High Level Languages

1
High Level Languages

• Lecture 4
• Regex

Chris Coleman cqc_at_cs.nott.ac.uk
2
Last lecture

• Last time we covered
• print
• Files and input
• Sub-routines
• Scoping

3
Today

• REGULAR
• EXPRESSIONS

4
Regular Expressions

• Also known as regex, these provide pattern
  matching capabilities
• Allows you to search through a string, or change
  a string in a number of different ways
• A little scary at first, but they are very
  usefull tools.

5
Regex Bold claim

• By the end of this lecture you'll understand
  this
• (word) /(a-z-a-z)/
• somehash (line /(\w)(\w)/g)
• s/(\w)(inged)/1/gi
• Well, at least with the aid of a reference book.

6
Regex - Matching

• The match operator is m/pattern/
• The m at the front is optional
• The operator says which string to perform the
  match on, e.g.
• somestring /pattern/
• If no variable is specified with the default
  variable is used instead, e.g.
• /pattern/
• True is returned if pattern is found in the string

7
Regex Patterns

• A pattern may simply be a sequence of letters
• if (input /perl/)
• print "line contains 'perl'\n"

8
Regex Wild cards

• You can use a full-stop as a wildcard meaning
  any character
• pattern example matches
• pe.l perl, peal, pe7l, pezl

9
Regex Quantifiers

• A means the previous character repeated zero or
  more times
• perl pel, perl, perrl, perrrl
• pe.l pel, peal, pedfl, perlperl
• A means the previous character repeated one or
  more times
• perl perl, perrl, perrrl
• pe.l peal, pedfl, perlperl

10
Regex Quantifiers

• You can specify an exact number of matches with
  
• bar3 barrr
• bar2,4 barr, barrr, barrrr
• WARNING! both of those would also match
  barrrrrrrrrr since the pattern barrr (for
  example) is present in barrrrrrrrrr.
• Info on how to avoid this in a bit

11
Regex Quantifiers

• and are greedy, i.e. they'll match as much
  as they can
• If you add a ? Afterwards they become minimal
  and match as little as they can
• _ "djm100020passwordDuncan Martin"
• /./ djm100020password
• /.?/ djm

12
Regex Grouping

• You can use brackets to grouping parts of a
  pattern together
• pe(rl) pe, perl, perlrl, perlrlrl
• p(e.l) pexl, pexle4l, pe9leDleXl
• means or
• pe(rR)l perl, peRl
• (perlpython) perl, python
• (heshe) said he said, she said

13
Regex Remembering values

• Using brackets has a side effect, that Perl will
  remember the part in brackets.
• These are stored in variables, 1, 2, 3 etc.
• 1 is the first bracket set from the left
• line "he ran"
• if (line /(heshe) (ranswam)/)
• print "person 1\n" he
• print "action 2\n" ran

14
Regex Remembering values

• In list context a match returns all the found
  values
• _ "hello duncan"
• _at_res /(hellogoodbye) (.)/
• print res0 hello
• print res1 duncan

15
Regex Remembering values

• Doing that when you're operating on a specific
  variable (not the default one) is a bit messier
• line "hello duncan"
• _at_res (line /(hellogoodbye) (.)/)
• print res0 hello
• print res1 duncan

16
Regex Anchors

• is a special character meaning beginning of
  the line
• means end of the line
• /hello/ line starts with hello
• /QED/ line ends with QED
• /piffle/ line consists entirely of piffle

17
Regex Finding borders

• Consider the text
• "The cat ate the mat"
• "The owner was irate and kicked it"
• We want to find the word ate and use
• /ate/
• Both lines match since irate contains ate

18
Regex Finding borders

• One way of solving this is to look for spaces
  either side of the word
• / ate /
• But this wouldn't match
• "ate a lot yesterday"
• "must be something I ate"
• Since there is no space before or after
  respectively

19
Regex Finding borders

• We use anchors to solve this problem
• /( )ate( )/
• Reads as
• Start of the line or a space
• Followed by the word 'ate'
• Followed by the end of line or a space

20
Regex Character classes

• A character class is like a selective wildcard
• Will match a single character from the choice
  given
• Defined with
• /bcabbage/ babbage, cabbage
• /a-zie/ aie, bie, cie, die
• /a-z0-9f/ af, bf, cf, 0f, 1f, 2f, 3f

21
Regex Character classes

• To match a number at the start of a line
• (number) /(0-9)/
• To find a hyphenated word
• (word) /(a-z-a-z)/
• means not
• /bcow/ match aow, dow, eow, fow

22
Regex Special characters

• There are a few of the special regex characters
• \s Whitespace either space, tab or new-line
• \S Any non-whitespace character
• \w Word character letter, number or underscore
• \W Any non word-character
• \d Digit 0 to 9
• \D Any non digit

23
Regex Modifiers

• A modifier is a character that comes after the
  pattern to affect the behaviour of the match
• Modifier i' means ignore case
• _ "Duncan"
• /duncan/ no match
• /duncan/i match

24
Regex Modifiers

• g returns all the patterns that match
• line "The cat ate the mat"
• _at_words (line /\w/g)
• foreach (_at_words)
• print "word _\n"
• We didn't use any ( ) in the regex so it returns
  the whole pattern

25
Regex Modifiers

• If we had used brackets it would return all of
  the groups in sequence. Consider
• line "animalcamel likessand humps2"
• We want to turn this into a hash, as if we'd
  said
• somehash'animal' "camel"
• somehash'likes' "sand"
• somehash'humps' 2

26
Regex Modifiers

• We can do this with the line
• somehash (line /(\w)(\w)/g)
• Why? The match will return a list like
• ('animal','camel','likes','sand','humps',2)
• And you can initialise a hash from a list

27
Regex Modifiers

• x ignores white space in your pattern and lets
  you use comments
• somehash (line /
• (\w) first word in pair
• the dividing equals
• (\w) second word in pair
• /gx)

28
Regex Substitution

• Matching is only half of regex's power...
• We can also get Perl to change what it matches to
  something else
• The format is
• s/pattern/replace with/
• Returns the number of changes made

29
Regex Substitution

• Easy example
• x "The cat sat on the mat. Cats are boring."
• changes (x s/cat/dog/i)
• print changes\n
• print x
• 1
• "The dog sat on the mat. Cats are boring"
• Only the first instance has been changed

30
Regex Substitution

• To change all occurrences, use the g modifier
• x "The cat sat on the mat. Cats are boring."
• x s/cat/dog/gi i ignores case
• print x
• "The dog sat on the mat. dogs are boring"

31
Regex Substitution

• You can access remembered patterns from the
  left-hand side
• Example, to remove the -ing and -ed endings from
  words
• s/(\w)(inged)/1/gi
• started becomes start
• swimming becomes swimm

32
Regex Substitution

• Or do something more complicated
• id ("djm" gt "Duncan", "jog" gt "Jim)
• text "djm sends mail to jog"
• text s/(djmjog)/id1/g
• "Duncan sends mail to Jim"
• print text

33
Regex Substitution

• The e modifier turns the right-hand side into an
  expression rather than a pattern
• x "The cat sat on the mat"
• x s/(\w)/reverse(1)/ge
• "ehT tac tas no eht tam"
• print x

34

• A few useful substitutions to finish with
• Remove white-space from start of line
• s/\s//
• Remove white-space from end of line
• s/\s//

35
Summary

• Today we covered
• Simple matching
• Wildcards
• Quantifiers
• Grouping
• Remembering values
• Anchors
• Modifiers
• Substitution

36
Next Lecture...

• Things get simpler again...
• We look at
• basic HTML
• How data is passed around the web

37
(No Transcript)