PHP RegEx and parts of string

Question

I have strings, for example:

bool dxDrawImage float posX, float posY, float width, float height, mixed image , float rotation = 0, float rotationCenterOffsetX = 0, float rotationCenterOffsetY = 0, int color = white, bool postGUI = false
bool dxUpdateScreenSource element screenSource, bool resampleNow =false

I need get parts from it, for example:

bool dxDrawImage
float posX
float posY
...

I wrote:

preg_match_all("/(bool|float|int)[a-zA-Z0-9\s=]+/", $txt, $ar);

print_r($ar):

Array ( [0] => Array ( [0] => bool dxDrawImage float posX [1] => float posY [2] => float width [3] => float height [4] => float rotation = 0 [5] => float rotationCenterOffsetX = 0 [6] => float rotationCenterOffsetY = 0 [7] => int color = white [8] => bool postGUI = false ) [1] => Array ( [0] => bool [1] => float [2] => float [3] => float [4] => float [5] => float [6] => float [7] => int [8] => bool ) )

Why this regexp capture bool dxDrawImage float posX instead

bool dxDrawImage
float posX

How to fix this?

hmm, is it ok that ther no comma between bool dxDrawImage and float posX in input string? — RiaD
– RiaD, Commented Jul 21, 2012 at 13:05
Oh, of course. I forgot that there is separator between other parts of string... Thank you! — user0103
– user0103, Commented Jul 21, 2012 at 13:10

Ωmega · Accepted Answer · 2012-07-21 13:15:25Z

1

preg_match_all("/(\b(?:bool|float|int|element)\s+.*?)(?=\b(?:bool|float|int|element|[,;\n\r]|$))/", $txt, $ar); 
print_r($ar[1]);

edited Jul 21, 2012 at 13:15

answered Jul 21, 2012 at 13:09

Ωmega

44k35 gold badges143 silver badges213 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

RiaD · Accepted Answer · 2012-07-21 13:10:07Z

1

You may disallow whitespaces in variable name:

preg_match_all("/(bool|float|int)\s+[a-zA-Z0-9=]+/", $txt, $ar);

What if you will use:

preg_match_all("/(bool|float|int)\s+[a-zA-Z0-9]+(\s+=\s+[chars,possible for value]+)?/", $txt, $ar);

edited Jul 21, 2012 at 13:10

answered Jul 21, 2012 at 13:00

RiaD

47.8k12 gold badges85 silver badges128 bronze badges

2 Comments

user0103 Over a year ago

But, this float rotationCenterOffsetX = 0 wouldn't be captured.

user0103 Over a year ago

Although I can replace this by making, for example, float rotationCenterOffsetX=0, but I need spaces between equals sign.

Bali Balo · Accepted Answer · 2012-07-21 13:02:14Z

0

Maybe you can use the ungreedy flag.

preg_match_all("/(bool|float|int)[a-zA-Z0-9\s=]+/U", $txt, $ar);

answered Jul 21, 2012 at 13:02

Bali Balo

3,4181 gold badge20 silver badges37 bronze badges

2 Comments

RiaD Over a year ago

AFAIK, It will match only bool, float or int

user0103 Over a year ago

Warning: preg_match_all() expects parameter 4 to be long, string given

hakre · Accepted Answer · 2012-07-21 15:59:39Z

The part that delimits one variable from the other is a bit unclear in your string. That makes it a bit hard to separate each from another. However, you can relatively simple still write the pattern as it occurs:

((?:bool|float|mixed|int|element) [a-zA-Z]+(?: = ?[a-z0-9]+)?)(?:\s?,\s?|\s|$)
|^-------- variable type -------^     |    ^--- optional ---^|^-- delimiter -^
|                                   name                     |
|                                                            |
`------------------ the interesting part --------------------´

Such a pattern works with preg_match_all:

$pattern = '(((?:bool|float|mixed|int|element) [a-zA-Z]+(?: = ?[a-z0-9]+)?)(?:\s?,\s?|\s|$))';
$variables = array();
if ($r = preg_match_all($pattern, $subject, $matches)) {
    $variables = $matches[1];
}
print_r($variables);

Output:

Array
(
    [0] => bool dxDrawImage
    [1] => float posX
    [2] => float posY
    [3] => float width
    [4] => float height
    [5] => mixed image
    [6] => float rotation = 0
    [7] => float rotationCenterOffsetX = 0
    [8] => float rotationCenterOffsetY = 0
    [9] => int color = white
    [10] => bool postGUI = false
    [11] => bool dxUpdateScreenSource
    [12] => element screenSource
    [13] => bool resampleNow =false
)

The other alternative is to only use the delimiter and split the string. However after the delimiter a variable type plus space follows (look-ahead):

$pattern = '((?:\s?,\s?|\s)(?=(?:bool|float|mixed|int|element)))';
$variables = preg_split($pattern, $subject);
print_r($variables);

It gives you the same result as above.

Collectives™ on Stack Overflow

PHP RegEx and parts of string

4 Answers 4

Comments

2 Comments

2 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

Comments

2 Comments

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related