BeautifulSoup - search by text inside a tag

beautifulsoup, python, regex

Solution

The problem is that your `<a>` tag with the `<i>` tag inside, doesn't have the `string` attribute you expect it to have. First let's take a look at what `text=""` argument for `find()` does.

NOTE: The `text` argument is an old name, since BeautifulSoup 4.4.0 it's called `string`.

From the docs:

Although string is for finding strings, you can combine it with arguments that find tags: Beautiful Soup will find all tags whose .string matches your value for string. This code finds the tags whose .string is “Elsie”:

soup.find_all("a", string="Elsie")
# [<a href="http://example.com/elsie" class="sister" id="link1">Elsie</a>]

Now let's take a look what `Tag`'s `string` attribute is (from the docs again):

If a tag has only one child, and that child is a NavigableString, the child is made available as .string:

title_tag.string
# u'The Dormouse's story'

(...)

If a tag contains more than one thing, then it’s not clear what .string should refer to, so .string is defined to be None:

print(soup.html.string)
# None

This is exactly your case. Your `<a>` tag contains a text and `<i>` tag. Therefore, the find gets `None` when trying to search for a string and thus it can't match.

How to solve this?

Maybe there is a better solution but I would probably go with something like this:

import re
from bs4 import BeautifulSoup as BS

soup = BS("""
<a href="/customer-menu/1/accounts/1/update">
    <i class="fa fa-edit"></i> Edit
</a>
""")

links = soup.find_all('a', href="/customer-menu/1/accounts/1/update")

for link in links:
    if link.find(text=re.compile("Edit")):
        thelink = link
        break

print(thelink)

I think there are not too many links pointing to `/customer-menu/1/accounts/1/update` so it should be fast enough.

Problem

Observe the following problem: ``` import re from bs4 import BeautifulSoup as BS soup = BS(""" <a href="/customer-menu/1/accounts/1/update"> Edit </a> """) # This returns the <a> element soup.find( 'a', href="/customer-menu/1/accounts/1/update", text=re.compile(".*Edit.*") ) soup = BS(""" <a href="/customer-menu/1/accounts/1/update"> <i class="fa fa-edit"></i> Edit </a> """) # This returns None soup.find( 'a', href="/customer-menu/1/accounts/1/update", text=re.compile(".*Edit.*") ) ``` For some reason, BeautifulSoup will not match the text, when the `<i>` tag is there as well. Finding the tag and showing its text produces ``` >>> a2 = soup.find( 'a', href="/customer-menu/1/accounts/1/update" ) >>> print(repr(a2.text)) '\n Edit\n' ``` Right. According to the Docs, soup uses the match function of the regular expression, not the search function. So I need to provide the DOTALL flag: ``` pattern = re.compile('.*Edit.*') pattern.match('\n Edit\n') # Returns None pattern = re.compile('.*Edit.*', flags=re.DOTALL) pattern.match('\n Edit\n') # Returns MatchObject ``` Alright. Looks good. Let's try it with soup ``` soup = BS(""" <a href="/customer-menu/1/accounts/1/update"> <i class="fa fa-edit"></i> Edit </a> """) soup.find( 'a', href="/customer-menu/1/accounts/1/update", text=re.compile(".*Edit.*", flags=re.DOTALL) ) # Still return None... Why?! ``` Edit My solution based on geckons answer: I implemented these helpers: ``` import re MATCH_ALL = r'.*' def like(string): """ Return a compiled regular expression that matches the given string with any prefix and postfix, e.g. if string = "hello", the returned regex matches r".*hello.*" """ string_ = string if not isinstance(string_, str): string_ = str(string_) regex = MATCH_ALL + re.escape(string_) + MATCH_ALL return re.compile(regex, flags=re.DOTALL) def find_by_text(soup, text, tag, **kwargs): """ Find the tag in soup that matches all provided kwargs, and contains the text. If no match is found, return None. If more than one match is found, raise ValueError. """ elements = soup.find_all(tag, **kwargs) matches = [] for element in elements: if element.find(text=like(text)): matches.append(element) if len(matches) > 1: raise ValueError("Too many matches:\n" + "\n".join(matches)) elif len(matches) == 0: return None else: return matches[0] ``` Now, when I want to find the element above, I just run `find_by_text(soup, 'Edit', 'a', href='/customer-menu/1/accounts/1/update')`

Original source