> Since we're talking about quoting functions, how likely do you think
> these patches are to break stuff?
>
> The first one doesn't even change the generated byte-code so I assume
> that's pretty safe.
Indeed, this one looks quite safe (the only risk I can think of is to
break some builds which treat compilation warnings as errors, tho).
> The second one is what you've suggested here but applied blindly to our
> entire code base.
Looks safe enough to me (see trivial comment below), but I'm known to be
a bit rash at times: I tend to think that software *should* be soft, so
if such an "obviously right" change breaks something, it still counts as
a success because it helped find a quirk in the code, which should be
ideally fixed or at least clearly labeled.
=== Stefan
> --- a/lisp/ob-sql.el
> +++ b/lisp/ob-sql.el
> @@ -414,13 +414,12 @@ org-babel-sql-expand-vars
> (let ((data-file (org-babel-temp-file "sql-data-")))
> (with-temp-file data-file
> (insert (orgtbl-to-csv
> - val (if sqlite
> - nil
> - '( :fmt (lambda (el)
> - (if (stringp el)
> - el
> - (format "%S" el)))
> - :with-special-rows nil)))))
> + val (unless sqlite
> + (list :fmt (lambda (el)
> + (if (stringp el)
> + el
> + (format "%S" el)))
> + :with-special-rows nil)))))
> data-file)
> (if (stringp val) val (format "%S" val))))
> body t t)))
FWIW, using
`( :fmt ,(lambda (el)
(if (stringp el)
el
(format "%S" el)))
:with-special-rows nil)))))
leads to slightly less consing because it shares the
(:with-special-rows nil)
tail.
