Join GitHub today
GitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.
Sign upGitHub is where the world builds software
Millions of developers and companies build, ship, and maintain their software on GitHub — the largest and most advanced development platform in the world.
Unable to extract post url, --ap-mso=Comcast_SSO does not work #11257
Comments
|
cc @raleeper |
|
Seems like reason is very simple |
|
Created PR, verified that with both links above after my change code works. |
|
The regex in adobepass.py now needs to be: |
|
It seems this issue is still a problem. Are there any updates on this. |
|
no
…On Fri, Feb 16, 2018 at 8:14 AM, halolordkiller3 ***@***.***> wrote:
It seems this issue is still a problem. Are there any updates on this.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#11257 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AINer6JqeqV-9Vfz-b3CWVqFvvqDTy38ks5tVPmtgaJpZM4K4UfC>
.
|
|
My comment was not clear I apologize. mekantor said in the previous comment that the code needs to be changed. does this fix the problem and if so where is it? I'm not finding those lines anywhere in adobepass.py |
|
It looks like all the regex and logic has been completely replaced, so my comment doesn't apply anymore, at least not to what I see in the current master branch |
|
Ah I see. Well hopefully this issue with comcast etc can be resolved. |
|
Just tried it, seems to work |
|
On Discovery or History channel etc? Any site that requires a comcast login doesnt seem to work for me. Example: youtube-dl --list-formats --ap-mso=Comcast_SSO --ap-username=email@comcast.net -ap-password=password https://www.tlc.com/tv-shows/counting-on/full-episodes/josephs-wedding I'm getting a error: batch file could not be read |
|
worked for me, you missed a - in --ap-password |
|
Actually I noticed an error saying http://www.history.com/shows/american-pickers/season-18/episode-8 the video url is invalid |
|
I do get an error on that one, but that turned out to be because I don't have History Channel in my comcast subscription. I tested with a Fox URL and it worked, as well as the tlc.com URL above |
|
Alright looks like subscriptions are still having in issues. Even when I tell it to use cookies I get ExtractorError: This video is only available via cable service provider subscription that is not currently supported. You may want to use --cookies. An example link is: https://www.tlc.com/tv-shows/my-600-lb-life/full-episodes/james-b-s-story |
|
Looks like TLC is not in my service, so unlocked episodes work fine, but anything requiring login gets the 403 response. @halolordkiller3 if you go directly to https://www.tlc.com/tv-shows/my-600-lb-life/full-episodes/james-b-s-story and "sign in to watch" does it work in the browser? |
|
@mekantor yes once i sign in it works. but passing the sign in credentials as shown earlier does not work. Example:
|
|
Charter/Spectrum here, also trying to download My 600-lb life, provider sign-in and play works fine in browser, with youtube-dl (latest version at time of this post) using --ap-mso=Charter_Direct or --cookies, get same error as halolordkiller3 ; "ERROR: This video is only available via cable service provider subscription that is not currently supported. You may want to use --cookies." Tried after logging in successfully in browser, still same error. I remembered that for discovery you should use the "go" urls, I didn't realize "go" had been discontinued. So that sucks. I did try running youtube-dl WITH the browser page opened AFTER successfully signing in, using both --ap-mso and --cookies, had no effect on issue. --ap-mso: --ap-cookies |
|
Doesn't seem to extract video data on FXNetworks / Simpsonsworld with DirecTV Now as well... |
|
Can we get an update on this issue? DirecTV is broken with this same issue also. Thanks! |
|
Any updates? Still broken as of April of this year or is all hope lost. |
|
Still running into the original ("Unable to extract post url") issue. For me it's with anything that's locked to your provider on http://disneynow.go.com. I just tried NBC and Simpsons World and they both work fine. Here's the output of my --version, just so you can see I'm on the most recent one: And here's my verbose output when I try to download. |
|
I found my error was occurring only on Linux (Raspbian Stretch, specifically) and not on MacOS. I opened another issue for that: #18577 |
|
I am having this same issue with Comcast_SSO. After reviewing the stack trace, the adobepass.py code, and using fiddler to view the actual data being pulled via HTTPS, I've come to the conclusion that the code is not properly detecting the automatically logged in page. The code appears to be searching the page for the text <!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge,chrome=1">
<meta http-equiv="Content-Type" content="text/html;charset=utf-8">
<meta name="viewport" content="width=device-width,initial-scale=1">
<title></title>
<!-- javascript references removed for brevity -->
<script type="tracking-data-digitalData">
{
"page" : {
"pageInfo" : {
"screenName" : "TVE interstitial",
"language" : "${context.locale.language}",
"referrerId" : "$partnerId"
},
"category" : {
"primaryCategory" : "login",
"designType" : "responsive",
"businessType" : "resi",
"siteType" : "selfservice"
},
"affiliate" : {
"name": "comcast",
"channel" : "web"
},
"codebase" : {
"name" : "cima login"
}
},
"user" : [{
"profile" : [{
"profileInfo" : {
"authenticationType" : "unauthenticated",
"recognizationType" : "unrecognized"
}
}],
"segment" : {
"isLocalized" : false
}
}],
"schemaVersion" : 0.11
}
</script>
<!-- CSS styling removed for brevity -->
</head>
<body>
<div id="background"></div>
<div class="scrim"></div>
<main class="main-content">
<section class="container">
<div class="content">
<div id="loader" class="loader xfinity-logo">
<svg version="1.1" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"
x="0px" y="0px" viewBox="0 0 80 80">
<circle fill="none" stroke-dasharray="251px" stroke-width="2" stroke-linecap="round" stroke-linejoin="round"
cx="40" cy="40" r="39" transform="rotate(0 10 10)">
</circle>
</svg>
</div>
<h1>You're automatically signed in with Xfinity and ready to watch.</h1>
<p>Access to this experience is an added benefit of your Xfinity TV service.</p>
</div>
<footer>Questions? Visit our <a id="FAQ" href="http://customer.xfinity.com/help-and-support/cable-tv/online-tv-through-partner-sites" target="_blank" data-tracking="{"eventAction" : "interstitial faq click"}">FAQs</a></footer>
</section>
</main>
<script type="text/javascript">
function initInterstitialRedirect(config) {
var loopTime = config.delay-1000,
circleLoader = new CircleLoader({
container: document.getElementsByClassName('loader')[0],
loopOnce: true,
loopTime: loopTime
});
circleLoader.startLooping();
setTimeout(function() {
circleLoader.stopLooping();
document.getElementById("loader").className += " checkmark";
}, loopTime);
setTimeout(function() { (config.window || window).location = config.continue; }, config.delay);
}
initInterstitialRedirect({
delay: 3000,
continue: "<removed in case ticket is sensitive>"
});
</script>
<script type="text/javascript">
document.dispatchEvent(new CustomEvent("c-tracking-init-end"));
</script>
</body>
</html>Also, based on the code in adobepass.py, it appears that it is looking for the oauth url as being assigned to I have not had a chance to try, but it appears pull request #17636 would fix the issue. |
|
For those running from source code or compiling your own, the quick fix for Comcast Xfinity is to use this version of adobepass.py: https://raw.githubusercontent.com/bhamboss/youtube-dl/patch-1/youtube_dl/extractor/adobepass.py That is from https://github.com/bhamboss/youtube-dl/blob/patch-1/youtube_dl/extractor/adobepass.py and I can confirm this works with Xfinity's new login if you are coming in from an Xfinity IP address. Just to be clear, Xfinity authenticates differently if you are coming in from one of their IP addresses vs. if you are coming in from the IP of some other provider. I tested the above patch coming from an Xfinity IP. I did NOT test it from another provider's IP. If I run diff -u on the original and patched files I get this: Note the use of "oauth.xfinity.com" rather than "login.comcast.net" which is what apparently worked back in 2016, but not in 2019. |
|
I'm also having the issue: the above fix works for youtube-dl from source, but is there a way to patch the pre-compiled version? |
|
I, too, would like to know if there's a way to resolve pre-compiled versions to handle Comcast changes... |
|
What I had done is take the pull request which fixes the issue, applied it to the most current source code at the time, then used docker to build it into a binary. I appear to have lost the docker information I used to build the code, but, given I have the time, unless someone else has the time and does it before I do, I could keep an up to date fork of the code, with the pull request applied, and have an Azure pipeline build process generate proper patched binaries as the main code is updated, at least until the proper code fix alluded in the comments of the pull request has been implemented. |
|
So, are you pulling the source code every time it's updated -and- then applying this patch? Trying to decide how to do this. Very surprised it wasn't merged in for the last update(s) of the binary. |
|
That is what I would plan to do. What I did was a one time implementation of that, with the intention of repeating it whenever I needed to upgrade the binary. I would imagine that it wouldn't be too difficult to automate this with the correct tools and hooks in place. |
|
So I cloned the latest git and the changes you used in the adobepass.py above. It fails for auth issues. Does yours still work? And if so, mind doing a -v so I can see what I may have different system wise? Using python 3, as required for cookies, and apmso flags. Very weird. |
|
@derekcentrico I apply this patch to my local copy of adobepass.py whenever I update from GitHub. It may help you. I've kinda given up on pull requests for Comcast support since it is impossible for non-Comcast folks to test. |
|
Cool. I'm not very git-smart here. Curious about the patch as I'm getting an error with it.... Is something missing? derek@ubuntu:~/Desktop/YoutubeDL/youtube-dl$ git apply --stat patch.for.comcast derek@ubuntu:~/Desktop/YoutubeDL/youtube-dl$ git am --signoff < patch.for.comcast Note that the patch.for.comcast contains your above pasted into it. |
|
I am probably less git-smart. I omitted necessary headers from the text I pasted. Let's try again. Sorry. |
|
Thanks. Sorry to be a real pain here, but what version of the app are you using that works for you? And, what command are you using if it still works? I have the latest source code and it's broken. I tried with and without the --cookies flag.
|
|
For me, I just pulled down the latest source, updated the adobepass.py and built it. It works without cookies.
Did you grab the latest source, update the adobepass.py and then run the python setup.py build then install? |
|
Yeah, I ran:
I just did it again and used the following command which failed. I wonder if you're downloading actual protected content or public-accessible content. Mind trying my download command?
|
|
That link actually gives a 404 error. I found it's actually "https://go.discovery.com/tv-shows/naked-and-afraid/full-episodes/stalked-on-the-savannah". However, even then, I get an unsupported error: [debug] System config: [] If I remember correctly, there was something with the "Go" websites and the way they either authenticated or the way the video was presented that made it difficult for youtube-dl to operate against it. |
|
Yeah, someone mentioned needing Python 3 and you have a lot of errors I had with Python 2.7. I copied the link which had the tailing ? and didn't think to drop it. Dropping it gives me the following which is still an error but much less than with yours:
So it is falling back to generic.py vs. using one of the several discoveryXXXX.py extractors. I'm not that techie by any stretch so looking at those lines aren't jumping out at me as a way to force it to use the discoverygo.py or one of those other options. |
|
So I may be making some progress. I edited the extractor for discoverygo.py to reflect the new subdomain. Without trying to define a file format I now get: [debug] System config: [] Defining it gets me more crap (-f bestvideo[ext!=webm]+bestaudio[ext!=webm]/best[ext!=webm]) or --all-formats: EDIT: note errors are the same with or without --cookies flag. |
|
Continue your discussions privately. |
Please follow the guide below
xinto all the boxes [ ] relevant to your issue (like that [x])Make sure you are using the latest version: run
youtube-dl --versionand ensure your version is 2016.11.18. If it's not read this FAQ entry and update. Issues with outdated version will be rejected.Before submitting an issue make sure you have:
What is the purpose of your issue?
Tried to use Comcast_SSO with fox and amc