Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Create Vice.js #1366

Merged
merged 4 commits into from Aug 4, 2018

Conversation

@owcz
Copy link
Contributor

commented Jul 10, 2017

Handles every text-focused content site in the network. Some subsites (mainly i-D, Amuse i-D, and Vice News) vary significantly from the others, so some selectors were uglier than I'd prefer so as to accommodate. The rest more or less follow the same format. While i-D uses OpenGraph and might have benefited from using EM, the rest needed custom coding, so I thought it was best to build i-D support into a single Vice translator.

Need help with two issues:

  1. Every snapshot I verified leads to a 404 page. Is there a workaround or should I just remove the snapshots altogether?

  2. Some multis from the network sites that follow the same format have a weird issue:
    https://broadly.vice.com/en_us
    https://broadly.vice.com/en_us/topic/occult
    https://waypoint.vice.com/en_us
    https://waypoint.vice.com/en_us/latest
    https://thump.vice.com/en_us/search?q=venetian
    For instance, try any one of the above (all multis) and the saved articles will lack both author and tags, even though both do save properly if you run the scraper on the individual page (without the multi). There's something about the multi that leaves those two parameters out.

    Scaffold is strange too. Add any one of those site's individual articles to the test, click "save", and it scrapes fine, as it normally does in the browser. But then "run" any one of those same individual articles and it'll ignore the site's authors and tags (same as above).

    I tried to isolate the issue by removing everything that wasn't necessary for, say, broadly.vice.com's multi, but the scraper still returned null for each page's author/tags (even though the scraper returned the right results if run on the same pages individually). So not sure if it's an issue with the multi code or something with text()/attr() (#1277)

Create Vice.js
Handles every text-focused content site in the network

Need help with two issues marked in the code
Vice.js Outdated
}
}

// archives appear to 404 on load?

This comment has been minimized.

Copy link
@owcz

owcz Jul 10, 2017

Author Contributor

Snapshots on every page I checked will succeed, but when later checked, they all 404. Not sure if there's a workaround or if it would be better to leave the snapshots out altogether.

This comment has been minimized.

Copy link
@zuphilip

zuphilip Jul 16, 2017

Collaborator

Saving this website also fails when I try to do this in the browser. They are using a lot of JavaScript and some iframe things which seems not so friendly for saving purposes. Moreover, I haven't found a link to a "print friendly version" which sometimes exists. I suggest to skip saving snapshots here.

@zuphilip

This comment has been minimized.

Copy link
Collaborator

commented Jul 16, 2017

I guess that the problems you are describing are connected to some dynamic loading of the website or building the website out of some JSON together with some JavaScript calls...

When coming from a multiple to a single article, then I see that some JSON data is loaded:

vice-loading

Actually, it looks promising to use such API calls directly and work with that JSON data rather than scrape from the page directly. In some the examples it is enough to add ?json=true to the url, e.g. https://www.vice.com/en_us/article/padaqv/anti-g20-activists-told-us-what-they-brought-to-the-protest-in-hamburg?json=true . Let me know if this helps.

Remove snapshots
as discussed
@zuphilip

This comment has been minimized.

Copy link
Collaborator

commented Jul 17, 2017

The news site on vice uses the WordPress API, i.e. e.g. https://news.vice.com/wp-json/wp/v2/posts/24702

@adam3smith

This comment has been minimized.

Copy link
Collaborator

commented Aug 14, 2017

Is this ready for review or are you still tweaking this?

@owcz

This comment has been minimized.

Copy link
Contributor Author

commented Aug 14, 2017

Still tweaking, thanks

@owcz

This comment has been minimized.

Copy link
Contributor Author

commented Jul 9, 2018

okay, ready for review

  • rewritten to use JSON
  • falls back to EM for News, i-D, and Amuse, and supplements with JSON-LD for News until it's officially built into EM

@zuphilip zuphilip removed the waiting label Jul 10, 2018

@adam3smith
Copy link
Collaborator

left a comment

A couple of suggestions & questions. Shou.ld be quick

Vice.js Outdated
"translatorID": "131310dc-854c-4629-acad-521319ab9f19",
"label": "Vice",
"creator": "czar",
"target": "^https?://((www|broadly|creators|i-d|amuse-i-d|impact|motherboard|munchies|news|noisey|sports|thump|tonic|waypoint)\\.)?vice\\.com",

This comment has been minimized.

Copy link
@adam3smith

adam3smith Jul 26, 2018

Collaborator

I'd be OK with the simpler ^https://(.+?\\.)vice\\.com unless you need to specifically exclude certain subdomains


function detectWeb(doc, url) {
if (/\/(article|story)\//.test(url)) {
return "blogPost";

This comment has been minimized.

Copy link
@adam3smith

adam3smith Jul 26, 2018

Collaborator

I would have thought of Vice as an online magazine, not as a series of blogposts -- not going to insist, but consider switching this.

To wits: Wikipedia has Vice as a magazine and it appears to have a ISSN

This comment has been minimized.

Copy link
@owcz

owcz Jul 28, 2018

Author Contributor

I could see potentially adding it for the base vice.com even though I don't know its overlap with the print circulation
But I'd consider all of its many subdomains to be more blog- than magazine-like in nature

Vice.js Outdated
function detectWeb(doc, url) {
if (/\/(article|story)\//.test(url)) {
return "blogPost";
} else if (/\.vice\.com\/?($|\w\w(\_\w\w)?\/?$)|\/(search\?q=)|topic\/|category\/|(latest|read)($|\?page=)/.test(url) && getSearchResults(doc, true) ) {

This comment has been minimized.

Copy link
@adam3smith

adam3smith Jul 26, 2018

Collaborator

I don't think you want that beginning \. before vice.com, do you? (doesn't match https://vice.com)

This comment has been minimized.

Copy link
@owcz

owcz Jul 28, 2018

Author Contributor

my understanding is that all vice.com addresses redirect to www.vice.com but I suppose the translator's target detection would have caught any bad addresses by this point

Vice.js Outdated


function scrape(doc, url) {
var jsonURL = url+'?json=true';

This comment has been minimized.

Copy link
@adam3smith

adam3smith Jul 26, 2018

Collaborator

You'll almost certainly want to clean the URL by doing url.replace(/#.+/, "") -- otherwise internal page navigation as well as social media origin info will get included in the jsonURL and will likely break it.

This comment has been minimized.

Copy link
@owcz

@adam3smith adam3smith merged commit 114abf4 into zotero:master Aug 4, 2018

1 check passed

continuous-integration/travis-ci/pr The Travis CI build passed
Details
@adam3smith

This comment has been minimized.

Copy link
Collaborator

commented Aug 4, 2018

Thanks!

@owcz owcz deleted the owcz:patch-4 branch Aug 4, 2018

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
3 participants
You can’t perform that action at this time.