quote:

---

Originally posted by Charles Capps:
Different languages have different ideas about operator precedence.

---

Well, that's brilliant! Probably another example of IT people ensuring their job security by making certain that no one can figure out what they're doing.

quote:

---

Knowing how often PHP breaks backwards compatability, it wouldn't surprise me if it changed from version to version.

---

quote:

---

Using parens is the best way to forcefully ensure that your idea of precedence is enforced.

---

code:

---

$dbh = mysql_connect("localhost", "user", "pass");
if(!$dbh) {
	echo "Could not connect to database: " . mysql_error();
	exit;
} // end if

// $dbh == our database handle
mysql_select_db("mydb", $dbh);

// Make a query...
$result = mysql_query("select * from table where col_1 = 3", $dbh);
if(!$result) {
	echo "Could not perform query: " . mysql_error();
	exit;
} // end if

// ... then pull the info out of the DB, line by line, into an array
$results = Array();
$i = 0;
while($resline = mysql_fetch_array($result)) {
	$results[$i] = $resline;
	$i++;
} // end while

// We now have our results array.  Loop through it and do stuff.
// (using the $i index from above as our array length rather than
// sizeof($results), just because I'm weird)
for($j = 0; $j < $i; $j++) {
	$returned_val = do_something_smart_with_row($results[$j]);
	$sql = "update table set col_2 = '$returned_val' "
		. where col_1 = '" . $results[$j]["col_1"] . "'";
	$thisres = mysql_query($sql);
	if(!$thisres) {
		echo "Could not perform query: " . mysql_error();
		exit;
	} // end if
} // end for

mysql_close($dbh);

---

code:

---

use DBI;
our $dbh = DBI->connect("DBD:mysql:mydb", "user", "pass")
	or die "Could not connect to database: $DBI::errstr";

# $dbh is the database handle

# Query from the DB and get all the results in a single hash
my $results = $dbh->selectall_hashref("select * from table where col_1 = 3")
	or die "Could not perform query: $DBI::errstr";

# Loop through the hash and do stuff
foreach my $key (keys %{$results}) {
	my $returned_val = &do_something_smart_with_row($results->{$key});
	my $sql = "update table set col_2 = '?' where col_1 = '?'"
	# the ?s get replaced with the 2nd+ arguments in the
	# do() method of the database handle...
	$dbh->do($sql, undef, $returned_val, $key) 
		or die "Could not perform query: $DBI::errstr";
} // end foreach

$dbh->disconnect();

---

The main difference between the code, other than the verbosity, is that all the database functions in the PHP script are database-specific, and all the database functions in Perl are generic. While PHP does have a database abstraction layer, it's not built into the language. (Then again, neither is Perl's DBI... but considering that DBI is all there is to connect to a DB in the first place, that's no big loss.)

Thus, moving the Perl script to another database system later is just a matter of making sure the SQL works, while every single command would need to be translated in the PHP script. This annoyed a coworker so much that, when he transitioned to PHP from Perl, he created a PHP object that duplicated the parts of DBI that he used.

Perl also has a number of other advantages in the convenience departments... for instance, having to escape quotes in prints/echos is a major annoyance in PHP... but in Perl, you can just define another quote character for that one string. I.e.:

$php = "There \" are \" too\"\" many" . '\'quotes\' in' . "\" this\" string";
$perl = qq~There " are " too"" many 'quotes' in" this " string~;

PHP does have its advantages.. often, hosts will compile everything and the kitchen sink into PHP, meaning often it can do graphics, PDF generation, and other nifty things "right out of the box"... Perl can do all of the same things, but hosts often don't compile the modules in by default... and compiling modules would require shell access.

PHP also often has a major speed advantage. Once compiled and running, PHP and Perl scripts are pretty much the same speed... but the startup/compile phase is a killer. PHP is often compiled right into the web server, meaning it practically stays resident in memory. PHP also has the ability to keep cached compiled copies of scripts in memory, which can be a major speedup.

Doing the same things in Perl requires the mod_perl Apache module, which isn't something many shared hosting providers offer. (Writing code for mod_perl is also rather complex. You can't just plug your script in and expect it to work without modification.. sometimes as little as two or three lines, thankfully)

Anyway. I've ranted enough. Personally, I prefer Perl because of the consistancy and the flexability of use and syntax. I don't care for PHP because it gets horrible inconsistant, forces the authors to be very verbose (huge function names) and has a number of forced syntax quirks.
 

---

Posted by Sol System (Member # 30) on :
 
Computer Science 110: Dropped.

No.

But, man. I do not have such strong or informed opinions on anything. I live a shadow life.
 

---

Posted by Jason Abbadon (Member # 882) on :
 
Computer....code?
Like Bible Code?

Can you predict a thread's future by decrypting this code?

This sucks: I forgot my decoder ring.

Stop me before my wise-ass refrences drift into codes of conduct like Bushido....
 

---

Posted by Harry (Member # 265) on :
 

quote:

---

Hell, even Java would be better to learn first.

---