py-parse 0.1.6

A simplest HTML parsing library.

README

A simplest html parsing library.

Key features:

• no third-party dependencies
• no need to know CSS, Xpath or complicated rules to find element
• interaction with native python lambda syntax or function-predicate
• opportunity to work with damaged html
• ability to use element relations (find ancestor, descendant, siblings)
• a standard find first element or find all by current filter

Installation

Via pip:

pip install py_parse

First example

Lets get src attribute (link) of the Google logo on google.com

import requests
from py_parse import parse

# get content of the google web page
content = requests.get('https://www.google.com/').text
# find first element with img-tag and 'alt' attribute equal to Google (logo)
google_logo = parse(content).find(lambda e: e.tag == 'img' and e.alt == 'Google')
# prints src attribute of the logo element
print(google_logo.src)

You will see following result

/images/branding/googlelogo/1x/googlelogo_white_background_color_272x92dp.png

If there is no element with current filter, you will get exception with filter text (if lambda was used): For code above lets say we use wrong filter

google_logo = parse(content).find(lambda e: e.tag == 'img' and e.alt == 'Wrong')

You will see following result

...traceback...
py_parse.exceptions.NoSuchElementError: No elements with current filter (e.tag == 'img' and e.alt == 'Wrong')

HOW IT WORKS

During parsing all html-elements in DOM converts to Node objects, which remains all relations (parent, child, sibling) and get all their attributes from html-element.

For example

<div class="some" type="submit">My text</div>

will return (after parsing):

from py_parse import parse

element = '<div class="some" type="submit">My text</div>'
html_element = parse(element)[0]
print(html_element.text)  # My text
print(html_element.tag)  # div
print(html_element.class_)  # some

As you can see, all html attributes became object attributes, so you can use it in your filters

But remember:

• Attribute tag is required, always present and can't be None
• Attribute text always present BUT can be None
• Attribute class became class_ in an object (html_element.class_), and it is not required

As you know, web page is a hierarchy, where html is a ancestor for all elements, and they all are nested in html. Function parse returns Nodes object, that is just container (like list) for Node objects. And for most cases that Nodes will have just one element (html), which contains all other elements inside (nested). So, for using search, you need to use methods like find or find_all of the Nodes.

Find methods

Method find_all of the Nodes objects returns all found elements. If you not specify a filter, then all elements be in result. With filter, you get only elements, that satisfying the condition in it. If there are no such elements, then empty Nodes container returns.

Method find based on find_all, but returns just first element with that filter. If there are no results, then exception will be raised.

Methods find_tags and find_tag are similar to find_all, find, but it makes search with tag more convinient. For example

find_elements(lambda e: e.tag=='div')

is equivalent to

find_tags('div')

Simple Filtering

For all examples we will use content of the python documentation page https://docs.python.org/3/

So, start of all code is

import requests
from py_parse import parse

content = requests.get('https://docs.python.org/3/').text

1. Find by tag

Let's find first element with 'strong' tag and get it text

first_strong = parse(content).find(lambda e: e.tag == 'strong')
print(first_strong.text)   #  Parts of the documentation:

or you can use special find_tag method:

first_strong = parse(content).find_tag('strong')
print(first_strong.text)   #  Parts of the documentation:

2. Find by tag and text (always present in any element)

tables = parse(content).find(lambda e: e.tag == 'strong' and e.text and e.text == 'Indices and tables:')
print(tables.text)   # Indices and tables:

or

tables = parse(content).find_tag('strong', lambda e: e.text == 'Indices and tables:')
print(tables.text)   # Indices and tables:

3. Find by containing text

copyright_ = parse(content).find(lambda e: 'pyri' in e.text)  # pyri is a part of Copyright
print(copyright_)  # <a class="biglink" href="copyright.html">Copyright</a>

In this example we print Node object itself, but not its text attribute.

4. Find element which has id

element_with_id = parse(content).find(lambda e: 'id' in e)  # 'id' in e - checks element has "id" attribute
print(element_with_id)  
# <script id="documentation_options" data-url_root="./" src="_static/documentation_options.js"></script>

5. Find element by tag and type, then get it value

Let's find 'Go' button to search on documentation page

# finds element with input tag, which has type and this type equal to submit
go = parse(content).find(lambda e: e.tag == 'input' and e.type == 'submit')
print(go.value)  # Go

or

go = parse(content).find_tag('input', lambda e: e.type == 'submit')
print(go.value)  # Go

6. Finds all script elements

scripts = parse(content).find_all(lambda e: e.tag == 'script') # Using find_all to finds all elements
for script in scripts:
    print(script)
# <script id="documentation_options" data-url_root="./" src="_static/documentation_options.js"></script>
# <script src="_static/jquery.js"></script>
# <script src="_static/underscore.js"></script>
# <script src="_static/doctools.js"></script>
# <script src="_static/language_data.js"></script>
# <script src="_static/sidebar.js"></script>
# <script type="text/javascript" src="_static/copybutton.js"></script>
# <script type="text/javascript">$('.inline-search').show(0);</script>
# <script type="text/javascript">$('.inline-search').show(0);</script>
# <script type="text/javascript" src="_static/switchers.js"></script>

or with special find_tags method:

scripts = parse(content).find_tags('script') # Using find_tags to finds all elements with that tag
for script in scripts:
    print(script)

7. Finds all elements by part of the class name

sphinxes = parse(content).find_all(lambda e: 'sphinx' in e.class_) # finds all elements containing 'sphinx' in class name
for sphinx in sphinxes:
    print(sphinx)
# <div class="sphinxsidebar" role="navigation" aria-label="main navigation"></div>
# <div class="sphinxsidebarwrapper"></div>

Relations filtering

Node object

Contact me

[e-mail]