[meta] Refine isProbablyReaderable algorithm

NEW
Unassigned

Status

()

Toolkit
Reader Mode
P3
normal
2 years ago
8 months ago

People

(Reporter: Margaret, Unassigned)

Tracking

(Depends on: 2 bugs, Blocks: 1 bug, {meta})

Trunk
Points:
---
Dependency tree / graph
Bug Flags:
qe-verify -

Firefox Tracking Flags

(Not tracked)

Details

(Whiteboard: [reader-mode-readability-algorithm])

(Reporter)

Description

2 years ago
In bug 1143844, we implemented an algorithm to only show the reader view button if a page has at least 5 <p> elements with a decent amount of text. It turns out that this might be a bit too aggressive, we should try to refine this.

Some testcases documented in https://github.com/mozilla/readability/issues/73:

1/ Wordpress:

Reader mode not activating:
https://blog.mozilla.org/blog/2015/03/16/please-welcome-david-bryant-vice-president-of-platform-engineering/

Reader mode activating:
https://blog.mozilla.org/blog/2015/03/03/unity-5-ships-and-brings-one-click-webgl-export-to-legions-of-game-developers/

2/ Dotclear

Reader mode not activating:
http://blog.mozfr.org/post/2015/03/La-lune-eclipse-le-soleil

Reader mode activating:
http://blog.mozfr.org/post/2015/03/Il-y-a-26-ans-naissait-le-Web-Coup-d-oeil-dans-le-retro
Setting qe-verify- for this bug, as the testing required here will be covered by the verification conducted for Bug 1139165.
Flags: qe-verify-
Priority: -- → P3
(Reporter)

Comment 2

2 years ago
markh has a patch over in bug 1149859. Let's morph this into a meta bug, since there may be other things we want to try to make this logic better.
Depends on: 1149859
Keywords: meta
(Reporter)

Updated

2 years ago
Depends on: 1150695
Blocks: 1149057

Updated

2 years ago
Depends on: 1164426

Updated

a year ago
Whiteboard: [reader-mode-readability-algorithm]

Updated

8 months ago
Depends on: 1242674

Updated

8 months ago
Depends on: 1167662

Updated

8 months ago
Depends on: 1260136

Comment 3

8 months ago
Just wanted to comment on this. The reader button seems to sometimes appear but only display part of the page when reader mode is activated. For example, the article found here: http://lolajournal.com/7/lola_montes.html appears to be perfect for reader mode. However, when it's activated, only three paragraphs of the article are displayed, the ones starting with "Although Lola's..." and ending with "capitalist industry." These paragraphs are in the middle of the page, not the first few or last few paragraphs. This behavior is definitely a bug.

Updated

8 months ago
Summary: Refine isProbablyReaderable algorithm → [meta] Refine isProbablyReaderable algorithm

Comment 4

8 months ago
(In reply to lord.of.the.flies.0 from comment #3)
> Just wanted to comment on this. The reader button seems to sometimes appear
> but only display part of the page when reader mode is activated. For
> example, the article found here: http://lolajournal.com/7/lola_montes.html
> appears to be perfect for reader mode. However, when it's activated, only
> three paragraphs of the article are displayed, the ones starting with
> "Although Lola's..." and ending with "capitalist industry." These paragraphs
> are in the middle of the page, not the first few or last few paragraphs.
> This behavior is definitely a bug.

Please file a separate bug for this - this bug is not the right venue to investigate this further.

Comment 5

8 months ago
(In reply to :Gijs (gone until 3 jan) from comment #4)
> (In reply to lord.of.the.flies.0 from comment #3)
> > Just wanted to comment on this. The reader button seems to sometimes appear
> > but only display part of the page when reader mode is activated. For
> > example, the article found here: http://lolajournal.com/7/lola_montes.html
> > appears to be perfect for reader mode. However, when it's activated, only
> > three paragraphs of the article are displayed, the ones starting with
> > "Although Lola's..." and ending with "capitalist industry." These paragraphs
> > are in the middle of the page, not the first few or last few paragraphs.
> > This behavior is definitely a bug.
> 
> Please file a separate bug for this - this bug is not the right venue to
> investigate this further.

Done, see 1325999. For some reason I do not see "Reader Mode" in the component field.
You need to log in before you can comment on or make changes to this bug.