Perl数组和哈希引用处理

Question

my problem is that i'm getting no Output at the end of Script.. All the prints while parsing are ok, but arrays at end are empty.

What i'm doing wrong? Isnt this the way to handle the refs?

thx 4 response

my %branches = ();
print "<pre>";
my %tmp_branch;

while (defined($_ = shift @bugs)) {
    my $bug_id = $_->id;
    my $bug_product = $_->product;
    my $content = $browser->get("http://****?ticket=".$bug_id);
    $content = $content->decoded_content;

    my @rows = split /\n/, $content;
    my $trigger = 0;

    while (defined($_ = shift @rows)) {
        chomp;
        if ($_ eq "") {
            $trigger = 0;
        }
        elsif (/Branch: (.*)/) {
            if (exists $branches{$1}) {
                my $branch_ref = $branches{$1};
                %tmp_branch = %$branch_ref;
                print "existing Branch: $1\n";
            } else {
                my %new_branch = ();

                my @sources = ();
                my @wfs = ();
                my @methods = ();

                $new_branch{'sources'} = \@sources;
                $new_branch{'methods'} = \@methods;
                $new_branch{'wfs'} = \@wfs;

                $branches{$1} = \%new_branch;
                %tmp_branch = %new_branch;
                print "new Branch: $1\n";
            }
        }
        elsif (/Sourcen.*:/) {
            $trigger = "sources";
        }
        elsif (/geaenderte Methoden.*:/) {
            $trigger = "methods";
        }
        elsif (/geaenderte Workflows.*:/) {
            $trigger = "wfs";
        }
        elsif ($trigger && $_ ne "") {
            my $tmp_array_ref = $tmp_branch{$trigger};
            my @tmp_array = @$tmp_array_ref;
            push @tmp_array, $_;
            print "find $trigger: $_\n";
        }
    }
}

print "\n\n\n";

while (my ($k,$v)=each %branches){
    my $branch_ref = $v;
    my %tmp_branch = %$branch_ref;
    my $sources_ref = $tmp_branch{'sources'};
    my @sources = @$sources_ref;
    my $methods_ref = $tmp_branch{'methods'};
    my @methods = @$methods_ref;
    my $wfs_ref = $tmp_branch{'wfs'};
    my @wfs = @$wfs_ref;

    print "Branch: $k\nSources:\n";
    print @sources;
    print "\nMethods:\n";
    print @methods;
    print "\nWorkflows:\n";
    print @wfs;
    print "\n";
}

print "</pre>";

Sample Input:

    Kontext Auswertung fuer Ticket: #12345 (xxxxSomeTextxxx)
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    HINWEIS: xxxxSomeTextxxx

    Branch: HEAD
    ~~~~~~~

    Sourcen (4):
    IamArow
    IamArow2
    IamArow3
    IamArow4

    geaenderte Methoden (1):
    IamArow

    geaenderte Workflows (2):
    IamArow
    IamArow2

Answer 1

It's quite hard to figure it out without any input data, because it means that we can't run a copy of the script on our own machines! It is generally pretty useful if your sample code is self-contained !

That said, I think your problem stems from doing this kind of thing:

my $branch_ref = $branches{$1};
%tmp_branch = %$branch_ref;

The second line does a shallow copy of the hash, so %tmp_branch is no longer the same hash as the one referenced by $branches{$1} . When you add data to the %tmp_branch hash, you are not adding data to the $branches{$1} hash.

@tmp_array suffers similarly.