* lisp/emacs-lisp/cl-preloaded.el (cl-struct-define): Don't allow
structures with the same names as builtin types.
(cl--typeof-types, cl--all-builtin-types): Move from cl-generic.el and
rename.
(cl--struct-name-p): New helper function.
* lisp/emacs-lisp/cl-macs.el (cl-defstruct): Don't allow structures
with the same names as builtin types.
* lisp/emacs-lisp/cl-generic.el (cl--generic-typeof-generalizer)
(cl-generic-generalizers): Adapt to name change.
* test/lisp/emacs-lisp/cl-macs-tests.el (cl-defstruct/builtin-type):
* test/lisp/emacs-lisp/cl-preloaded-tests.el
(cl-struct-define/builtin-type): New unit tests.
* etc/NEWS: Document changed behavior.
Consider the expansion of `cl-loop' with a `for' clause and more
than one internal variables, X, Y, processed in parallel.
Each step updates X and Y right after update the loop variable, K; if
either X or Y depend on K, then some forms of the body are
evaluated with the wrong K (Bug#29799).
For instance, consider the following code:
(cl-loop for k below 2
for x = (progn (message "k = %d" k) 1)
and y = 1)
This code should show in *Messages*:
k = 0
k = 1
Instead, the code shows:
k = 0
k = 1
k = 2
To prevent this we must ensure that the loop condition is still
satisfied right after update the loop variable.
In the macro expansion of the example above, right after:
(setq k (+ k 1))
evaluate the rest of the body forms iif the condition
(< k 2)
is still valid.
* lisp/emacs-lisp/cl-macs.el (cl--loop-guard-cond): New variable.
(cl--parse-loop-clause): Set it non-nil if the loop contains
a for/as clause.
(cl-loop): After update the loop variable, evaluate the remaining of
the body forms just if the loop condition is still valid (Bug#29799).
* test/lisp/emacs-lisp/cl-macs-tests.el (cl-macs-loop-for-as-equals-and):
New test.
* lisp/emacs-lisp/cl-preloaded.el (cl-struct-define): Don't allow
structures with the same names as builtin types.
(cl--struct-name-p): New helper function.
* lisp/emacs-lisp/cl-macs.el (cl-defstruct): Don't allow structures
with the same names as builtin types.
* test/lisp/emacs-lisp/cl-macs-tests.el (cl-defstruct/builtin-type):
* test/lisp/emacs-lisp/cl-preloaded-tests.el
(cl-struct-define/builtin-type): New unit tests.
* etc/NEWS: Document changed behavior.
The changes to cl argument parsing are not backwards compatible, and
cause inconvenience when writing macros (e.g., instead of doing '&aux
,@auxargs', some more complicated conditionals would be required).
The `cl-defstruct' macro makes use of this convenience when defining
empty structs (Bug#29728).
* lisp/emacs-lisp/cl-macs.el (cl--transform-lambda):
(cl--do-&aux, cl--do-arglist): Undo strict checking of &rest, &key,
and &aux.
* test/lisp/emacs-lisp/cl-macs-tests.el (cl-macs-bad-arglist): Remove
test.
* etc/NEWS: Explain that '&optional' not followed by a variable is now
an error.
* lisp/emacs-lisp/cl-macs.el (cl--transform-lambda, cl--do-&aux)
(cl--do-arglist): Also reject '&optional', '&rest', or '&aux' not
followed by a variable for consistency.
* test/lisp/emacs-lisp/cl-macs-tests.el (cl-macs-bad-arglist): New
test.
Most of this change is to boilerplate commentary such as license URLs.
This change was prompted by ftp://ftp.gnu.org's going-away party,
planned for November. Change these FTP URLs to https://ftp.gnu.org
instead. Make similar changes for URLs to other organizations moving
away from FTP. Also, change HTTP to HTTPS for URLs to gnu.org and
fsf.org when this works, as this will further help defend against
man-in-the-middle attacks (for this part I omitted the MS-DOS and
MS-Windows sources and the test tarballs to keep the workload down).
HTTPS is not fully working to lists.gnu.org so I left those URLs alone
for now.
