Re^2: Trouble skipping lines using Perl

Replies are listed 'Best First'.
Re^3: Trouble skipping lines using Perl by haukex (Archbishop) on Nov 21, 2017 at 16:19 UTC
I went with `next if ($chromosome =~ "chrM");` That works, but personally I wouldn't write it that way, because writing a regex like for example `/chrM/` or `m{chrM}` makes it more visually clear what you want to do (and also allows you to add modifiers). I was using both Strict and Warnings and using the quotes instead of the regex doesn't produce any warnings Are you sure? `next if ($chromosome = "chrM");` should have given you the warning "`Found = in conditional, should be ==`". Perhaps you're not enabling warnings correctly? Update: My apologies about the input data format You can edit your posts (please mark updates as such), see How do I change/delete my post?	[reply] [d/l] [select]
Re^4: Trouble skipping lines using Perl by LeBran (Initiate) on Nov 21, 2017 at 16:55 UTC
Sorry, I meant no warnings were produced at the end result, I got the warning you mentioned initially before using the ~ Cheers	[reply]
Re^3: Trouble skipping lines using Perl by roboticus (Chancellor) on Nov 21, 2017 at 16:30 UTC
LeBran: It didn't give you any warnings because the expression `$chromosome = /^chrM/` is perfectly fine. It just doesn't do what you want it to. Instead of checking whether $chromosome starts with "chrM", it instead checks whether $_ starts with "chrM", and then sets $chromosome to a true value if it does, and a false value otherwise. Since you're not using $_ while parsing your lines, it never starts with "chrM" and always returns a false value. It's a common enough mistake that I could see a case being made for "if ($var = /rex/)" generating a warning, as I expect that "if ($var = ($_ =~ /rex/))" is pretty uncommon (at least, when looking at my code). ...roboticus When your only tool is a hammer, all problems look like your thumb.	[reply] [d/l]
Re^4: Trouble skipping lines using Perl by LeBran (Initiate) on Nov 21, 2017 at 17:03 UTC
Ah, thank you, So presumably `next if ($_ = /^chrM/);` would also be correct? Cheers	[reply] [d/l]
Re^5: Trouble skipping lines using Perl by AnomalousMonk (Archbishop) on Nov 21, 2017 at 17:23 UTC
So presumably `next if ($_ = /^chrM/);` would also be correct? No! The `$_ = /^chrM/` expression matches the regex against `$_` (update: by default, since the `//` match is not explicitly bound to any other scalar by a `=~` binding operator) and then assigns the result of the ~~comparison~~ \| match to `$_` (which at least gives `$_` some defined value, so I guess it's not all bad :) Matching against `$_` (or any other scalar variable) is only semantically correct if that variable has first been given some meaningful value as in a `while (<FILE>) { ... }` loop. Stick to `next if ($line =~ /^chrM/);` Give a man a fish: `<%-{-{-{-<`	[reply] [d/l] [select]
Re^3: Trouble skipping lines using Perl by AnomalousMonk (Archbishop) on Nov 21, 2017 at 17:49 UTC
I went with `next if ($chromosome =~ "chrM");` Also note that `$chromosome =~ "chrM"` matches if `"chrM"` is found anywhere in the `$chromosome` string: `c:\@Work\Perl\monks>perl -wMstrict -le "my $chromosome = 'foo xx chrM yy bar'; print 'found a chrM' if $chromosome =~ 'chrM'; " found a chrM` [download] This is because you no longer anchor the match to the start of the string (as you do in the code in the OP) with the `^` match anchor regex operator. The match `/^chrM/` would IMHO be better for what you seem to want. Another point is that the data posted in the OP has a leading space or spaces in some cases. Leading whitespace will cause the `/^chrM/` match to fail. If leading whitespace may be present in real data, I would recommend something along the lines of `/^\s*chrM/` instead. Give a man a fish: `<%-{-{-{-<`	[reply] [d/l] [select]
Re^4: Trouble skipping lines using Perl by LeBran (Initiate) on Nov 22, 2017 at 11:03 UTC
Ah ok, I'm following you, thanks very much :)	[reply]