Re: Strange compile error?

Welcome to the Monastery.

"Sorry for a stupid question! But I want to learn: ..."

Not a "stupid question"!
++ for "I want to learn".
Here's a description of what's going on.

A '#' introduces a comment. See "perlsyn: Comments". So, when compiling, Perl sees your lines 9 to 12 as:

my $N= 


say "N = $N";
[download]

As Perl doesn't care about the intervening whitespace, that's basically:

my $N= say "N = $N";
[download]

Assignment is a right-to-left operation. See "perlop - Perl operators and precedence". That's a fairly long page; the parts most pertinent to your question are:

The table near the end of the "Operator Precedence and Associativity" section.
The "Assignment Operators" section.

So, Perl attempts to compile 'say "N = $N"' first. It hasn't evaluated the left side of the assignment ('my $N') yet, so it sees $N as a "Global symbol" and hence the error message.

There is a subtle point here that isn't immediately, or intuitively, obvious. A statement such as 'my $x = ...;' has two distinct parts:

'my $x' declares $x as a lexical variable at compile time.
The whole statement 'my $x = ...;' is compiled at compile time; however, the actual assignment is a runtime operation.

I see a couple of posts showing how to get length and last index of an array. Another way to get the length is:

my $N= 0+@Org;
[download]

As a further tangential point, a "use VERSION" statement, where VERSION is 5.12 or greater, you get 'use strict;' automatically. I like to put the use VERSION statement first: if the user does not have a sufficiently high enough version, you might as well stop here rather than doing other (effectively) pointless processing. Accordingly, instead of

use strict;
use warnings;

use 5.020;
[download]

I'd write

use 5.020;
use warnings;
[download]

Just something to consider for future reference.

— Ken

Comment on Re: Strange compile error? Select or Download Code

Replies are listed 'Best First'.
Re^2: Strange compile error? by afoken (Chancellor) on Apr 30, 2024 at 13:46 UTC
A '`#`' introduces a comment. Oh no! You triggered that annoying little bean counter! So here we go ... A '`#`' does not ALWAYS introduce a comment: `$#array` is the last index of `@array` `#` can be used in place of `/` in `s///`, `m//`, `q//`, `qq//`, `qx//`, `qw//`, `y///`, `tr///`. `$#` was a magic variable, deprecated long ago, removed in Perl 5.10.0 (see perlvar) `#` inside a comment is ignored like any other character until end of line And of course, a `#` in a string constant is just that. It does not start a comment inside a string constant. Alexander -- Today I will gladly share my knowledge and experience, for there are no sweeter words than "I told you so". ;-)	[reply] [d/l] [select]
Re^3: Strange compile error? by LanX (Saint) on Apr 30, 2024 at 22:43 UTC
BUT ... Inside POD, you forgot POD. WAIT... After `__DATA__` and `__END__` STOP ... could be inside a here-doc marker too MOMENT ... A source filter could transform them (see half of f*ck!ng `ACME::` ...) HOLD ... what if C-preprocessor macros are activated? AND ... (fill in more nitpicking, yawn 🥱) Cheers Rolf _{(addicted to the Perl Programming Language :) see Wikisyntax for the Monastery} UPDATES Inserted 3 more criteria in hindsight... 🙃	[reply] [d/l] [select]
Re^2: Strange compile error? by LanX (Saint) on Apr 29, 2024 at 19:47 UTC
> Assignment is a right-to-left operation That's right, but sorry you are overcomplicating things. Precedence isn't relevant here: `DB<1> use strict; my $N = $N; Global symbol "$N" requires explicit package` [download] Declarations are only effective after the statement, in this case behind the semicolon. This was done to allow accessing equally named variables from a surrounding scope. See perlsub#Private-Variables-via-my() The declared variable is not introduced (is not visible) until after the current statement. Thus, my $x = $x; can be used to initialize the new `$x` with the value of the old `$x`, and the expression my $x = 123 and $x == 123 is false unless the old `$x` happened to have the value `123`. Cheers Rolf _{(addicted to the Perl Programming Language :) see Wikisyntax for the Monastery}	[reply] [d/l] [select]
Re^3: Strange compile error? by kcott (Archbishop) on Apr 29, 2024 at 23:08 UTC
I was talking about associativity (as in "right-to-left" operation). The table to which I referred lumps associativity and precedence together: there's not a lot I can do about that. — Ken	[reply]
Re^4: Strange compile error? by LanX (Saint) on Apr 30, 2024 at 00:19 UTC
But it doesn't matter which part of a statement is compiled first, be it by associativity or precedence or quantum flux. The docs are clear: The declared variable is not introduced (is not visible) until after the current statement. Cheers Rolf _{(addicted to the Perl Programming Language :) see Wikisyntax for the Monastery}	[reply]

UPDATES