Python Programming for Biology: Bioinformatics and Beyond

Appendix 5

Regular expressions

The  term  ‘regular  expression’  refers  to  the  specification  of  a  particular  pattern  of

characters which may occur within a text string. Given a pattern a program can then look

for certain kinds of character being present in particular locations. As an example we may

wish  to  search  for  email  addresses  within  a  larger  piece  of  text.  In  reality  this  is  fairly

complex,

1

 but  the  basic  pattern  for  this  would  consist  of  a  leading  group  of  (mostly)

alphanumeric  characters,  without  spaces,  before  the  ‘@’  symbol,  then  more  characters

before  ending  with  a  dot  ‘.’  and  a  final  group  of  letters.  Such  a  pattern  would  expect  to

identify strings like ‘[e-mail]’ or ‘[e-mail]’.

In  practice,  regular  expressions  in  Python  are  specified  using  a  system  where  special

codes are used to represent general or ambiguous kinds of character (it is actually a very

similar  system  to  what  is  used  in  other  languages  like  Perl  and  C).  For  example,  ‘\d’  is

used  to  match  any  digit  character  in  the  range  ‘0’  through  to  ‘9’.  Given  a  regular

expression  the  essential  task  is  to  determine  whether  a  pattern  does  or  does  not  occur

within  a  string.  If  it  does,  often  we  will  also  want  to  extract  the  substrings  that  matched

and perhaps replace them with some other text. It should be noted that some simple text

string queries and manipulations, which certainly could be done with regular expressions,

are  often  more  easily  performed  using  the  standard  string  methods.  Thus,  when  dealing

with  only  a  few  exact  substrings  we  recommend  considering  string  methods  like

text.find(subString), text.replace(old, new) in preference to regular expressions.