Bug 43800 - [KeyBindings] "Null" and strange characters in key sequence entry widget
Summary: [KeyBindings] "Null" and strange characters in key sequence entry widget
Status: RESOLVED FIXED
Alias: None
Product: Platform
Classification: Eclipse Project
Component: UI (show other bugs)
Version: 3.0   Edit
Hardware: All All
: P1 blocker (vote)
Target Milestone: 3.0 M4   Edit
Assignee: Chris McLaren CLA
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2003-09-27 17:16 EDT by Nick Edgar CLA
Modified: 2003-09-29 11:14 EDT (History)
1 user (show)

See Also:

Attachments
KeySupport (1.34 KB, patch)
2003-09-29 11:00 EDT, Douglas Pollock CLA 		no flags Details | Diff
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Nick Edgar CLA 2003-09-27 17:16:10 EDT 
latest from head as of Sept. 27 2003
on top of build I20030719

- launch target
- set key bindings into emacs mode
- open a text file
- click in a word
- Esc D
- has no effect (should delete next word)

- go to key bindings page
- Text Editing / Delete Next Word shows Esc D as one of the bindings
- type Esc D in the key sequence field
- you get: Esc Null
Comment 1 Nick Edgar CLA 2003-09-27 22:10:27 EDT 
Note that I was running with an older SWT (I20030719), so this may not be a 
problem in the latest.
Comment 2 Douglas Pollock CLA 2003-09-29 10:23:45 EDT 
Nick, I'm going to hijack this bug for my own purposes.  The specific bug you
mention does not occur (anymore).  However, in general, something has changed
recently that has caused rather spectacular breakage.

These are the things I've noticed:
+ Typing arrow keys into the entry widget inserts strange characters.
+ Typing "Ctrl+" inserts "Null".


Problems occur on a self-hosted workbench running the latest head of
platform-ui, but do not occur on I20030925.
Comment 3 Douglas Pollock CLA 2003-09-29 11:00:18 EDT 
Created attachment 6272 [details]
KeySupport

The mistake was a blind cast from int to char without checking bounds.	Leaving
the value as an int solves the problem.

Test suites all pass (except for the seg faults described in Bug 43312).
Comment 4 Douglas Pollock CLA 2003-09-29 11:01:36 EDT 
nick/chris: review/apply
Comment 5 Chris McLaren CLA 2003-09-29 11:14:24 EDT 
patch applied.