Rec/quoted quote #143
Rec/quoted quote #143
Conversation
julia-mode.el
Outdated
| "\\\\") | ||
| (or (not (any "\\'")) | ||
| (seq "\\" (or (repeat 1 9 (not (any "\\'"))) | ||
| (any "\\'")))) |
There was a problem hiding this comment.
Does this match e.g. '\n', \0, \x7f, \177? It doesn't seem like it would. Probably worth adding a test for those.
There was a problem hiding this comment.
They all fit into the "(1 to 9 (not (backslash or singlequote))) after backslash" clause. I could be more prescriptive, if you like.
There was a problem hiding this comment.
That makes sense. This is fine. If you wanted to, you could add a couple comments with more explanation of the different forms of char literals and how they're matched, but I don't think that should block merging this.
There was a problem hiding this comment.
Yes, explanations would be great if @recri has the time. Also, indentation of (seq ... looks a bit off.
| @@ -482,6 +482,18 @@ identity" | |||
| a = \"#\" # |> | |||
| identity")) | |||
|
|
|||
| (ert-deftest julia--test-indent-quoted-single-quote () | |||
There was a problem hiding this comment.
Please add julia--should-font-lock tests for examples from #142.
julia-mode-tests.el
Outdated
| @@ -620,6 +632,10 @@ identity")) | |||
| (julia--should-font-lock s1 4 nil) | |||
| (julia--should-font-lock s1 10 nil))) | |||
|
|
|||
| (ert-deftest julia--test-char-const-font-lock () | |||
| (dolist (c '("'\\''" "'\\\"'" "'\\\\'" "'\\010'" "'\\xfe'" "'\\uabcd'" "'\\Uabcdef01'" "'\\n'" "'\\alpha'" "'a'" "'z'")) | |||
| (julia--should-font-lock c 1 font-lock-string-face))) | |||
There was a problem hiding this comment.
Need to test that the string-face ends at the end of the char literal.
julia-mode.el
Outdated
| (or ; two alternatives | ||
| (not (any "\\'")) ; one character, not single quote or backslash | ||
| (seq "\\" ; sequence of a backslash followed by ... | ||
| (or ; four alternatives | ||
| (any "\\'\"?abfnrtv") ; single character escape | ||
| (repeat 1 3 (any "0-7")) ; octal escape | ||
| (seq 'x' (repeat 1 8 hex)) ; hex escape | ||
| (seq 'u' (repeat 1 4 hex)) ; unicode escape | ||
| (seq 'U' (repeat 1 8 hex))))) ; extended unicode escaple |
There was a problem hiding this comment.
| (or ; two alternatives | |
| (not (any "\\'")) ; one character, not single quote or backslash | |
| (seq "\\" ; sequence of a backslash followed by ... | |
| (or ; four alternatives | |
| (any "\\'\"?abfnrtv") ; single character escape | |
| (repeat 1 3 (any "0-7")) ; octal escape | |
| (seq 'x' (repeat 1 8 hex)) ; hex escape | |
| (seq 'u' (repeat 1 4 hex)) ; unicode escape | |
| (seq 'U' (repeat 1 8 hex))))) ; extended unicode escaple | |
| (or ; two alternatives | |
| (not (any "\\'")) ; one character, not single quote or backslash | |
| (seq "\\" ; sequence of a backslash followed by ... | |
| (or ; four alternatives | |
| (any "\\'\"?abfnrtv") ; single character escape | |
| (repeat 1 3 (any "0-7")) ; octal escape | |
| (seq 'x' (repeat 1 8 hex)) ; hex escape | |
| (seq 'u' (repeat 1 4 hex)) ; unicode escape | |
| (seq 'U' (repeat 1 8 hex))))) ; extended unicode escaple |
(suggested indentation)
|
Sorry, I seem to have crossed the git gods somehow. No, it was the emacs gods, running tests with out of date .elc. |
|
Thanks for all the work! LGTM, I will merge when @non-Jedi approves. |
julia-mode.el
Outdated
| (group "'")))) | ||
| (group "'") ; start single quote of character constant | ||
| (or ; two alternatives | ||
| (not (any "\\'")) ; one character, not single quote or backslash |
There was a problem hiding this comment.
''' is a valid Char constant for '\''. This doesn't match that does it?
There was a problem hiding this comment.
Amazing, that must have been the only quoted single quote that the original code recognized, and the only one that's not documented in the julia manual.
This last pull also tightens up the \x syntax, julia only permits two hex digits, C permits as many as you write though not guaranteeing what will happen to the excess..
I think this solves #142 by simply rewriting
julia-char-regex. The regex didn't actually recognize'\'' as a character constant as it was written. It could still be wrong. Added a test for my particular case.Sorry, I should have rebased or something to squash the intermediate commits.