Patrick Fong <patrickf3139@xxxxxxxxx> writes: > What did you do before the bug happened? (Steps to reproduce your issue) > touch "this is escaped" > touch "this is not escaped" > git add "this is escaped" > git status --short > > What did you expect to happen? (Expected behavior) > I expected that git status --short would string escape both "this is > escaped" and "this is not escaped" since they both contain spaces that > need escaping. > > What happened instead? (Actual behavior) > git status did not string escape "this is not escaped" but it does if > you add it to the index. It indeed is a disturbing inconsistency, and the code that shows the untracked paths should be fixed. I wonder what is done to ignored paths when the "--ignored" option is given. wt-status.c::wt_shortstatus_status() has a strange special case that says "whitespace alone does not usually deserve quoting, but we special case it and manually quote"; this discrepancy/special casing was introduced at dbfdc625 (status: Quote paths with spaces in short format, 2010-11-08). A fix would probably be to port the same fix to wt-status.c::wt_shortstatus_other(). In the longer term, quote.c::quote_path() may want to learn an optional flag to tell it to do the the special casing. Thanks.
