How to Access an Array in Perl for Regex

Question

I have two inputs reading into my command prompt, the first being a series of words that are to be searched by the program I'm writing, and the second being the file that contains where the words are to be found. So, for instance, my command prompt reads perl WebScan.pl word WebPage000.htm

Now, I have no trouble accessing either of these inputs for printing, but I am having great difficulty accessing the contents of the webpage so I can perform regular expressions to remove html tags and access the content. I realize that there is a subroutine available to do this without regular expressions that is far more effective, but I need to do with with regular expressions :(.

I can access the html file for printing with no trouble:

open (DATA, $ARGV[1]);
my @file = <DATA>;
print @file;

Which prints the entire code of the html page, but I am unable to pass regular expressions in order to remove html blocks. I keep receiving an error that says "Can't modify array dereference in s/// near," which is where I have my specific regular expression. I'm not sure how to get around this- I've tried converting the array into a scalar but then I am unable to access any of the data in the html at all (and no, it doesn't just print the number of values in the array :P)

How do I access the array's contents so I can use regular expressions to refine the desired output?

Answer 1

It sounds like you are doing something like @file =~ s/find/replace/;. You are getting that error because the left hand side of the regex binding operator imposes scalar context on its argument. An array in scalar context returns its length, but this value is read only. So when your substitution tries to perform the replacement, kaboom.

In order to process all of the lines of the file, you could use a foreach loop:

foreach my $line (@file) {$line =~ s/find/replace/}

or more succinctly as:

s/find/replace/ for @file;

However, if you are running regular expressions on an HTML file, chances are you will need them to match across multiple lines. What you are doing above is reading the entire file in, and storing each line as an element of @file. If you use one of Perl's iterative control structures on the array, you will not be able to match multiple lines. So you should instead read the file into a single scalar. You can then use $file =~ s/// as expected.

You can slurp the file into a single variable by temporarily clearing the input record separator $/:

my $file = do {local $/; <DATA>};

In general, regular expressions are the wrong tool for parsing HTML, but it sounds like this is a homework assignment, so in that case its just practice anyway.

And finally, in modern Perl, you should use the three argument form of open with a lexical file handle and error checking:

open my $DATA, '<', $ARGV[1] or die "open error: $!";

my $file = do {local $/; <$DATA>};