1094485 - Properly handle exceptions in readerWorker.js

Status: RESOLVED FIXED

(Firefox for Android Graveyard :: Reader View, defect)

Description

[:Margaret Leibovic]

Right now if there's a JS exception in readerWorker.js, we just eat it and never get back to Reader.js with a response. In the case of _downloadAndParseDocument, this means we'll never release a loaded <browser> from memory.

Comment 1

[:Margaret Leibovic]

Attachment: Properly handle exceptions in readerWorker.js

Adding an onerror handler will make sure that we can handle JS exceptions in the worker.

I looked through the promise chains, and a rejection here will be caught and logged upstream. However, I realized that our getArticle in _addTabToReadingList expects that null will be returned, not an exception thrown, if there's a problem getting the article, so I decided to make getArticle handle the exception itself.

Comment 2

[Brian Nicholson (:bnicholson)]

Comment on attachment 8517805 [details] [diff] [review]
Properly handle exceptions in readerWorker.js

Review of attachment 8517805 [details] [diff] [review]:
-----------------------------------------------------------------

Nice, thanks for fixing this!

Comment 3

[Dave Townsend [:mossop]]

https://hg.mozilla.org/integration/fx-team/rev/f2237c6455ab

Comment 4

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/f2237c6455ab