folder='path/to/downloads',
    **kwargs
)

To save full website (This could overload the target server, So, be careful)

from pywebcopy import save_website

kwargs = {'project_name': 'some-fancy-name'}

save_website(
    url='http://example-site.com/index.html',
    project_folder='path/to/downloads',
    **kwargs
)

1.3 Running Tests

Running tests is simple and doesn't require any external library. Just run this command from root directory of pywebcopy package.

$ python -m pywebcopy run-tests

1.4 Command Line Interface

pywebcopy have a very easy to use command-line interface which can help you do task without having to worrying about the inner long way.

• Getting list of commands

  $ python -m pywebcopy -- --help
  

• Using apis

  $ python -m pywebcopy save_webpage http://google.com E://store// --bypass_robots=True
  or
  $ python -m pywebcopy save_website http://google.com E://store// --bypass_robots
  

• Running tests

  $ python -m pywebcopy run_tests
  

1.5 Authentication and Cookies

Most of the time authentication is needed to access a certain page. Its real easy to authenticate with pywebcopy because it usage an requests.Session object for base http activity which can be accessed through pywebcopy.SESSION attribute. And as you know there are ton of tutorials on setting up authentication with requests.Session.

Here is a basic example of simple http auth -

import pywebcopy

# Update the headers with suitable data

pywebcopy.SESSION.headers.update({
    'auth': {'username': 'password'},
    'form': {'key1': 'value1'},
})

# Rest of the code is as usual
kwargs = {
    'url': 'http://localhost:5000',
    'project_folder': 'e://saved_pages//',
    'project_name': 'my_site'
}
pywebcopy.config.setup_config(**kwargs)
pywebcopy.save_webpage(**kwargs)

2.1 WebPage class

WebPage class, the engine of this saving actions. You can use this class to access many more methods to customise the process with.

• Creating the instance

  You can directly import this class from pywebcopy package.

  from pywebcopy import WebPage    
  wp = WebPage()
  

• fetching the html source from internet

  You can tell it to fetch the source from the internet, it then uses requests module to fetch it for you.

  You can pass in the several params which requests.get() would accept e.g. proxies, auth etc.

   from pywebcopy import WebPage    
   wp = WebPage()
  
   # You can choose to load the page explicitly using 
   # `requests` module with params `requests` would take
  
   url = 'http://google.com'
   params = {
       'auth': '[email protected]',
       'proxies': 'localhost:5000',
   }
   wp.get(url, **params)
  

• providing your own opened file You can also provide opened source handles directly

  from pywebcopy import WebPage    
  wp = WebPage()
  
  # You can choose to set the source yourself
  handle = open('file.html', 'rb')
  wp.set_source(handle)
  

2.1.2 WebPage properties and methods

Apis which WebPage object exposes after creating through any method described above

• .file_path property Read-only location at which this file will end up when you try to save the parsed html source

  To change this location you have to manipulate the .utx property of the WebPage class. You can look it up below.

• .project_path property Read-only location at which all the files will end up when you try to save the complete webpage.

  To change this location you have to manipulate the .utx property of the WebPage class. You can look it up below.

• .save_assets method This methods saves all the css, js, images, fonts etc. in the folder you setup through property .project_path.

  
  from pywebcopy import WebPage
  wp = WebPage()
  wp.get('http://google.com')
  
  wp.save_html()
  #> a .html file would be saved at
  

• .save_html method After setting up the WebPage instance you can use this method to save a local copy of the parsed and modified html at .file_path property value.

  
  from pywebcopy import WebPage
  wp = WebPage()
  wp.get('http://google.com')
  
  wp.save_html()
  #> a .html file would be saved at location which
  #> `.file_path` property returns
  

• .save_complete method This is the important api which you would be using frequently for saving or cloning a webpage for later reading or whatever the use case would be.

  This methods saves all the css, js, images, fonts etc. in the same order as a most browser would do when you will click on the save page option in the right click menu.

  if you want complete webpage with css, js and images

  from pywebcopy import WebPage
  wp = WebPage()
  wp.get('http://google.com')
  
  wp.save_complete()
  

3.1 Scrapings Support

Multiple scraping packages are wrapped up in one object which you can use to unlock the best of all those libraries at one go without having to go through the hassle of instantiating each one of those libraries

To use all the methods and properties documented below just create a object once as described

from pywebcopy import MultiParser

import requests

req = requests.get('http://google.com')

html = req.content

# You can skip the encoding declaration
# it is start enough to auto-detect :)
encoding = req.encoding

wp = MultiParser(html, encoding)

# done

All code follows above code

• BeautifulSoup methods are supported

you can also use any beautiful_soup methods on it

```python
>>> links = wp.bs4.find_all('a')

['//docs.python.org/3/tutorial/', '/about/apps/', 'https://github.com/python/pythondotorg/issues', '/accounts/login/', '/download/other/']

```

• lxml is completely supported

  You can use any lxml methods on it. Read more about lxml at http://lxml.de/

  >>> wp.lxml.xpath('//a', ..)
  [<Element 'a'>,<Element 'a'>]
  
  

• pyquery is Fully supported

  You can use PyQuery methods on it .Read more about pyquery at https://pythonhosted.org/pyquery/

  >>> wp.pq.select(selector, ..)
  ...
  

• lxml.xpath is also supported

  xpath is also natively supported which retures a :class: requests_html.Element See more at https://html.python-requests.org

  >>> wp.xpath('a')
  ['<Element 'a' class='btn' href='https://help.github.com/articles/supported-browsers'>']
  

• select only elements containing certain text

  Provided through the requests_html module.

  >>> wp.find('a', containing='kenneth')
  >>> [<Element 'a' href='http://kennethreitz.com/pages'>, ...]
  

Crawler object

This is a subclass of WebPage class and can be used to mirror any website.

>>> from pywebcopy import Crawler, config
>>> url = 'http://some-url.com/some-page.html'
>>> project_folder = '/home/desktop/'
>>> project_name = 'my_project'
>>> kwargs = {'bypass_robots': True}
# You should always start with setting up the config or use apis
>>> config.setup_config(url, project_folder, project_name, **kwargs)

# Create a instance of the webpage object
>>> wp = Crawler()

# If you want to you can use `requests` to fetch the pages
>>> wp.get(url, **{'auth': ('username', 'password')})

# Then you can access several methods like
>>> wp.crawl()

Common Settings and Errors

You can easily make a beginners mistake or could get confuse, thus here are the common errors and how to correct them if you are facing them.

1. pywebcopy.exceptions.AccessError

   If you are getting pywebcopy.exceptions.AccessError Exception. then check if website allows scraping of its content.

   >>> import pywebcopy
   >>> pywebcopy.config['bypass_robots'] = True
   
   # rest of your code follows..
   
   

2. Overwrite existing files when copying

   If you want to overwrite existing files in the directory then use the over_write config key.

   import pywebcopy
   pywebcopy.config['over_write'] = True
   
   # rest of your code follows..
   
   

3. Changing your project name

   By default the pywebcopy creates a directory inside project_folder with the url you have provided but you can change this using the code below

   >>> import pywebcopy
   >>> pywebcopy.config['project_name'] = 'my_project'
   
   # rest of your code follows..
   
   

How to - Save Single Webpage

Particular webpage can be saved easily using the following methods.

Note: if you get pywebcopy.exceptions.AccessError when running any of these code then use the code provided on later sections.

Method 1 : via api - save_webpage()

Webpage can easily be saved using an inbuilt funtion called .save_webpage() which takes several arguments also.

>>> from pywebcopy import save_webpage
>>> save_webpage(project_url='http://google.com', project_folder='c://Saved_Webpages/',)

Method 2

This use case is slightly more powerful as it can provide every functionallity of the WebPage class.

>>> from pywebcopy import WebPage, config
>>> url = 'http://some-url.com/some-page.html'

# You should always start with setting up the config or use apis
>>> config.setup_config(url, project_folder, project_name, **kwargs)

# Create a instance of the webpage object
>>> wp = WebPage()

# If you want to use `requests` to fetch the page then
>>> wp.get(url)

# Else if you want to use plain html or urllib then use
>>> wp.set_source(object_which_have_a_read_method, encoding=encoding)
>>> wp.url = url   # you need to do this if you are using set_source()

# Then you can access several methods like
>>> wp.save_complete()
>>> wp.save_html()
>>> wp.save_assets()

# This Webpage object contains every methods of the Webpage() class and thus
# can be reused for later usages.

Method 2 using Plain HTML

I told you earlier that Webpage object is powerful and can be manipulated in any ways.

One feature is that the raw html is now also accepted.

>>> from pywebcopy import WebPage, config

>>> HTML = open('test.html').read()

>>> base_url = 'http://example.com' # used as a base for downloading imgs, css, js files.
>>> project_folder = '/saved_pages/'
>>> config.setup_config(base_url, project_folder)

>>> wp = WebPage()
>>> wp.set_source(HTML)
>>> wp.url = base_url
>>> wp.save_complete()

How to - Clone Whole Websites

Use caution when copying websites as this can overload or damage the servers of the site and rarely could be illegal, so check everything before you proceed.

Method 1 : via api - save_website()

Using the inbuilt api .save_website() which takes several arguments.

>>> from pywebcopy import save_website

>>> save_website(project_url='http://localhost:8000', project_folder='e://tests/')

Method 2 -

By creating a Crawler() object which provides several other functions as well.

>>> from pywebcopy import Crawler, config

>>> config.setup_config(project_url='http://localhost:5000/', 
project_folder='e://tests/', project_name='LocalHost')

>>> crawler = Crawler()
>>> crawler.crawl()

1.3 Configuration

pywebcopy is highly configurable. You can setup the global object using the methods exposed by the pywebcopy.config object.

Ways to change the global configurations are below -

• Using the method .setup_config on global pywebcopy.config object

  You can manually configure every configuration by using a .setup_config call.

  
  >>> import pywebcopy
  
  >>> url = 'http://example-site.com/index.html'
  >>> download_loc = 'path/to/downloads/'
  >>> project = 'my_project'
  
  >>> pywebcopy.config.setup_config(url, download_loc, project, **kwargs)
  # done!
  
  # Now check
  >>> pywebcopy.config.get('project_url')
  'http://example-site.com/index.html'
  
  >>> pywebcopy.config.get('project_folder')
  'path/to/downloads'
  
  >>> pywebcopy.config.get('project_name')
  'example-site.com'
  
  ## You can also change any config even after
  ## the `setup_config` call
  
  pywebcopy.config['url'] = 'http://url-changed.com'
  # rest of config remains unchanged
  
  
  Done!
  
  

• Passing in the config vars directly to the

  global apis e.g. .save_webpage

  To change any configuration, just pass it to the api call.

  Example:

  from pywebcopy import save_webpage
  
  kwargs = {
      'project_url': 'http://google.com',
      'project_folder': '/home/pages/',
      'project_name': 
      ...
  }
  
  save_webpage(**kwargs)
  
  

  List of available configurations

  below is the list of config keys with their default values :

  
  # writes the trace output and log file content to console directly
  'DEBUG': False  
  
  # make zip archive of the downloaded content
  'zip_project_folder': True
  
  # delete the project folder after making zip archive of it
  'delete_project_folder': False
  
  # to download css file or not
  'LOAD_CSS': True
  
  # to download images or not
  'LOAD_IMAGES': True
  
  # to download js file or not
  'LOAD_JAVASCRIPT': True
  
  
  # to overwrite the existing files if found
  'OVER_WRITE': False
  
  # list of allowed file extensions
  # shortend for readability
  'ALLOWED_FILE_EXT': ['.html', '.css', ...]
  
  # log file path
  'LOG_FILE': None
  
  # name of the mirror project
  'PROJECT_NAME': website-name.com
  
  # define the base directory to store all copied sites data
  'PROJECT_FOLDER': None
  
  
  # DANGER ZONE
  # CHANGE THESE ON YOUR RESPONSIBILITY
  # NOTE: Do not change unless you know what you're doing
  
  # requests headers to be shown on requests made to server
  'http_headers': {...}
  
  # bypass the robots.txt restrictions
  'BYPASS_ROBOTS' : False
  
  

4.1 Contribution

You can contribute in many ways

• give it a star on github repo
• reporting bugs on github repo: https://github.com/rajatomar788/pywebcopy/ or at my email.
• creating pull requests on github repo: https://github.com/rajatomar788/pywebcopy/
• sending a thanks mail

If you have any suggestions or fixes or reports feel free to mail me :)

5.1 Undocumented Features

I built many utils and classes in this project to ease the tasks I was trying to do.

But, these task are also suitable for general purpose use.

So, if you want, you can help in generating suitable documentation for these undocumented ones, then you can always create and pull request or email me.

6.1 Changelog

[version 6.0.0]

• Breaking Change New command-line interface using Python Fire library.
• Implemented type checks and path normalising in the config.setup_paths.
• added new dynamic pywebcopy.__all__ attr generation.
• WebPage class now doesnt take any argument (breaking change)
• WebPage class has new methods WebPage.get and WebPage.set_source
• Queuing of downloads is replaced with a barrier to manage active threads

[version 5.x]

• Optimization of existing code, upto 5x speed ups in certain cases
• Removed cluttering, improved readability

[version 4.x]

• A complete rewrite and restructuring of core functionality.

[version 2.0.0]

[changed]

• core.setup_config function is changed to config.setup_config.

[added]

• added utils.trace decorator, which will print function_name, args, kwargs and return value when debug config key is True.
• new html-parsers ('html5lib', 'lxml') are supported for better webpages.
• html-parser is now defaulted to 'lxml'. You can use any through new config.config key called parser

[fixed]

• fixed issue while changing user-agent key cracked webpages. You can now use any browser's user-agent id and it will get exact same page downloaded.
• fixed issue in generators.extract_css_urls which was caused by str and bytes difference in python3.
• fixed issues in modules importing. (Thanks "Илья Игоревич").
• added errorhandling to required functions

[version 2.0(beta)]

• init function is replaced with save_webpage
• three new config automation functions are added -
  • core.setup_config (creates every ideal config just from url and download location)
  • config.reset_config (resets the configuration to default state)
  • config.update_config (manual-mode version of core.setup_config)
• object structures.WebPage added
• merged generators.generate_style_map and generators.generate_relative_paths to a single function generators.generate_style_map
• rewrite of majority of functions
• new module exceptions added

[version 1.10]

• url is checked and resolved of any redirection before starting any work functions.
• init vars : mirrors_dir and clean_up were fixed which cleaned the dir before the log was completely written.
• init call now takes url arg by default and could raise a error when not supplied
• professional looking log entries
• rewritten archiving system now uses zipfile and exceptions handling to prevent errors and eventual archive corruption

[version 1.9]

• more redundant code
• modules are now separated based on type e.g. Core, Generators, Utils etc.
• new helper functions and class structures.WebPage
• Compatible with Python 2.6, 2.7, 3.6, 3.7