Closed hannesm closed 5 years ago
Mmmh actually I'd rather not have that ~empty, neither in the API nor in the post-process of ocp-indent
's output (it sometimes outputs largish stuff, I'm not sure there might not be empty lines there). Can't we rather do without the trim
? I guess I put it there as a reflex.
I force-pushed another solution that strips empty results in complete_word
, and does not modify Txt.lines
.
I force-pushed another solution that strips empty results in
complete_word
,
As I said I rather not arbitrarily strip's ocp-indent
pretty-printed output. I have something in 04b55d7ece3f8a6f9e90993 tell me if it doesn't work.
Thanks for the analysis.
thanks daniel, your solution works fine for me as well.
this fixes #21:
the issue was mainly the
String.trim cs
inid_complete
(which ate too much, namely all the whitespaces and the tab of the last completion), but then I needed a way to skip the last empty line (by enhancingTxt.lines
with an optional namedempty
argument, similar toAstring.String.cuts
).