On snippet failure, tab key is ignored

[Michael Henry]

When SuperTab is not present, UltiSnips suppresses the <tab> key on snippet expansion failures. This prevents the user from using the <tab> key for indentation.

A patch which corrects this behavior may be found here:
https://code.launchpad.net/~drmikehenry/ultisnips/tab-failure

The patch is also attached for convenience.

Note that the test.py file has an additional hack to retry tests up to twice before reporting a failure. This is required on my system to get the out-of-the-box tests to complete reliably. I've tried adjusting sleeptime in several ways, including doing time.sleep(sleeptime + 0.10) in the type() function to globally increase the sleep time, but I still get occasional errors and the tests run extremely slowly. So far, having a couple of retries has made the tests reliable for me. Since each test starts execution cleanly, it seems reasonable to me to allow any single passing run of a test to indicate overall success, given how strongly timing influences test success.

Note that the patch doesn't fix the problem for arbitrary mappings of g:UltiSnipsExpandTrigger; only the <tab> key is detected and passed through. This is because I don't know how to generically convert :map-style strings like "<tab>" into feedkeys()-style strings like "\<tab>" (with the backslash-escape). But given the importance of the <tab> key and the fact that <tab> is the default trigger, it seems worthy of a special-case solution.

Michael Henry

[SirVer (sirver)]

Hi. I fully agree; your patch was applied and you have been added to the contributor list. Keep em coming ;)

[Ovidiu Visan (visan-ovidiu)]

still not fixed, i am experiencing this behaviour

[SirVer (sirver)]

#3: This is a completely unhelpful comment. Please provide complete steps to reproduce your problem like key presses to do or a screencast that shows the problem. Also your installed plugins and vimrc.