1259763 - Reader mode omits opening paragraph on CNN articles

Status: RESOLVED FIXED

(Toolkit :: Reader Mode, defect, P3)

Description

[Adam Roach [:abr]]

Articles on CNN frequently have an opening paragraph that is styled differently than others. Reader mode does not include this lead paragraph in its view.

See, for example, http://money.cnn.com/2016/02/01/news/economy/poverty-inequality-united-states/index.html

Comment 1

[:Margaret Leibovic]

Thanks for the report. This sounds like an issue with the Readability library, so I filed an issue here: https://github.com/mozilla/readability/issues/281

Comment 2

[Evan Tseng [:evanxd]]

Cannot reproduce anymore. The webpage[1] seems already changed. There is no opening paragraph there and the reader mode result is good.

[1]: http://money.cnn.com/2016/02/01/news/economy/poverty-inequality-united-states/index.html

Comment 3

[:Gijs (he/him)]

(In reply to Evan Tseng [:evanxd][:愛聞插低] from comment #2)
> Cannot reproduce anymore. The webpage[1] seems already changed. There is no
> opening paragraph there

I see:

The U.S. has long been heralded as a land of opportunity -- a place where anyone can succeed regardless of the economic class they were born into.

on the page in a different font, and that paragraph does not make it into the reader mode result. Are you seeing something else?

> and the reader mode result is good.
> 
> [1]:
> http://money.cnn.com/2016/02/01/news/economy/poverty-inequality-united-
> states/index.html

Comment 4

[Kevin Brosnan [Ex-Mozilla]]

Gijs I see the same. I do know of cases where large sites to serve different markup to different regions.

Comment 5

[Evan Tseng [:evanxd]]

I can reproduce the issue mentioned on Comment 3. Somehow the `<h2>The U.S. has long been heralded as a land of opportunity -- a place where anyone can succeed regardless of the economic class they were born into.</h2>` node is just removed by some kind of reason.

Good thing is the algorithm chooses correct `topCandidate` (`<div id="storytext">`).

Continue investigate the issue...

Comment 6

[Evan Tseng [:evanxd]]

Added tests[1] to investigate the issue.

[1]: https://github.com/mozilla/readability/pull/347/commits/077bca8721975efa607839f8c2756d2eee323f29

Comment 7

[Evan Tseng [:evanxd]]

Sent a PR[1] with the solution. Let's discuss it there.

[1]: https://github.com/mozilla/readability/pull/347/commits/a0f94b1869b5188dfad66d1d5cc8b6270c5bc4f2

Comment 8

[Evan Tseng [:evanxd]]

Updated the patch to use a new solution to fix the issue[1].

[1]: https://github.com/mozilla/readability/pull/347/commits/64e97fead34ed567025109c2b6df0ae2d8a40db4

Comment 9

[Evan Tseng [:evanxd]]

Fixed all test failures[1].

[1]: https://github.com/mozilla/readability/pull/347/commits/73a020d56675ba649a272d95e025404a9d7936f5

Comment 10

[Evan Tseng [:evanxd]]

Updated patch for review comments: https://github.com/mozilla/readability/pull/347/commits/e6ae86bd9c4dc9d14f87d16f99f4a57b319ce4bb

Comment 11

[Evan Tseng [:evanxd]]

Landed in GitHub: https://github.com/mozilla/readability/commit/498a7b2bf6b6c3460cb85bed57aad0f8196fd532

Comment 12

[Evan Tseng [:evanxd]]

We'll land it in m-c in the MozReview patch[1].

[1]: https://reviewboard.mozilla.org/r/109976/diff/2#index_header

Comment 13

[Evan Tseng [:evanxd]]

Landed in m-c: https://hg.mozilla.org/mozilla-central/rev/d1bef3268b21