I have a bad formed HTML who contains something like

  <a style="text-decoration: none;" class="smalltitle" href="xxxxxxxxxx">
  <font style="margin: 0px; padding: 0px;" color="#000000">
  U.R.P.</font></a>

First I pass the whole HTML to the prettify() function and I obtain

  <a class="smalltitle" href="xxxxxxxxxx" style="text-decoration: none;">
   <font color="#000000" style="margin: 0px; padding: 0px;">
    U.R.P.
   </font>
  </a>

Then I need to search for *URP* or *U.R.P.* and check that it's in an <a>
(hyperlink) tag. The problem is that the find_all() function does a ==
match, but the prettify() function add always whitespaces for indentation.

This is what I do without using regular expression and just to check if my
code is right:

  from bs4 import BeautifulSoup
  import re

  html = BeautifulSoup(BeautifulSoup(open("index.html"), "lxml").prettify())

  for parent in html.find_all("font", text="\n  U.R.P.\n  ")[0].parents:
      if parent.name == "a":
          print 'found!'
          break

The next step is to substitute that ugly text="\n  U.R.P.\n  " with a
regular expression:

  from bs4 import BeautifulSoup
  import re

  html = BeautifulSoup(BeautifulSoup(open("index.html"), "lxml").prettify())

  print html.find_all(text=re.compile("\s*U\.R\.P\.\s*"))  # first try  

  print html.find_all("a", text=re.compile("\s*U\.R\.P\.\s*"))  # second try

The first find_all(), find correctly the string, BUT I need the tag, not
the string.
The second find_all() should works exactly the way I need BUT it doesn't
find anything.

What am I doing wrong?

If I was not too clear, I need to search for *URP* or *U.R.P.* in the whole
page (including eventual whitespaces and tabs) and check that it's in an
<a> (hyperlink) tag. There is a better way to do, instead of the one I'm
trying?

Cheers

I saw that this is a common problem. A partial solution can be

  print
html.find_all(text=re.compile("\s*U\.R\.P\.\s*"))[0].find_parent("a")

Using find_parent() function. This does not find an <a> tag with the
expression in it but find the regex and then try to check if it has a
parent tag named "a" (at any superior level).

The good part is that find_parent() return an bs4.element.Tag instead of a
bs4.element.NavigableString (returned by find_all + regular expression)