eleksir/hexchat-scripts
1
Fork 0
Code Activity

fix mimetype handling, make script syncronous

Browse Source
This commit is contained in:
Sergey Fedosov
2016-09-04 21:20:45 +03:00
parent a98fc36c6f
commit 86b0857f9f
1 changed files with 57 additions and 49 deletions
Download Patch File Download Diff File Expand all files Collapse all files

106
img_url_wget.pl
View File

@@ -2,20 +2,23 @@
# HexChat Perl script
# Automatically save images that are linked in a channel to [hexchat-dir]/imgsave on linux/unices or %USERPROFILE%\Downloads on Windows
# You may need to create this directory first
# The bad_domains array may be used to exclude image URLs with domains matching any of a list of patterns. Set to an empty list to disable.
# TODO: enhance urlencode with punicode
# TODO: suppress any errors (errors only in debug mode)
# TODO: create menu and plugin-own ini-file or some sort of config (or add one more section to hexchat config, which is possible via api)
# TODO: fix bad behaving picture hostings that put wrong mime-types on content, tt-rss is one of them
# NOTE: Hexchat run scripts in syncronously, all script' children should exit before main thread, so massive file download can cause
# "stop whe world" effect for a while, so to effectively aviod it downloading part should _fork_ childern as separate parentless
# processes.
# TODO: Metaprogamming stuff (via eval) with downloader and proper mime-type handling.
use strict;
use warnings "all";
use threads;
use URI::URL;
use Image::Magick;
use File::Copy;
my $wgetpath;
@@ -27,14 +30,17 @@ if (-f "/bin/wget") {
$wgetpath = '/usr/local/bin/wget';
}
sub hookfn; # in defaults we trust: dunno arrayname/args name, but default values looks good :)
sub dlfunc(@); # download thread
sub cdlfunc(@); # thread, that checks and downloads stuff
sub is_picture($); # sub thar checks the mime-type
sub cdlfunc($); # thread, that checks and downloads stuff
sub is_picture($); # sub that checks mime-type given by the web server, this check should be used only to
# detect that given url contain picture or video, after download we should check mime-type
# again and rename file if needed
sub urlencode($); # correctly encode url :)
sub hookfn; # in defaults we trust: dunno arrayname/args name, but default values looks good :)
my $script_name = "Image URL Auto Grabber and Downloader, wget flavour";
HexChat::register($script_name, '0.6-alpha1', 'Automatically grabs and downloads image URLs via wget');
HexChat::register($script_name, '0.6-alpha2', 'Automatically grabs and downloads image URLs via wget');
HexChat::print("$script_name loaded\n");
HexChat::hook_print('Channel Message', \&hookfn);
@@ -45,56 +51,33 @@ HexChat::hook_print('Channel Action Hilight', \&hookfn);
my @bad_domains = (); #array of regexs for domain name exclusion
#my @bad_domains = (qr/^(?:.*\.)?example\.com$/, qr/\bbannedword\b/, qr/^name\./); #example, exclude: example.com and subdomains, domains containing the term bannedword, and domains starting with name.
sub hookfn {
my ($nick, $text, $modechar) = @{$_[0]};
my @words = split(/\s+/, $text);
sub dlfunc(@) {
my $url = shift;
my $file = shift;
foreach (@words) {
# here we try to catch string that can contain http or https at beginning (assume http if none)
# and after all ends with known extension
if ($_ =~ m{^(?:https?://)?([a-zA-Z0-9.-]+\.[a-zA-Z]+)/(?:.*)\.(?:jpe?g|png|gif|mp4|webm)$}i) {
$url = urlencode($url);
`$wgetpath --no-check-certificate -q -T 20 -O '$file' -o /dev/null '$url'`;
if ($file =~ /(png|jpe?g|gif)$/i){
my $im = Image::Magick->new();
my $rename = 1;
my (undef, undef, undef, $format) = $im->Ping($file);
foreach my $re (@bad_domains) {
return HexChat::EAT_NONE if ($1 =~ $re);
if (defined($format)) {
$rename = 0 if (($format eq 'JPEG') and ($file =~ /jpe?g$/i));
$rename = 0 if (($format eq 'GIF') and ($file =~ /gif$/i));
$rename = 0 if (($format =~ /^PNG/) and ($file =~ /png$/i));
if ($rename == 1) {
mv($file, sprintf("%s.%s", $file, lc($format)));
}
# reconstruct url if it does not contain proto schema
if ( $_ !~ m{^https?://} ) {
$_="http://".$_;
}
# create filename
my $fn;
unless (-d sprintf("%s/imgsave", HexChat::get_info("configdir"))) {
mkdir (sprintf("%s/imgsave", HexChat::get_info("configdir")));
}
$fn = HexChat::get_info("configdir") . "/imgsave/" . s/[^\w!., -#]/_/gr;
my $th = threads->create(\&dlfunc, $_, $fn);
$th->detach();
} elsif ($_ =~ m{https?://([a-zA-Z0-9.-]+\.[a-zA-Z]+)/(?:.*)}) {
# here we try to download _url_ that begins with https and does not ends with known file type
my $full_url = $_;
my $fqdn = $1;
my $uri_part = $2;
my $thr = threads->create(\&cdlfunc, $full_url);
$thr->detach();
}
}
return HexChat::EAT_NONE;
}
sub dlfunc(@) {
my $file = shift; # 0
my $url = shift; # 1
$url = urlencode($url);
`$wgetpath -q -T 20 -O '$file' '$url'`;
return 0;
}
sub cdlfunc(@) {
sub cdlfunc($) {
my $url = shift;
my $extension = is_picture($url);
@@ -106,6 +89,14 @@ sub cdlfunc(@) {
my $fname = HexChat::get_info("configdir") . "/imgsave/" . s/[^\w!., -#]/_/gr . ".$extension";
if (lc($url) =~ /\.(gif|jpeg|png|webm|mp4)$/) {
if ($1 eq $extension) {
$fname = HexChat::get_info("configdir") . "/imgsave/" . s/[^\w!., -#]/_/gr;
}
}
dlfunc($url, $fname);
}
}
@@ -115,7 +106,7 @@ sub is_picture($) {
$url = urlencode($url);
my $r = undef;
$r = `$wgetpath --no-check-certificate -q --method=HEAD -S --timeout=15 "$url" 2>&1`;
$r = `$wgetpath --no-check-certificate -q --method=HEAD -S --timeout=5 "$url" 2>&1`;
foreach (split(/\n/, $r)) {
next unless($_ =~ /Content\-Type: (.+)/);
@@ -142,4 +133,21 @@ sub urlencode($) {
return $url;
}
sub hookfn {
my ($nick, $text, $modechar) = @{$_[0]};
my @words = split(/\s+/, $text);
foreach (@words) {
# here we try to catch string that can contain http or https at beginning (assume http if none)
# and after all ends with known extension
if ($_ =~ m{https?://([a-zA-Z0-9.-]+\.[a-zA-Z]+)/(?:.*)}) {
my $full_url = $_;
my $fqdn = $1;
my $uri_part = $2;
cdlfunc($full_url);
}
}
return HexChat::EAT_NONE;
}
__END__