clearing the memory

[richardko]

hi, i am trying to parse a bunch of html files in a directory and i need help clearing the memory.
I get the same output from "$table_html" for all the files.

Seems like table_html doesnt get emptied out. I tried the delete and delete_content() methods but it did not work.
Any idea?

foreach $file (@dir_contents)
{
        if(!(($file eq ".") || ($file eq "..")))
        {
                $filename = "/var/[URL unfurl="true"]www/cgi-bin/pages/$file";[/URL]           
                $te->parse_file($filename);                
                $table = $te->first_table_found;
                $table_tree = $table->tree;                
                $table_html = $table_tree->as_HTML;
                print $table_html."\n";            
         }
}

 

[KevinADC]

my $table_html = $table_tree->as_HTML;

------------------------------------------
- Kevin, perl coder unexceptional!

 

[richardko]

Thanks for the reply. However, defining

my $table_html = $table_tree->as_HTML;

still gives me the same result. I think the problem lies with $table->tree being the same every time but I am not sure.

 

[MillerH]

Your problem lines in your parser "$te". Without knowing any about what it is or how it was created, we can't possible suggest how to flush the buffer. I would expect that there will be some way of clearing it's buffer though.

If all else fails, look into recreating the parser each time.

- Miller

 

[KevinADC]

maybe:

my $te->parse_file($filename);

------------------------------------------
- Kevin, perl coder unexceptional!

 

[richardko]

redeclaring the object does seem to work. Thanks for the response.

 

[MillerH]

Glad we could help.

Just as a note concerning stylistic coding. I personally do not like to include a large if block directly within a while. Most often you can invert or move the logic in such a way as to remove the block entirely and make the code a lot more readable.

For example, take a look at the following examples:

Using next to specifically handle those special cases.

foreach $file (@dir_contents) {
	[COLOR=green]next if $file eq "." || $file eq "..";[/color]

	$filename = "/var/[URL unfurl="true"]www/cgi-bin/pages/$file";[/URL]
	$te->parse_file($filename);
	$table = $te->first_table_found;
	$table_tree = $table->tree;
	$table_html = $table_tree->as_HTML;
	print $table_html."\n";
}

Using a regex instead:

foreach $file (@dir_contents) {
	[COLOR=green]next if $file =~ /^\.+$/;[/color]

	$filename = "/var/[URL unfurl="true"]www/cgi-bin/pages/$file";[/URL]
	$te->parse_file($filename);
	$table = $te->first_table_found;
	$table_tree = $table->tree;
	$table_html = $table_tree->as_HTML;
	print $table_html."\n";
}

Moving the filtering to the while by adding a grep

foreach $file ([COLOR=green]grep {!/^\.+$/}[/color] @dir_contents) {
	$filename = "/var/[URL unfurl="true"]www/cgi-bin/pages/$file";[/URL]
	$te->parse_file($filename);
	$table = $te->first_table_found;
	$table_tree = $table->tree;
	$table_html = $table_tree->as_HTML;
	print $table_html."\n";
}

As I hope you can see, each of the above coding styles is a lot more readable then having a giant if statement. Another example would be to add the grep to the statement creating @dir_contents itself, as this is a common type of filtering that is needed.

- Miller

 

[stevexff]

Useful tips from Miller. While we are on the subject of grepping out things we don't want, what happens if @dir_contents contains other directories?

Steve

[small]"Every program can be reduced by one instruction, and every program has at least one bug. Therefore, any program can be reduced to one instruction which doesn't work." (Object:erlDesignPatterns)[/small]

 

[richardko]

On the subject of grep I have another dilemma regarding html i need to replace.
Part of the html contains "img src" tag like this:

<img alt="pad" border="0" height="5" src="[URL unfurl="true"]http://some.example.com/Img/trans_1x1.gif"[/URL] width="1" />

The html contains more than one "img src" tags. I need to replace the

[URL unfurl="true"]http://some.example.com/Img/[/URL]

with a different url. (something like

http://website.com/images)

I am not even sure where to start looking in regex to match all the way from http to the last backslash "/". I could do that once and replace with a url not once but where ever this occurs. I am guessing backreference is the way to go on this one.
Any idea how to tell regex to match all the backslash upto the last one?
thanks

 

[MillerH]

On the subject of grep I have another dilemma regarding html i need to replace.
Part of the html contains "img src" tag like this:

It is certainly possible to create complicated regex'es to parse and translate HTML, but it is generally a waste of your time. Instead, learn one of the HTML parsers available via CPAN, and make it do the translation.

It's a much better investment of your time (let alone ours), then making single use regex's simply as a programming challenge.

- Miller