Dana Structure:

& lt; td class = '1234_sfsfs' & gt; & lt; span eu = '333' & gt; & lt; / span & gt ; & lt; / td & gt;
& lt; td class = '1234_fddfd' & gt; & lt; span eu = 'fkdjnv' & gt; & lt; / span & gt; & lt; / td & gt;
& lt; td class = '1234_bjnieujrini' & gt; & lt; span eu = 'kdfjnvk' & gt; & lt; / span & gt; & lt; / td & gt;
& lt; TD Class = '11234_fddfd' & gt; & lt; span eu = 'dkdv' & gt; & lt; / span & gt; & lt; / td & gt;
& lt; td class = 'fddfd' & gt; & lt; span eu = 'vd' & gt; & lt; / span & gt; & lt; / td & gt;

As you can use BeautifulSoup to find all the TD tags, which in the name of the class there are '1234 _' , then inside them find the & lt tag; span & gt; He already gets the name of the attribute EU

Soup.find_all ('TD', Class _ = '?') Find ('Span'). Get ('EU')

Answer 1, Authority 100%

from bs4 import beautifulsoup
Import Re.
html_doc = "" "& lt; td class = '1234_sfsfs' eu = '333' & gt; & lt; / td & gt;
& lt; td class = '1234_fddfd' eu = 'fsvs' & gt; & lt; / td & gt;
& lt; td class = '1234_bjnieujrini' eu = 'BBBBC' & GT; & LT; / TD & GT;
& lt; td class = '11234_fddfd' eu = 'fsvs' & gt; & lt; / td & gt;
& lt; td class = 'fddfd' & gt; & lt; / td & gt; "" "
Soup = beautifulsoup (HTML_DOC, 'HTML.PARSER')
Links = [Item ['EU'] for item in soup.findall ('TD', ATTRS = {'Class': Re.compile ("1234_")})]
For i in Links:
  Print (I)