Ordering of taint option -T

pbeckingham has asked for the wisdom of the Perl Monks concerning the following question:

My code failed code review because of the claim that when the -T taint flag is used, it must always be the first flag on the shebang line. I did not know this, or in other words, I don't believe this, because I cannot find any supporting documentation.

Does anyone have any reasons why I would either accept this requirement, or challenge it?

Update: I am also required to remove the -w option from the shebang line for production code. Is this also a good idea?

Comment on Ordering of taint option -T Select or Download Code

Replies are listed 'Best First'.
Re: Ordering of taint option -T by jeffa (Bishop) on Jun 07, 2004 at 19:37 UTC
According to perlrun: For security reasons, this option must be seen by Perl quite early; usually this means it must appear early on the command line or in the #! line ... ~~And that makes sense, since the switches are parsed first come first serve.~~ Also, many places require warnings to be "turned off" for production code, as using them does require a few more CPU cycles -- some places, however, turn them off because of plethora of warnings being sent out to their error logs is overwhelming. :P jeffa L-LL-L--L-LL-L--L-LL-L-- -R--R-RR-R--R-RR-R--R-RR B--B--B--B--B--B--B--B-- H---H---H---H---H---H--- (the triplet paradiddle with high-hat)	[reply]
Re: Ordering of taint option -T by chromatic (Archbishop) on Jun 07, 2004 at 19:51 UTC
If you look in `perl.c` at the `S_parse_body()` function, you'll see the loop that checks for switches. There's no particular ordering required, at least for taint and baby taint. However, if this function only sees a taint flag after you've already invoked `perl` and it's already processed command-line arguments, the too late for taint macro will take over and exit. As far as I can tell, however, this has absolutely nothing to do with the position of a taint switch within the shebang line and everything to do with when and how you invoke the interpreter.	[reply] [d/l] [select]