"406 not acceptable" errors with LWP::UserAgent::get

jimhenry has asked for the wisdom of the Perl Monks concerning the following question:

I'm writing a podcatcher and I ran into a problem where one particular podcast (whose RSS feed downloads fine in Firefox or wget) was failing to download when I used the default HTTP headers I used with all the other podcasts I've tested with. It would give a "406 Not Aceptable" error, which research suggested was caused by bad Accept headers. I ran wget -d to see what headers it was using and copied them into my script (or rather the config file for my script), and got the same error with this podcast. Below, find a simplified version of the code that reproduces the problem.

#! /usr/bin/perl

use strict;
use warnings;

use LWP::UserAgent;

my $ua = LWP::UserAgent->new;
$ua->agent('Mozilla/5.0');
$ua->show_progress( 1 );

my %headers = ( 'Accept' => '*/*',
        'Accept-Encoding' => 'identity',
        'Connection' => 'Keep-Alive',
#        'Host' => 'feed.podbean.com' );
        'Host' => 'makeoursmarvel.com' );

#my $url = 'https://feed.podbean.com/betheserpent/feed.xml';
my $url = 'http://makeoursmarvel.com/feed/podcast/';


my $response = $ua->get( $url, %headers );

if ( $response->is_success ) {
    print "Okay!\n";
} else {
    my $status = $response->status_line;
    print "failed to download $url: $status\n";
}
[download]

If I use the commented out lines instead for the Host and $url value (checking a different podcast's RSS feed), everything works fine. I also tried using the default Firefox Accept: header, based on https://developer.mozilla.org/en-US/docs/Web/HTTP/Content_negotiation/List_of_default_Accept_values, and got the same 406 error. The same podcast also give me a 406 error when I try to get an individual mp3 file.

Any ideas how to narrow the problem down further, if not fix it?

I'm using Perl v5.26.1 and LWP::UserAgent version 6.31.

Comment on "406 not acceptable" errors with LWP::UserAgent::get Download Code

Replies are listed 'Best First'.
Re: "406 not acceptable" errors with LWP::UserAgent::get by Your Mother (Archbishop) on Aug 17, 2020 at 23:44 UTC
Seems the main thing it doesn’t like is the agent. Your code gave me the same error, even tweaked and pruned down until I swapped the agent for something that isn’t, apparently, blacklisted. `use 5.10.0; use LWP::UserAgent; my $ua = LWP::UserAgent->new; $ua->agent('DomoArigato/3.0'); my $url = 'http://makeoursmarvel.com/feed/podcast/'; my $response = $ua->get($url); say $response->is_success ? "OK!" : $response->as_string; __END__ OK!` [download]	[reply] [d/l]
Re^2: "406 not acceptable" errors with LWP::UserAgent::get by thomas895 (Deacon) on Aug 18, 2020 at 00:34 UTC
Usually to get around these kinds of blocks the easiest way is to set the user agent string to be the same one your browser uses -Thomas "Excuse me for butting in, but I'm interrupt-driven..."	[reply]
Re^3: "406 not acceptable" errors with LWP::UserAgent::get by Your Mother (Archbishop) on Aug 18, 2020 at 01:14 UTC
Disagree in the main. :P People blacklist agents, not whitelist, and a one-off for an agent that is NEVER abusive to a service is less likely to get crapcanned than one that shares a (base) name with 30% of the traffic.	[reply]
Re^3: "406 not acceptable" errors with LWP::UserAgent::get by jimhenry (Acolyte) on Aug 18, 2020 at 20:18 UTC
Thanks for the replies. I tried setting the agent to the user-agent string used by my current version of Firefox and had no trouble. (I'm still not sure why I was getting a 406 error for this podcast instead of a 403 Forbidden error, which I was getting on a number of podcasts with the default agent of "libwww-perl".)	[reply]
Re^4: "406 not acceptable" errors with LWP::UserAgent::get by Your Mother (Archbishop) on Aug 18, 2020 at 23:57 UTC
Re: "406 not acceptable" errors with LWP::UserAgent::get by tobyink (Canon) on Aug 19, 2020 at 11:42 UTC
As an aside, there's no need to set a `Host` header. Yes, it's required by HTTP (unless you're still using HTTP 1.0), but LWP::UserAgent will do that for you. As would HTTP::Tiny or any other HTTP client library worth its salt. The only benefit you get from specifying it manually is the wonderful opportunity to screw things up. toby döt ink	[reply] [d/l]


Think about Loose Coupling
	PerlMonks