Tutorial on Regular Expressions

1
Tutorial on Regular Expressions
2
Matching Characters

• ab - a followed by b
• . wild-card character
• a. - a followed by any character

3
Character Sets

• xyz single character, only character in set
  x, y, z is allowed
• Hhello either Hello or hello
• 0-9 specify range of character allowed, from
  0 to 9
• xyz not x, y, z
• a-zA-Z character other than alphabet
• include as the 1st character in a set
• ?()\\ no need to escape inside bracket
  character set, except backslash

4
Quantifiers

• zero or more
• one or more
• ? zero or one
• ba b followed by zero or more a
• (ab) one or more ab, specified within group ()
• . match everything, including empty string
• .\n all lines in input (greedy match)
• \n\n zero or more non-newline followed by a
  newline, i.e. the first line

5
Quoting RE with Braces

• \n\n wont work, but
• \n\n will do
• \\n\\n will also do

6
Alternation

• Hellohello either Hello or hello
• (Hh)ello same as above
• Hh also same as above

7
Anchoring a Match

• beginning of the string
• end of the string
• \t leading spaces tabs (at least one)
• tion string end up with tion
• abc.xyz string begin with abc, end up with
  xyz

8
Backslash Quoting

• To turn off special characters . ? ( )
  \
• \ matching
• No backslash inside a bracketed expression
• (\\?) is the same as ?
• \\ to match a single backslash, you need two
• \B aslo backslash as in advanced R.E.

9
Matching Precedence

• Earliest in the string
• Longest in the string
• Nongreedy quantifier ? shortest
• ab will match abbb on abbbx
• ab will match ab on ababbbx

10
Capturing Subpatterns

• Use () parentheses to capture a subpattern
• lttdgt(lt)lt/tdgt get everything between lttdgt
  and lt/tdgt tag
• (0-9)\.(0-9) will match subpattern before
  after decimal point

11
Advanced RE, Escape Sequences

• Quote the R.E. pattern with to us advanced RE
• \s space-like characters (white spaces)
• \w letters, digit, and the underscore
• \B backslash
• \d digit
• \S Non-space
• \W Non-alphanumeric
• \D Non-digit
• \A Beginning of a string
• \Z Ending of a string

12
Advanced RE, Character Class

• identifier valid only within bracketed
  character set
• A-Za-z alpha
• digit 0-9 \d
• space \b\f\n\r\t\v \s
• digitalpha_ \dalpha_
  alnum_ \w

13
Advanced RE, Nongreedy Quantifiers

• .?\n making shortest match, otherwise longest
• \n\n .?\n
• lttdgt(lt)lt/tdgt lttdgt(.?)lt/tdgt
• ? Will match zero or one, ?? Will make it match
  only zero

14
Advanced RE, Bound Quantifiers

• m,n at least m, at most n
• m exact m
• m, m or more
• Nongreedy ? can be added after them

15
Advanced RE, Back References

• \1, \2, \3, (count by left parentheses)
• () quoted string (single or
  double quote)
• ().?\1 same as above, but simpler

16
Advanced RE, Look-ahead

• (?) peek without consume
• A.(?\.txt) matching filename before .txt
• A.(?!\.txt) matching filename does not end
  with .txt

17
Advanced RE, Character Codes

• \nn or \mmm 8 bit Octal
• \unnnn 16 bit Unicode
• \uyyyyyyyy 32 bit Unicode
• \xnn 8 bit Hex (Pitfall! Will consume all hexa
  digit truncate to 8 bit Hex)

18
Advanced RE, Newline Sensitive Matching

• lineanchor option makes the and anchors work
  relative to newlines
• With or w/o lineanchor, one can use \A and \Z to
  match the beginning and end of the string
• linestop will make . and match stop at
  newline character, unless \n explicitly
  specified.

19
Advanced RE, Expanded Syntax

• Leading (?x) will make the RE pattern capable of
  embedding comments within RE syntax
• E.g.
• regexp (?x) A pattern to match URLs
• () The protocol before
  initial colon
• //(/) The server name
• ((0-9))? The optional port
  number
• (/.) The trailing pathname
• input

20
Regular Expression Commands

• regexp Regular Expression
• regsub Regular expression Substitution

21
Sample Scripts

• Url_Decode
• proc Url_Decode url
• regsub all \ url url
• regsub all (xdigit2 url \
• format c 0x\1 url
• return subst url

22
References

• http//www.lib.uchicago.edu/keith/tcl-course/topic
  s/regexp.html
• http//wiki.tcl.tk/10842 (Tcl Intro) link to
  http//wiki.tcl.tk/986 (regexp)
• Brent Welch et al, Practical Programming in Tcl
  and Tk, 4th ed., Chapter 11