Hunspell 1.4.0 had a couple bugs that led to a 1.4.1 release shortly after we'd landed the 1.4.0 update. I'd like to update to 1.4.1 before uplifting it all around.
Created attachment 8748447 [details] [diff] [review] update Hunspell to version 1.4.1 Aryeh, I hear you're getting more involved in Editor-related things these days. Can I interest you in getting your feet wet with a small Hunspell update? https://treeherder.mozilla.org/#/jobs?repo=try&revision=2e7fd29e08bf
Comment on attachment 8748447 [details] [diff] [review] update Hunspell to version 1.4.1 Except I was looking at the wrong Try push at the time. *sigh* The correct Try push for this update passes tests just fine, I swear!
Comment on attachment 8748447 [details] [diff] [review] update Hunspell to version 1.4.1 I'm not really sure I'm qualified to review this, sorry.
Attachment #8748447 - Flags: review?(ehsan)
For the purposes of this test I have used the Bugzilla textbox, the one from GitHub, the Facebook “Post” textbox, the Twitter “Tweet” textbox, the text box from wordcounter.net and a basic one created in HTML. Tested on the following platforms: Windows 10 x64 Ubuntu 14.04 x32 Mac OS X 10.11 The following dictionaries were used during testing: English (US) English (UK) Romanian Hindi Spanish Conclusions: I haven't observed any odd behavior after the Hunspell update. All text boxes behave the same way on the tested platforms. Worked correctly for most of the cases with respect to correctly spelled and misspelled words. However, Romanian, Spanish and Hindi words using special characters are not correctly flagged by the spell checker (English US dictionary) as incorrect. Whenever a number or special character is appended to any word, spell check ignores it and it is not flagged. This is contrary to behavior seen in Chrome browser. For the first issue I will log a separate bug, but I'm not sure if the second one is a valid one. What are your thoughts on this?
Thanks for all the testing! (In reply to Ciprian Muresan [:cmuresan] from comment #5) > Whenever a number or special character is appended to any word, spell check > ignores it and it is not flagged. Certainly sounds like a bug to me.
Status: NEW → RESOLVED
Last Resolved: 2 years ago
status-firefox49: affected → fixed
Resolution: --- → FIXED
Target Milestone: --- → mozilla49
Created attachment 8753977 [details] [diff] [review] roll-up patch for uplift Approval Request Comment [Feature/regressing bug #]: n/a [User impact if declined]: numerous issues found by Coverity and fuzzing (see bug 1257902 deps) [Describe test coverage new/current, TreeHerder]: v1.4.0 has been on trunk for over two weeks now. The v1.4.1 update has baked for over a week now without issue. Also had a QA pass with multiple languages and no new issues found. We also have automated spellchecker tests in CI. No obvious crash spikes on crash-stats that I can see. [Risks and why]: Fuzzing hasn't turned up any new issues and no known regressions in 2+ weeks on trunk makes me feel pretty confident about this being low risk. [String/UUID change made/needed]: none
Comment on attachment 8753977 [details] [diff] [review] roll-up patch for uplift Spellchecker update after almost a year (incl. sec fixes from those guys), no regressions found in 2 weeks of baking on central. Aurora48+, Beta47+, ESR45+
Attachment #8753977 - Flags: approval-mozilla-esr45?
Attachment #8753977 - Flags: approval-mozilla-esr45+
Attachment #8753977 - Flags: approval-mozilla-beta?
Attachment #8753977 - Flags: approval-mozilla-beta+
Attachment #8753977 - Flags: approval-mozilla-aurora?
Attachment #8753977 - Flags: approval-mozilla-aurora+
status-firefox47: --- → affected
status-firefox48: --- → affected
status-firefox-esr45: --- → affected
status-firefox48: affected → fixed
status-firefox47: affected → fixed
status-firefox-esr45: affected → fixed
You need to log in before you can comment on or make changes to this bug.