? Parsing strings to arrays ?

glerma

I need some help with the following:

Need to find out how I can parse a long string (up to 1024 k) that has the following characteristics.

{tag1 "value", tag2 "value", tag3 "value"}

Having so much difficulty trying to get the tag "value" pair to match each other. I don't know how to parse things between two double-quotes. Would prefer convert the string to something like:

(tag1=>value,tag2=>value,tag3=>value)

This would allow me to use the powerful PHP built-in array functions. Much more flexible.

Anyone have any suggestions?

Thanks in advance.

Weedpacket

Something similar came up recently, with someone wanting to convert a big block of HTML - but only the bits between tags, not inside the tags themselves. A search should turn it up

In this case it's even easier, however so I'm going to make it sound tedious. One possible complication: can " appear inside a double-quoted value? If so, how is it prevented from choking things up?

First off, trim() the string so's I don't have to worry about leading or trailing whitespace. Then use substr($bigstring,1,-1) to knock off the {} characters (I'm presuming that there's only one of these things in the entire string.)

If it can't appear the problem is pretty simple. $bits = explode('"', $bigstring);
All the odd-numbered elements are tag names and sundry punctuation (, and occasional whitespace) and all the even-numbered elements are the values.

If " can appear, and it's escaped with (say) [/b], then $bits = preg_split('/(?<!\)/"', $bigstring); instead. (Note that \ is escaped because PHP uses it as an escape character, too. If the string uses something else, then it will have to be converted before turning it into a PHP-array-like thing.);

Strip the sundry punctuation from the odd-numbered bits and you'll have an array going tag,value,tag,value,...

I think there may be an empty "tag" right on the end - easy enough to check - there is if the array has an odd number of elements. array_pop and throw it away if it's there.

Now put " characters around all the elements (yes, you'll want to do this if it's to be a PHP-style array).

Put => on the end of all the odd-numbered elements, and , on the even-numbered ones.

Now $bigstring = join('', $bits);
If yo're worried about the stray comma on the very end of the resulting string, substr() it off, but PHP won't mind if it's there.

And to wrap up, $bigstring = '('.$bigstring.')';

glerma

RE: Now put " characters around all the elements .

Not quite sure how to do that. Tried sr_pad(), but no dice.

Weedpacket

Same way I put () around $bigstring at the end 🙂

foreach($bits as $key=$item)
  $bits[$key]='"'.$item.'"';

glerma

Never mind. I figured out a way. Here is custom padding function for anyone looking for something similar:

/ add quotes to each element in an array /

function add_quotes(&$value, $key){
$value = trim("$value"); // Strip whitespace.
$len = strlen($value); // Capture string lenghth.
$len = $len+2; // Add two to total for padding.
print str_pad($value,$len,"\"",STR_PAD_BOTH); //Pad both ends with double quotes.
}

just apply the above function to array_walk like this:

array_walk($array,'add_quotes');

glerma

Didn't think of that one! Thanks! 😃

glerma

Well. I think I'm almost there. I'm still having one last problem though, which is the most important for me.

For some reason, I cannot build my final array based on the string variable "$bigstring".

Consider the following:
$bigstring = " test1=>'01',test2=>'02',test3=>'03' ";

echo $bigstring;

outputs .... test1=>'01',test2=>'02',test3=>'03'

(should be correct syntax for arrays)

But if I try to do :
$myArray = array($bigstring);

print_r($myArray);

outputs... Array([0] => test1=>'01',test2=>'02',test3=>'03')

(not what I'm looking for)

If I do:
$myArray (test1=>'01',test2=>'02',test3=>'03');

will output... Array([test1] =>01 [test2] => 02 [test3] => 03)

This last one is what I'm looking for.

I'm a bit stumped as to why the variable string won't get passed correctly into the array.

Weedpacket

That's because, as far as PHP is concerned, $bigstring is just a string - it has no reason to treat it as an array. (A good thing it doesn't do this, of course: otherwise the string "2002-11-29" would always end up as "1962"!)

To explicitly tell PHP to evaluate a string as a piece of PHP code, use the eval() function. Remember that the contents of the eval() function call has to be syntactically valid PHP code.

$bigstring = "array('key1'=>'value1','key2'=>'value2','key3'=>'value3',)";
$myarray = eval("return $bigstring;");
var_dump($myarray);