1866910 - In reader view, certain pages will lose accented characters when switching tabs or reloading

Status: NEW

(Toolkit :: Reader Mode, defect, P3)

Description

[Cecília Lopes]

User Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:109.0) Gecko/20100101 Firefox/119.0

Steps to reproduce:

Open this page, it's the official government page for the constitution of Brazil: https://www.planalto.gov.br/CCIVIL_03/Constituicao/Constituicao.htm

Enter reader view.

Switch to a different tab or press F5.

(happens for me on Firefox for Windows, Linux and Android, including Nightly, on various devices. Conspicuously, once in a while it doesn't happen. Not sure why.)

Actual results:

Accented characters are replaced with

Example: the phrase "A República Federativa do Brasil, formada pela união indissolúvel dos Estados" becomes "A Repblica Federativa do Brasil, formada pela unio indissolvel dos Estados"

Check the images

Comment 1

[Cecília Lopes]

Attachment: Before reloading / switching tabs

Comment 2

[Cecília Lopes]

Attachment: After reloading / switching tabs

Comment 3

[BugBot [:suhaib / :marco/ :calixte]]

The Bugbug bot thinks this bug should belong to the 'Toolkit::Reader Mode' component, and is moving the bug to that component. Please correct in case you think the bot is wrong.

Comment 4

[:cmkm]

Reproducible every time for me on macOS Nightly when refreshing the page, but haven't reproduced when switching tabs.

Comment 5

[Alice0775 White]

I can reproduce the issue on Nightly121.0a1 Windows10.

Comment 6

[Alice0775 White]

Regression window:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=cc71a688509fa8a5174c9963b7af7f521d532854&tochange=d4e1b984239b7eac8f5b462dac354627a897b5a4

Suspect: Bug 1701828

Comment 7

[BugBot [:suhaib / :marco/ :calixte]]

:hsivonen, since you are the author of the regressor, bug 1701828, could you take a look?

For more information, please visit BugBot documentation.

Comment 8

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1701828

Comment 9

[Henri Sivonen (:hsivonen)]

I can't repro with tab switch that keeps the reader mode page alive. I can repro with reload.

It looks like the reader mode doesn't deal with the underlying page undergoing the encoding reload. The reader mode should not react to an encoding change Stop, but there isn't currently a flag to indicate that a stop happens ahead of an encoding reload.

Perhaps the Reader Mode should either take into account any page-triggered reload or somehow be attentive to the an encoding reload.