COMP284 Scripting Languages
Lecture 6: PHP (Part 4)
Handouts
Ullrich Hustadt
Department of Computer Science
School of Electrical Engineering, Electronics, and Computer Science
University of Liverpool
Regular Expressions Overview
PHP and regular expressions
•
Input validation is an important step in preventing incorrect or
dangerous data entering an application
•
Regular expression matching is a useful technique for input validation
•
Regular expressions are also useful for
•
data extraction
•
data clearning and data transformation
•
Over time, PHP has supported different variants of regular expressions
•
POSIX-extended regular expressions
; deprecated in PHP 5.3.0, removed in PHP 7.0.0
•
Shell-style wildcard patterns
; still supported, but not particularly powerful
•
Perl-compatible regular expressions (PCRE)
; closely resemble Perl regular expressions
COMP284 Scripting Languages Lecture 6 Slide L6 – 2
Regular Expressions PCRE
Perl-compatible regular expressions
•
In PHP, a regular expression is a string the content of which resembles a
Perl regular expression
' /(\ d +\.\ d +) seconds / ' " /[ bc ][b -e ][^ bcd ]/"
•
The character / acts as a delimiter indicating the start and end of the
regular expression
; carried over from Perl
; ‘closing’ delimiter can be followed by modifiers
•
In PHP any non-alphanumeric, non-backslash, non-whitespace character
can be used as delimiter
•
If the delimiter also occurs inside the pattern it must be escaped using
a backslash \
•
Backslash behaviour and differences between single- and double-quoted
strings carry over
COMP284 Scripting Languages Lecture 6 Slide L6 – 3
Regular Expressions Characters
Regular expressions: Characters
The simplest regular expression just consists of a sequence of
•
alphanumberic characters and
•
non-alphanumeric characters escaped by a backslash
that matches exactly this sequence of characters occurring as a substring
in the target string
"/cbc/"
matches "ababcbcdcde"
'/1\+2/'
matches "21+23=44"
•
Strictly speaking, not all non-alphanumeric characters need to be
escaped by a backslash, only those that are ‘reserved’
COMP284 Scripting Languages Lecture 6 Slide L6 – 4
Regular Expressions Escape sequence and meta-characters
Regular expressions: Escape sequences and
meta-characters
There are various meta-characters and escape sequences that match
characters other than themselves:
. Matches any character except \n
\w Matches a ‘word’ character (alphanumeric plus ‘_’)
\W Matches a non-‘word’ character
\s Matches a whitespace character
\S Matches a non-whitespace character
\d Matches a decimal digit character
\D Matches a non-digit character
"/\w.\d/"
matches "-N$1.00"
'/\W.\D/'
matches "-N$1.00"
COMP284 Scripting Languages Lecture 6 Slide L6 – 5
Regular Expressions Character classes
Regular expressions: Character class
•
A character class, a list of characters and escape sequences enclosed in
square brackets, matches any single character from within the class,
for example, [ad\t\n\-\\09]
•
One may specify a range of characters with a hyphen - ,
for example, [b-u]
•
A caret ^ at the start of a character class negates/complements it,
that is, it matches any single character that is not from within the class,
for example, [^01a-z]
"/[bc][b-e][^bcd]/"
matches "abacdecdc"
'/[$£][1-9]\.[0-9]/'
matches "$2.1"
COMP284 Scripting Languages Lecture 6 Slide L6 – 6
Regular Expressions Anchors
Regular expressions: Anchors / assertions
Anchors (assertions) allow us to fix where a match has to start or end
\A Match only at string start
^ Match only at string start (default)
Match only at a line start (in multi-line matching)
\Z Match only at string end modulo a preceding \n
\z Match only at string end
$ Match only at string end modulo a preceding \n
Match only at a line end (in multi-line matching)
\b Match word boundary (between \w and \W)
\B Match except at word boundary
Targ et stri ng " one two \ n three four \ nfive \n"
Do match Do not match
'/\Aone/' (at string start) '/^three/' (not a string start)
'/^three/m' (at line start) '/four\Z/' (not at string end mod \n)
'/four$/m' (at line end) '/five\z/' (not a string end)
'/five\Z/' (at string end mod \n)
'/\bone\b/' (a word in string) '/\bour\b/' (not a word in string)
COMP284 Scripting Languages Lecture 6 Slide L6 – 7
Regular Expressions Quantifiers
Quantifiers
•
The constructs for regular expressions that we have seen so far are not
sufficient to match, for example, natural numbers of arbitrary size
•
Also, writing a regular expressions for, say, a nine digit number
would be tedious
This is made possible with the use of quantifiers
regexpr* (Greedily) Match regexpr 0 or more times
regexpr+ (Greedily) Match regexpr 1 or more times
regexpr? (Greedily) Match regexpr 1 or 0 times
regexpr{n} Match regexpr exactly n times
regexpr{n,} (Greedily) Match regexpr at least n times
regexpr{n,m} (Greedily) Match regexpr between n and m times
Quantifiers are greedy by default and match the longest leftmost sequence
of characters possible
COMP284 Scripting Languages Lecture 6 Slide L6 – 8
Regular Expressions Quantifiers
Quantifiers
regexpr* (Greedily) Match regexpr 0 or more times
regexpr+ (Greedily) Match regexpr 1 or more times
regexpr? (Greedily) Match regexpr 1 or 0 times
regexpr{n} Match regexpr exactly n times
regexpr{n,} (Greedily) Match regexpr at least n times
regexpr{n,m} (Greedily) Match regexpr between n and m times
"/\d+\.\d+/"
matches "Delay 10.486 sec"
'/\d+/'
matches "No 12 345 6789"
'/[A-Z]0{2}\d{6}/'
matches "ID E004813709"
'/\d+/'
matches "ID E004813709"
•
The examples on the right illustrate that a regular expressions \d+
•
matches as early as possible
•
matches as many digits as possible
COMP284 Scripting Languages Lecture 6 Slide L6 – 9
Regular Expressions Quantifiers
Quantifiers
•
There are situations where greedy matching is not appropriate,
for example, matching a single comment in a program
'=/\*.*\*/='
matches "x = /* one */ y*z; /* three */"
•
By adding ? to a quantifier, we can make it match lazily, that is,
it will match the shortest leftmost sequence of characters possible
'=/\*.*?\*/='
matches "x = /* one */ y*z; /* three */"
COMP284 Scripting Languages Lecture 6 Slide L6 – 10
Regular Expressions Quantifiers
(Lazy) Quantifiers
regexpr*? Lazily match regexpr 0 or more times
regexpr+? Lazily match regexpr 1 or more times
regexpr?? Lazily match regexpr 1 or 0 times
regexpr{n,}? Lazily match regexpr at least n times
regexpr{n,m}? Lazily match regexpr between n and m times
"/\d+?\.\d+?/"
matches "Delay 10.486 sec"
'/\d+?/'
matches "No 12 345 6789"
'/[A-Z]0{2}\d{6}/'
matches "ID E004813709"
'/\d+?/'
matches "ID E004813709"
COMP284 Scripting Languages Lecture 6 Slide L6 – 11
Regular Expressions Capturing subpatterns
Regular expressions: Capturing subpatterns and
backreferences
•
We often encounter situations where we want to identify the repetition
of the same or similar text, for example, in HTML markup:
<strong > ... </ strong >
<h1 > ... </h1 >
•
We might also not just be interested in the repeating text itself,
but the text between or outside the repetition
•
We can characterise each individual example above
using regular expressions:
'= < strong >.*? </ strong >= '
'= <h1 >.*? </ h1 >= '
but we cannot characterise both without losing fidelity, for example:
' = <\ w + >.*? </\ w+ >= '
does not capture the ‘pairing’ of HTML tags
COMP284 Scripting Languages Lecture 6 Slide L6 – 12
Regular Expressions Capturing subpatterns
Regular expressions: Capturing subpatterns and
backreferences
The solution are capturing subpatterns and backreferences
(regexpr) creates a capturing subpattern
(opening parentheses are counted starting from 1
to give an index number)
(?<name>regexpr) creates a named capturing subpattern
(?:regexpr) creates a non-capturing subpattern
\N, \gN, \g{N} backreference to a capturing subpattern N
(where N is a natural number)
\g{name} backreference to a named capture group
"=<(\w+)>.*?</\1>=" "=<(?<c1>\w+)>.*?</\g{c1}>="
matches "<li><b>item</b></li>" matches "<li><b>item</b></li>"
"=<(\w+)>.*?</\1>="
do es not match "<li>item</em>"
COMP284 Scripting Languages Lecture 6 Slide L6 – 13
Regular Expressions Alternations
Regular expressions: Alternations
•
The regular expression regexpr1|regexpr2 matches
if either regexpr1 or regexpr2 matches
This type of regular expression is called an alternation
•
Within a larger regular expression we need to enclose alternations
in a subpattern or capturing subpattern
(regexpr1|regexpr2) or (?:regexpr1|regexpr2)
to indicate where regexpr1 and regexpr2 start/end
'/(Mr|Ms|Mrs|Dr).*?([\w\-]+)/'
matches "Dr Michele Zito"
matches "Mr Dave Shield"
matches "Mrs Judith Birtall"
COMP284 Scripting Languages Lecture 6 Slide L6 – 14
Regular Expressions Alternations
Regular expressions: Alternations
The order of expressions in an alternation only matters
if one expression matches a sub-expression of another
'/cat|dog|bird/'
matches "cats and dogs"
'/dog|cat|bird/'
matches "cats and dogs"
s
0
s
1
s
2
s
3
s
d
s
c
s
b
s
o
s
a
s
i
s
g
s
t
s
r
s
2
d
d o
g
c a t
b i r t
'/dog|dogs/'
matches "cats and dogs"
s
0
s
1
s
2
s
1
d
s
2
d
s
1
o
s
2
o
s
1
g
s
2
g
s
2
s
d
d
o
g
o
g
s
'/dogs|dog/'
matches "cats and dogs"
s
0
s
1
s
2
s
1
d
s
2
d
s
1
o
s
2
o
s
1
g
s
2
g
s
1
s
d
d
o
g
o
g
s
COMP284 Scripting Languages Lecture 6 Slide L6 – 15
Regular Expressions Modifiers
Regular expressions: Modifiers
Modifiers change the interpretation of certain characters in a regular
expression or the way in which PHP finds a match for a regular expression
/ / Default
‘.’ matches any character except ‘\n’
‘^’ matches only at string start
‘$’ matches only at string end modulo preceding \n
/ /i perform a case-insensitive match
Target string "Hillary\nClinton"
/clinton/ do es not match
/clinton/i do es match
COMP284 Scripting Languages Lecture 6 Slide L6 – 16
Regular Expressions Modifiers
Regular expressions: Modifiers
Modifiers change the interpretation of certain characters in a regular
expression or the way in which PHP finds a match for a regular expression
/ /s Treat string as a single long line
‘.’ matches any character including ‘\n’
‘^’ matches only at string start
‘$’ matches only at string end modulo preceding \n
/ /m Treat string as a set of multiple lines
‘.’ matches any character except ‘\n’
‘^’ matches at a line start
‘$’ matches at a line end
Target string "Hillary\nClinton"
/(bill|hillary).clinton/mi does not match
/(bill|hillary).clinton/si does match
/(bill|hillary).^clinton/mi does not match
/(bill|hillary).^clinton/si does not match
COMP284 Scripting Languages Lecture 6 Slide L6 – 17
Regular Expressions Modifiers
Regular expressions: Modifiers
Modifiers change the interpretation of certain characters in a regular
expression or the way in which Perl finds a match for a regular expression
/ /sm Treat string as a single long line, but detect multiple lines
‘.’ matches any character including ‘\n’
‘^’ matches at a line start
‘$’ matches at a line end
Target string "Hillary\nClinton"
/(bill|hillary).^clinton/smi do es match
COMP284 Scripting Languages Lecture 6 Slide L6 – 18
PCRE functions
PCRE functions: preg_match (1)
int preg_match(rx, str [,&$matches [,flags [,offset]]])
•
Attempts to match the regular expression rx against the string str
starting at offset
•
Returns 1 if there is match; 0 if there is not; FALSE in case of error
•
$matches is an array containing at index 0 the full match and at the
remaining indices the matches for any capture groups
•
flags modify the behaviour of the function
$txt = " Yabba dabba doo ";
if ( p r e g _ m a t c h ( ' /^[a -z\s ]* $ /i ', $ txt )) {
echo " ' $ txt ' only consis t s of letters and spaces "
}
' Yabba dabba doo ' only cons i sts of lette rs and spaces
COMP284 Scripting Languages Lecture 6 Slide L6 – 19
PCRE functions
PCRE functions: preg_match (2)
int preg_match(rx, str [,&$matches [,flags [,offset]]])
•
Attempts to match the regular expression rx against the string str
starting at offset
•
Returns 1 if there is match; 0 if there is not; FALSE in case of error
•
$matches is an array containing at index 0 the full match and at the
remaining indices the matches for any capture groups
•
flags modify the behaviour of the function
$t = " Yabba dadda doo ";
if ( p r e g _ m a t c h ( ' /((? < c1 >\ w)(? < c2 >\w )\ g{ c2 }\ g{ c1 })/ ',$t , $m )){
foreach ( $m as $ key => $ value ) {
prin tf ( " % -2s => % s \n" ,$key , $ value );
} }
0 => abba
1 => abba
c1 = > a c2 = > b
2 => a 3 => b
COMP284 Scripting Languages Lecture 6 Slide L6 – 20
PCRE functions
PCRE functions: preg_match_all
int preg_match_all(rx, str [,&$matches [,flags [,offs]]])
•
Retrieves all matches of the regular expression rx against the string
str starting at offs
•
Returns the number of matches; FALSE in case of error
•
$matches is a multi-dimensional array containing the matches found
•
flags modify the behaviour of the function
$txt = " Yabba dadda doo ";
if ( p reg_match_all ( ' /((? < c1 >\ w )(? <c2 >\w )\ g{ c2 }\ g{ c1 })/ ',
$txt , $m , PREG_S E T _ O R DER )){
print_r ( $m ); }
Arr ay ( [0] =>
Array (
[0] = > abba
[1] = > abba
[ c1 ] => a [c2 ] => b
[2] = > a [3] => b )
[1] = >
Arr ay (
[0] = > adda
[1] = > adda
[ c1 ] => a [c2 ] => d
[2] = > a [3] => d ))
COMP284 Scripting Languages Lecture 6 Slide L6 – 21
PCRE functions
PCRE functions: preg_replace
mixed preg_replace(rx, rpl, str [, lmt [, &$num]])
•
Returns the result of replacing matches of rx in str by rpl
•
lmt specifies the maximum number of replacements
•
On completion, $num contains the number of replacements
performed
•
rpl can refer back to capturing subpatterns in rx via $N where N is
a natural number
$old = " Dr Ul l rich Hustadt ";
$new = preg_replace ( ' /( Mr |Ms| Mrs | Dr )?\ s *(\ w +)\ s +(\ w +)/ ' ,
'$3 , $2 ' , $old );
echo $new ;
Hustadt , Ullrich
COMP284 Scripting Languages Lecture 6 Slide L6 – 22
PCRE functions
PCRE functions: preg_replace
mixed preg_replace_callback(rx,fun,str [,lmt [,&$num]])
•
Returns the result of replacing matches of rx in str by the return
values of the application of fun to each match
•
lmt specifies the maximum number of replacements
•
On completion, $num contains the number of replacements
performed
$old = " 105 degre es F a h r e n h eit is quite warm ";
$new = pre g _ replace_c a l lback ( ' /(\ d +) deg rees Fahrenheit / ' ,
functio n ( $matc h ) {
retu rn round (( $ match [1] - 32) * 5 / 9) .
" degrees Cel cius " ;
}, $old );
echo $new ;
41 degrees Celcius is quite warm
COMP284 Scripting Languages Lecture 6 Slide L6 – 23
PCRE functions
PCRE functions: preg_split
array preg_split(rx, str [,lmt [,flags]])
•
Splits str by the regular expression rx and returns the result as array
•
lmt specifies the maximum number of split components
•
flags modify the behaviour of the function
# E s t ablish the freque n c y of words in a s tring
$string = " peter paul mary paul jim mary paul " ;
$cou nt = [];
forea ch ( preg_split ( '/\ s+/ ', $s tring ) as $ word )
if ( array_key_ex i s t s ( $ word , $ coun t ))
$cou nt [ $ word ]++;
else
$cou nt [ $ word ] = 1;
forea ch ( $count as $key => $ value )
print (" $key = > $value ; ");
pet er = > 1; paul = > 3; mary = > 2; jim => 1;
COMP284 Scripting Languages Lecture 6 Slide L6 – 24
PCRE functions
Revision
•
Read
•
R. Cox: Regular Expression Matching Can Be Simple And Fast (but is slow
in Java, Perl, PHP, Python, Ruby, . . . ). swtchboard, Jan 2007.
https://swtch.com/
~
rsc/regexp/regexp1.html [accessed 28 Dec 2019]
•
Read
•
PHP Manual: Regular Expressions (Perl-Compatible)
http://uk.php.net/manual/en/book.pcre.php
of P. Cowburn (ed.): PHP Manual. The PHP Group, 24 Dec 2019.
http://uk.php.net/manual/en [accessed 29 Dec 2019]
COMP284 Scripting Languages Lecture 6 Slide L6 – 25
