request for improvement suggestions on existing (non-commercial) code block

hedgehog

I'm looking for suggestions on how i might improve a bit of code i've included.

essentially it adds a random number of tags to random indexes within an array of random words that varies between 1 and 300 indexes (inclusive).

in other words:

i have an array of strings, which are words randomly chosen from a larger list contained in a mysql database. the array index can range anywhere from 1 to 300, but is confined to that max and min.

what i'd like to do is insert html tags into a random number of the array's indexes, picking the indexes, and how many indexes are picked, randomly.

for example if i have an array that contains:

0 = then
1 = when
2 = i
3 = printed
4 = my
5 = it
6 = would
7 = display
8 = like
9 = this
10 = .

and, as the contents state, ideally a print statement
would give me something along the lines of:

"then when
i
printed
out my array, it
would
display like
this
."

(continued in reply)

hedgehog

what i have does work, with a few minor dissapointments. i also have the feeling it's being kind of a server hog. obviously i'd like
to trim the burden down as much as possible.

i've got a database query that generates my array. as you might have guessed, it picks the words at random from a larger list 😛

########################### begin code

$result=mysql_db_query($database, "SELECT word FROM $wordtable ORDER BY rand() LIMIT $limit");
while ($data=mysql_fetch_array($result)) {
$word[] = $data['word'];
}

// next i'll pick a random number of indexes to add the s to.
// the limit var is user-defined, and can range from 1-300.
// since i don't want only 1 or 2 breaks, and i don't want a break
// for each word to occur, i've thrown in some rather lazy math.
// that's probably the first thing i'm unhappy with.

srand((double)microtime()*1000000);
$howmany = (rand($limit%10,$limit%-5));

(continued in reply)

hedgehog

// here i'll make sure i don't have duplicate array indexes in my new
// random picking array

$get = array();
while(count($get)<$howmany) {
$where = (rand(1,$limit-1));
if (!in_array($where, $get)) {
$get[]=$where;
}
}

// here we actually change the array

$getlimit = count($get);

$i=0;
while ($i <= $getlimit) {
$i++;
$g = $get[$i];
$word[$get] = "$word[$get] ";
}

// now we'll print. but here's another problem:
// it won't print the 0 index of $word[],
// so i've got this spotty workaround.

echo "$word[0] ";
$wordlimit = count($word);
$i=0;
while ($i <= $wordlimit) {
$i++;
echo "$word[$i] ";
}

(continued in reply)

hedgehog

while this seems to give me pretty desireable results, i'm nervous that this is putting some horrible strain on the server, since
i'm really more of a designer than programmer. i also wish i could figure out a way to get around the constraints i've mentioned
above, but i seem to be stumped.

any advice for a lowly cheap hack designer type would be greatly, greatly appreciated. 🙂

-a

diego25

echo "$word[0] ";
$wordlimit = count($word);
$i=0;
while ($i <= $wordlimit) {
$i++;
echo "$word[$i] ";
}

can be improved to

$wordlimit = count($word);
for ($i=0; $i<$wordlimit; $i++) {
echo "$word[$i] ";
}

And are you sure your code works? 'cause $get is an array, so you can't do
$word[$get] = "$word[$get] ";

Try
$word[$g] = "$word[$g] ";
instead. Also, the while loop that has that line will also skip the index 0. Move the $i++ after everything in that while loop (but keep it in the while loop - or use a for loop instead)

Diego

hedgehog

indeedily-do, i was recopying some of the code (i tend to choose really outlandish variable names, heh), and accidently threw in $gets where the $gs should have been.

for loops have a speed advantage i take it?

$i++ at the end... doh. i guess that's why it's like that in every while loop i've ever seen 😛

-a

diego25

I don't think for loops are faster, but, for me at least, they are in some cases clearer. The reason I think it's clearer is because you have everything that controls the loop in one line, i.e. one place. In the while loop, you have the initialization outside the loop, the terminating condition at the start of the loop, and the status variable (I forgot its formal name) can be anywhere in the loop.

e.g.

for ($i=0; $i < 100; $i++) {
}

vs.

$i = 0;
while ($i < 100) {
$i++;
}

Diego

hedgehog

yeah, i s'pose this is true. i think i've probably gotten too dependant on using one type of loop for everything.

the thing i think i've become particularly concerned on the more i look at it is how to handle the rand() for $howmany.

the min %10 and the max-5 work okay past 10, but of course once you go below that it errors with greater and greater frequency.

i suppose it doesn't work all that bad with the higher numbers (though towards 300 there's too many line breaks sometimes), so then i started thinking maybe i'd do a series of three ifs, where it would be

if limit <= 25
if limit > 25 && <= 160
if limit > 160 && <= 220
if limit > 220 && <= 300

and having some mins and maxes that are more relative.

this doesn't really feel like the best way to go about this though i'm thinking. i just don't want the number to be too close to 0 or the limit, but it still needs to be relative to the limit.

-a