Parsing and tokenising a css stylesheet

[thestonefox]

Hello all, im new to these boards but i hope someone can help me with this one.

Basically i was wondering if anyone could help me write a function that can parse a css stylesheet document and return an array of the stylesheets

so lets say the sheet is like this

/comment field one/

.style1
{
font-family: arial;
color: white;
}

/comment field two/

.style2
{
font-family: verdana;
color: black;
overflow: hidden;
}

so then the function would basically strip out the comments as i dont want them, so id need some way of saying

everything between this point / and this point / is equal to ""
so it gets rid of the comment

then id need to parse by the . i guess and then parse by ; to get each line, then parse by : to split the styles i guess

so the array would be a multidimensional one and you will get an array like this maybe

array( ('.sheet1','{','font-family:','arial;','color:','white;','}'),
('.sheet2','{','font-family:','verdana;','color:','black;','overflow:','hidden;','}'));

then i can use this array then to populate a database field holding the stored styles

so it could be like

UPDATE $array[1] SET $array[3]=$array[4] etc.....

so perhaps it would be better to just ignore things like { } in the array as theyre not needed by the database

anyway it would be great if someone could help me along with this function so basically you give the function a filename

function importCSS($file)

it parses it and returns a multidimensional array that can then be traversed and create UPDATE queries out of it, or maybe even return the queries in an array, i think that would be better, to return all the created queries in an array, then i could just loop through the array calling the mysql_query command.

also my tables are named as the style name

so i have a table called style1 for instance and each field is every style that you can have

so:

font_size | font_family | color etc......

note that the fields are spaced with and not a - like in a stylesheet, so they would need to replace the - with a in the array

anyway thats what i need help with, i hope im not asking for too much. But id sure appreciate the help.

Kind regards

Harvey

[thestonefox]

Hi all

ive managed to strip the css comments out of the stylesheet

by

$mytext="/ some random comments\n blah blah blah / additional stuff";
print "o:$mytext<br>";
$mytext=preg_replace("/*[\d\D]*?*/", "", $mytext);
$mytext=str_replace("/","",$mytext);
print "n:$mytext<br>";

this will just leave me with additional stuff so additional stuff would be my stylesheets

now what i need to do is strip out everything between

.styleName
{
styles.........
}

and stick them in an array

(style1,color,white,family,arial) etc....
(style2,color,black.........etc.......

i think this preg function may be the key to it, but im finding it difficult to wrap my head around writing these expressions.

Has anyone had experience with these expressions?

[thestonefox]

i have now split each of the tags into its own array now

by pregging between the . and the }

preg_match_all('/.[\d\D]*?}/', $mytext, $matches);

for ($i=0; $i< count($matches[0]); $i++) {
echo "matched $i: " . $matches[0][$i] . "<br><br>";

so now i have

array 1= .style1{color: white; font-family: arial;}
array 2= .style2{color: black; font-family: verdana;}

so now what i need to do is to get those arrays split down even more

to get a sub array of

'.style1','{','color:','white;'.ect........

then i can get the name of the style by

stylename=str_repace(".",""array[0])

ignore array[1] cos i dont want the {

so start a loop at array 2 and traverse through the fields

i have to assume that the stylesheet is in good order

ie like this

color: white;
font-family: arial;

and not

color:
font-family;

because this will mean that the array wont match and wont create a proper insert query

because the array would be like

'.style1','{','color:','font-family'......

so if i was looping the array and said

do while ($c<count($array) || $array[$c]!="}")
{
$c+=2; (to loop over the value to the next style)

//check maybe to see if u can make the query
if (strstr('$array[$c]',':') && strstr('$array[$c+1]',';')
//this would check to see if the to elements have the right endings and also check to see if the next array is also a value and not a tag name

make the query >> SET $array[$c]='$array[$c+1]'
}

although thinking about it if i jump by 2, then it would miss out a tag name again.

hmmmm.......

by the way youre all being extremely helpful (thats sarcasm)

[thestonefox]

dunno why im posting this as no one even bothered to lend a hand to it, bit of crappy community dont you think?

anyway here is the function to parse a css into a useable update query

//bit of test code
$t=parseCss("test.css",1);

for($a=0;$a<count($t);$a++)
{
print $t[$a]."<br><br>";
}
//

function parseCss($file,$id)
{
//open file
$handle = fopen($file, "r");
$mytext="";
while (!feof($handle))
{
$buffer = fgets($handle, 4096);
$mytext.=$buffer;
}

//strip comments
$mytext=preg_replace("/\*[\d\D]*?\*/", "", $mytext); 
$mytext=str_replace("/","",$mytext); 
//end strip comments


//Get styles into an array
preg_match_all('/\.[\d\D]*?}/', $mytext, $matches);

//Get the titles into an array
for ($i=0; $i< count($matches[0]); $i++) {
preg_match_all('/\.[\d\D]*?{/', $matches[0][$i] , $titles[$i]);
}
for ($i=0; $i< count($titles); $i++) {
$titles[$i][0][0]=preg_replace("/,[\d\D]*?{/", "", $titles[$i][0][0]); 

$findthis = array(".", "{");
$titles[$i][0][0]=str_replace($findthis,"",$titles[$i][0][0]);
$mytitles[$i]= $titles[$i][0][0];
}
// end get titles

for ($i=0;$i<count($mytitles);$i++)
	{
	//print "$mytitles[$i]<br>";
	}


// Strip tags to retrieve inner data
for ($i=0; $i< count($matches[0]); $i++) {
$matches[0][$i]=preg_replace("/\.[\d\D]*?{/", "", $matches[0][$i]); 
$matches[0][$i]=str_replace("}","",$matches[0][$i]);
$mydata[$i]=$matches[0][$i];
}
//end strips of inner data

//build the array for query generation
for ($i=0;$i<count($mydata);$i++)
	{
	//if (strlen($mydata[$i])>3)
	//print "$mytitles[$i]<br>$mydata[$i]<br><br>";
	$dpos=0;
	$thedata[$i][$dpos]=$mytitles[$i];
	$tok = strtok($mydata[$i], ";");

	while ($tok) {
	$dpos+=1;
	$thedata[$i][$dpos]=$tok;
	$tok = strtok(";");
	}
	}

//do a lot of shizzle to parse and format the strings to build the query, surweet
for ($i=0;$i<count($thedata);$i++)
	{
	$query[$i]="";
	$query[$i]= "UPDATE ".$thedata[$i][0]." SET ";
	for ($a=1;$a<count($thedata[$i]);$a++)
		{
		$revthis=$thedata[$i][$a];
		$thedatarev=strrev($revthis);
		$thedatatmp=strstr($thedatarev,":");
		$thetemprep=strrev($thedatatmp);
		$thedatatmp1=str_replace("-","_",$thedatatmp);
		$thedatanorm=strrev($thedatatmp1);
		$thedata[$i][$a]=str_replace($thetemprep,$thedatanorm,$thedata[$i][$a]);

		$thedata[$i][$a]=str_replace(":","='",$thedata[$i][$a]);
		$thedata[$i][$a]=str_replace("\t","",$thedata[$i][$a]);
		if ($a<count($thedata[$i])-1)
		$thedata[$i][$a].="'";
		if ($a<count($thedata[$i])-2)
		$thedata[$i][$a].=",";
		$query[$i].= $thedata[$i][$a];
		}
	$query[$i].=" WHERE ID=$id";
	}

return $query;
}

[The_Chancer]

Not so much a crappy community - but a global one. Some people may not be around in the 16 hours from your first post until the last.... (either asleep or working...)

Some posts may take days to get answered - but you have done the right thing and posted your answer.. it may well help another person to get to their solution that much faster.

[thestonefox]

if you look at the part of the code that grabs the variables and values, ull see that i do some string reversing

in retrospect this is stupid, i only did this so i could get the value of the variable and replace the - with a _ but i couldnt use strstr or strrepace on it because some values have a - in them and need the -.

anyway it would have been better to simply explode the string with a : delimeter and then just apply the changes on the 0 array of the explode then stitch the 2 back together for my query string

not too sure if this would be more efficient, i suppose it would be but then i dont really know the speed of doing it either way.

anyway thats just a note

by the way

does anyone know how to parse any given xml form to strip out the form names and values and then display them on the screen or even store them in a database

i know php has built in xml parsing functions but ive tried a few scripts and they dont seem to do anything.

[kkobashi]

Harvey,

Maybe the reason why noone bothered to answer your question is because we just didn't feel like it? We all got jobs or work to do ourselves. What you wanted is a full solution to a problem that wasn't going to be done without much thought and coding.

Ask yourself, how long did it take you to solve your problem. If it took you 3 hours to solve it, do you expect us to take 3 hours out of our lives (maybe even less) to help you? What incentive is there for us to do so?

Was your problem so interesting that someone would want to help you? Can't you see from looking at all the questions posed and answered that there is a community of people here, helping one another?

It's wrong to call this community crappy just because you failed to get a timely response. If you truly think that the world halts for you just because you posted a question, you are looking at this place all wrong... more so if you are a newbie that is going to need future help.

People don't forget aliases you know...

[thestonefox]

shut up