RE: [PHP] Odd Strpos Behavior
| From: | John W. Holmes | Date: | Sat, 14 Dec 2002 05:50:32 +0000 |
| Subject: | RE: [PHP] Odd Strpos Behavior | ||
| References: | 1 | Groups: | php.general |
| Request: | Send a blank email to php-general+get-128256@lists.php.net to get a copy of this message | ||
And here's a good example of why you should always test each solution
and time it to see what's better. I was recommending a
preg_replace_callback solution which I thought, as a lot of other people
would also think, is a lot faster that your own "method" of doing it.
So, I wrote the preg_replace_callback method and timed it against your
method using the following input and replacement strings:
//test string
$string = "This is a [test] of something [that] will hopefully [work].";
//replacement strings
$test = "test_one, test_two [foo], test_three";
$that = "that_one, that_two, that_three [foo]";
$work = "work_one [foo], work_two, work_three";
$foo = "foo_one [test], foo_two [that], foo_three [work]";
That's actually some complex replacement because [foo] can be replaced
with a replacement that refers back to [test] which refers again to
[foo], etc... So actually each run will result in a different string
entirely and would take different amounts of times based on what was
randomly chosen as a replacement. So, I ran each method 1000 times to
kind of equal out the randomness...
Anyway, here are the results. The test file I used is also attached. You
may be able to tweak it further, I don't know. But, since you're opening
and reading files, the small difference between the two may be negated
by having to deal with files... always test your solutions.
Original String: This is a [test] of something [that] will hopefully
[work].
Replaced String: This is a test_two foo_one test_two foo_two that_one of
something that_two will hopefully work_three.
Preg_replace_callback Time: 0.226889014244
------------------------------------------------------------------------
----
Original String: This is a [test] of something [that] will hopefully
[work].
Replaced String: This is a test_two foo_one test_three of something
that_two will hopefully work_two.
String-parser Time: 0.187005996704
(The "replaced string" shown is just the last result of running it 1000
times)
---John W. Holmes...
PHP Architect - A monthly magazine for PHP Professionals. Get your copy
today. http://www.phparch.com/
> -----Original Message-----
> From: Steve Keller [mailto:skeller@healthtvchannel.org]
> Sent: Thursday, December 12, 2002 5:09 PM
> To: php-general@lists.php.net
> Subject: RE: [PHP] Odd Strpos Behavior
>
> At 12/11/2002 08:09 PM, you wrote:
>
> >Okay, so how do you know what to replace something like [author]
with?
> >What exactly are you doing again? I've forgotten the original
question.
> >:)
>
> Ok, got a sentence, like:
>
> a pile of [metal] 600 feet wide and 30 feet tall. On top of
it is
> a [monster].
>
> The items in the [] represent the names of files containing lists of
> possible results. So I need to grab the first one, "metal," open up
> "metal.php," grab a random item from it, such as "gold," and replace
> [metal] in the original sentence with the result. I should now have:
>
> a pile of gold 600 feet wide and 30 feet tall. On top of it
is a
> [monster].
>
> Now, what I was doing before was strpos'ing the []'s and grabbing what
as
> in between to be the file name, then repeating in a loop until I've
> eliminated all of the [] items. The reason I'm doing it in a loop is
> because the results of the "metal.php" random item may include their
own
> []
> items, which also need to be run exactly the same way. A result might
be
> "gold [objects]" and then I have to run through the objects.php file
and
> get a result from that.
>
> This is what I started with:
>
> <?
> $a = "[adjective] [beginning]"; // temporary item for testing
>
> $e = strpos($a,"]");
> while ($e) {
> echo "A: ".$a."<br />";
> echo "E: ".$e."<br />";
> $f = strpos($a, "[");
> echo "F: ".$f."<br />";
> $tmp = substr($a, 0, $f);
> $table=substr($a, $f+1, $e-1);
> echo "Table: ".$table."<br />";
> $a = substr($a, $e+1, strlen($a));
> $dataFile = $table.".php";
> //$b = getFileElement($dataFile);
> $tmp .= $b;
> $tmp .= $a;
> $a = $tmp;
> $e = strpos($a,"]");
> }
> echo $a;
> ?>
>
> That should work just fine. I cut out everything up to the first [ and
add
> it to $tmp. Then I get the next sequence of characters up to the first
]
> and use it as a file name for the "getFileElement" function. Add the
> result
> to $tmp, add what was left after the first ], and viola.
>
> My problem is that, on the first loop through, strpos returns exactly
> where
> the first ] is, so I can chop up to that no problem. However, the
second
> time through the loop, it's one off, which breaks the logic of the
loop, I
> end up with "beginning]" as my file name.
> --
> S. Keller
> UI Engineer
> The Health TV Channel, Inc.
> (a non - profit organization)
> 3820 Lake Otis Pkwy.
> Anchorage, AK 99508
> 907.770.6200 ext.220
> 907.336.6205 (fax)
> Email: skeller@healthtvchannel.org
> Web: www.healthtvchannel.org
>
>
> --
> PHP General Mailing List (http://www.php.net/)
> To unsubscribe, visit: http://www.php.net/unsub.php