[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Error-Free Navigation in Trees of Sub-Expressions (sexp) in Font-Lockin
From: |
Nordlöw |
Subject: |
Error-Free Navigation in Trees of Sub-Expressions (sexp) in Font-Locking Function Matchers |
Date: |
Wed, 8 Oct 2008 05:24:16 -0700 (PDT) |
User-agent: |
G2/1.0 |
How can I programmatically iterate over all sexps in an emacs-lisp
expression without risking getting an error? I have tried forward-
sexp() and backward-sexp() and up/down-list() but these generate an
error when I get to the beginning or end in each "direction". Why, on
earth, doesn't these functions have an optional argument, say no-
error, that inhibits errors in these cases and instead indicates this
event by returning nil?
I believe I need these functions because I am currently extending font-
locking in emacs-lisp-mode to highlight variables in let and defun-
like statements.
Here is my moccup so far:
(defun pnw-fancy/setq-args-matcher (limit)
(let ((start (point))) ; remember beginning
(if (looking-at (concat "[[:blank:]\n]*" "'?\\(\\w+\\)")) ; One
more SYM?
(progn (goto-char (match-end 1))
(forward-sexp); skip VAL
t)); signal hit
))
(defun pnw-fancy/emacs-lisp-variables-font-locking ()
(font-lock-add-keywords
nil
(list
;; setq-statements: (setq SYM VAL ...)
`(;; MATCHER: (SETQ
,(concat "(" (regexp-opt '("set"
"setq"
"setq-default"
"setq-mode-local") t)
"[[:blank:]\n]*")
;; SUBEXP-HIGHLIGHTER
(1 'font-lock-function-call-face keep)
;; ANCHORED-HIGHLIGHTER: (setq SYM VAL ...)
(;; ANCHORED-MATCHER
pnw-fancy/setq-args-matcher
nil ; PRE-FORM
nil ; POST-FORM
(1 'font-lock-variable-name-face prepend) ; SUBEXP-HIGHLIGHTERS
))
) t))
(add-hook 'emacs-lisp-mode-hook 'pnw-fancy/emacs-lisp-variables-font-
locking)
Test on these examples:
(setq a 1 b 2 c 3)
(setq a 1
b 2
c 3)
(setq x-1 '(1 1) x-2 '(2 2) x-3 '(3 3))
It currently doesn't work for multi-line variants of setq-form. Is
this beyond the capabilities of font-lock?
As I have said above it siliently errors when the last VAL is missing
like in the following statement:
(setq a 1 b 2 c)
What is the most clever way of checking when we are at the last sexp
(forward-sexp will fail as mentioned above)?
Thanks in advance,
Nordlöw
- Error-Free Navigation in Trees of Sub-Expressions (sexp) in Font-Locking Function Matchers,
Nordlöw <=