paritybit.ca

commit 18412b032e4de0b35db284082ca53017f287a236
parent 9cf1f5319f0caa38c59f9001627de324316189ce
Author: Jake Bauer <jbauer@paritybit.ca>
Date:   Thu,  4 Jun 2020 01:02:27 -0400

Gopherize progress

Diffstat:
M
gopherize
 | 
26
++++++++++++++++++++------

1 file changed, 20 insertions(+), 6 deletions(-)
diff --git a/gopherize b/gopherize
@@ -22,19 +22,25 @@ use strict;
 use warnings;
 our $VERSION = "0.1.0";
 
-use HTML::TreeBuilder 5 -weak;
-use HTML::Strip;
+use HTML::TreeBuilder;
 use File::Basename;
 
 my $tree = HTML::TreeBuilder->new();
-my $hs = HTML::Strip->new();
 my @files = @ARGV;
 my $output_path="public/gopher";
-my @links;
 my $fh;
 
 # Prints links to the bottom of the page;
 sub print_links {
+    my (@links) = @_;
+    print "Found ", scalar(@links), " links.\n";
+    foreach my $link (@links) {
+        print "Link: ", $link->[0], ".\n";
+        print "Obj: ", $link->[1], ".\n";
+        print "Prop ", $link->[2], ".\n";
+        print "Tag: ", $link->[3], ".\n";
+        print "=========\n"
+    }
     return;
 }
 
@@ -60,6 +66,13 @@ foreach my $file (@files) {
         next;
     }
 
+    print "Opening input file: $file\n";
+    open (my $orig_file, "<:encoding(UTF-8)", $file)
+        or do {
+            warn "Could not open input file $file: $!";
+            next;
+        };
+
     print "Creating output file: $output_path/$file_name.gph...\n";
     open($fh, ">", "$output_path/$file_name.gph")
         or do {
@@ -68,8 +81,9 @@ foreach my $file (@files) {
         };
 
     print "Parsing HTML from: $file...\n";
-    $tree->parse_file($file);
-    $tree->dump;
+    $tree->parse_file($orig_file)->elementify();
+    my $links_ref = $tree->extract_links("a", "img", "audio", "video");
+    print_links(@$links_ref);
 
     close($fh);
 }