unofficial mirror of bug-gnu-emacs@gnu.org 
 help / color / mirror / code / Atom feed
* bug#14332: 24.3; Octave indentation bug
@ 2013-05-02  4:04 Leo Liu
  2013-05-05  5:29 ` Stefan Monnier
  0 siblings, 1 reply; 12+ messages in thread
From: Leo Liu @ 2013-05-02  4:04 UTC (permalink / raw)
  To: 14332

Indent the following lines in octave mode

if (true)
  a = (1 \(2-3));
else
  a = 0;
endif

gives me

if (true)
  a = (1 \(2-3));
                else
                  a = 0;
                endif


It seems treating \ as escape outside strings is incorrect. Is the
following fix is correct?


diff --git a/lisp/progmodes/octave.el b/lisp/progmodes/octave.el
index a58fdefb..1e2c2425 100644
--- a/lisp/progmodes/octave.el
+++ b/lisp/progmodes/octave.el
@@ -233,7 +233,7 @@ (defvar octave-mode-syntax-table
     (modify-syntax-entry ?& "."   table)
     (modify-syntax-entry ?| "."   table)
     (modify-syntax-entry ?! "."   table)
-    (modify-syntax-entry ?\\ "\\" table)
+    (modify-syntax-entry ?\\ "." table)
     (modify-syntax-entry ?\' "."  table)
     ;; Was "w" for abbrevs, but now that it's not necessary any more,
     (modify-syntax-entry ?\` "."  table)





^ permalink raw reply related	[flat|nested] 12+ messages in thread

* bug#14332: 24.3; Octave indentation bug
  2013-05-02  4:04 bug#14332: 24.3; Octave indentation bug Leo Liu
@ 2013-05-05  5:29 ` Stefan Monnier
  2013-05-05  5:35   ` Leo Liu
  0 siblings, 1 reply; 12+ messages in thread
From: Stefan Monnier @ 2013-05-05  5:29 UTC (permalink / raw)
  To: Leo Liu; +Cc: 14332

> It seems treating \ as escape outside strings is incorrect. Is the
> following fix is correct?
[..]
> -    (modify-syntax-entry ?\\ "\\" table)
> +    (modify-syntax-entry ?\\ "." table)

Wouldn't that mess up the case of

  a = "hello\"there";

?

        Stefan





^ permalink raw reply	[flat|nested] 12+ messages in thread

* bug#14332: 24.3; Octave indentation bug
  2013-05-05  5:29 ` Stefan Monnier
@ 2013-05-05  5:35   ` Leo Liu
  2013-05-05  7:17     ` Stefan Monnier
  0 siblings, 1 reply; 12+ messages in thread
From: Leo Liu @ 2013-05-05  5:35 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: 14332

On 2013-05-05 13:29 +0800, Stefan Monnier wrote:
> Wouldn't that mess up the case of
>
>   a = "hello\"there";

Yes it does. How to fix this bug properly?

Leo





^ permalink raw reply	[flat|nested] 12+ messages in thread

* bug#14332: 24.3; Octave indentation bug
  2013-05-05  5:35   ` Leo Liu
@ 2013-05-05  7:17     ` Stefan Monnier
  2013-05-05  7:28       ` Leo Liu
  0 siblings, 1 reply; 12+ messages in thread
From: Stefan Monnier @ 2013-05-05  7:17 UTC (permalink / raw)
  To: Leo Liu; +Cc: 14332

>> Wouldn't that mess up the case of
>> a = "hello\"there";
> Yes it does. How to fix this bug properly?

You'll probably need to add some syntax-propertize case.
What does "(1 \(2-3))" mean in Octave?  Where else can \ appear in
Octave code?


        Stefan





^ permalink raw reply	[flat|nested] 12+ messages in thread

* bug#14332: 24.3; Octave indentation bug
  2013-05-05  7:17     ` Stefan Monnier
@ 2013-05-05  7:28       ` Leo Liu
  2013-05-06  1:04         ` Stefan Monnier
  0 siblings, 1 reply; 12+ messages in thread
From: Leo Liu @ 2013-05-05  7:28 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: 14332

On 2013-05-05 15:17 +0800, Stefan Monnier wrote:
> You'll probably need to add some syntax-propertize case.
> What does "(1 \(2-3))" mean in Octave?  Where else can \ appear in
> Octave code?

A \ 1 ≡ inv(A) # one over matrix A, with some optimisation

if A is a scalar, A \ 1 = 1 / A.

so \ is an operator.

Leo





^ permalink raw reply	[flat|nested] 12+ messages in thread

* bug#14332: 24.3; Octave indentation bug
  2013-05-05  7:28       ` Leo Liu
@ 2013-05-06  1:04         ` Stefan Monnier
  2013-05-06  3:29           ` Leo Liu
  0 siblings, 1 reply; 12+ messages in thread
From: Stefan Monnier @ 2013-05-06  1:04 UTC (permalink / raw)
  To: Leo Liu; +Cc: 14332

>> You'll probably need to add some syntax-propertize case.
>> What does "(1 \(2-3))" mean in Octave?  Where else can \ appear in
>> Octave code?

> A \ 1 ≡ inv(A) # one over matrix A, with some optimisation

> if A is a scalar, A \ 1 = 1 / A.

> so \ is an operator.

Then your best bet is probably to put in the mode's syntax-table the
syntax corresponding to the most common use of \ (either as an escape
char in strings or as an operator), and then use syntax-propertize to
put the other syntax on the other case (where you can check (nth
3 (syntax-ppss (math-beginning 0))) to see if you're within a string).


        Stefan





^ permalink raw reply	[flat|nested] 12+ messages in thread

* bug#14332: 24.3; Octave indentation bug
  2013-05-06  1:04         ` Stefan Monnier
@ 2013-05-06  3:29           ` Leo Liu
  2013-05-06  8:26             ` Leo Liu
  0 siblings, 1 reply; 12+ messages in thread
From: Leo Liu @ 2013-05-06  3:29 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: 14332

[-- Attachment #1: Type: text/plain, Size: 1611 bytes --]

On 2013-05-06 09:04 +0800, Stefan Monnier wrote:
> Then your best bet is probably to put in the mode's syntax-table the
> syntax corresponding to the most common use of \ (either as an escape
> char in strings or as an operator), and then use syntax-propertize to
> put the other syntax on the other case (where you can check (nth
> 3 (syntax-ppss (math-beginning 0))) to see if you're within a string).

What I would like is to have \ mean "." unless it is in double-quoted
strings where it should be "\\". I have the following non-working patch.
Ideas?

diff --git a/lisp/progmodes/octave.el b/lisp/progmodes/octave.el
index 2e848600..d5a1af2b 100644
--- a/lisp/progmodes/octave.el
+++ b/lisp/progmodes/octave.el
@@ -137,6 +137,8 @@ (defun octave-syntax-propertize-function (start end)
   (goto-char start)
   (octave-syntax-propertize-sqs end)
   (funcall (syntax-propertize-rules
+            ("\\\\" (0 (when (eq (nth 3 (syntax-ppss (match-beginning 0))) ?\")
+                         "\\")))
             ;; Try to distinguish the string-quotes from the transpose-quotes.
             ("\\(?:^\\|[[({,; ]\\)\\('\\)"
              (1 (prog1 "\"'" (octave-syntax-propertize-sqs end)))))
@@ -242,7 +244,7 @@ (defvar octave-mode-syntax-table
     (modify-syntax-entry ?& "."   table)
     (modify-syntax-entry ?| "."   table)
     (modify-syntax-entry ?! "."   table)
-    (modify-syntax-entry ?\\ "\\" table)
+    (modify-syntax-entry ?\\ "."  table)
     (modify-syntax-entry ?\' "."  table)
     ;; Was "w" for abbrevs, but now that it's not necessary any more,
     (modify-syntax-entry ?\` "."  table)


[-- Attachment #2: oct_ind.m --]
[-- Type: text/plain, Size: 83 bytes --]

## -*-Octave-*-

"abc\"abc"

if (true)
  a = (1 \(2-3));
		else
		  a = 0;
		endif

^ permalink raw reply related	[flat|nested] 12+ messages in thread

* bug#14332: 24.3; Octave indentation bug
  2013-05-06  3:29           ` Leo Liu
@ 2013-05-06  8:26             ` Leo Liu
  2013-05-06 14:48               ` Stefan Monnier
  0 siblings, 1 reply; 12+ messages in thread
From: Leo Liu @ 2013-05-06  8:26 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: 14332

On 2013-05-06 11:29 +0800, Leo Liu wrote:
> What I would like is to have \ mean "." unless it is in double-quoted
> strings where it should be "\\".

This patch seems working. This is the first time I am using the
syntax-table property. Stefan, do you see any problem with the patch?

diff --git a/lisp/progmodes/octave.el b/lisp/progmodes/octave.el
index 2e848600..66fda302 100644
--- a/lisp/progmodes/octave.el
+++ b/lisp/progmodes/octave.el
@@ -134,6 +134,12 @@ (defvar octave-font-lock-keywords
   "Additional Octave expressions to highlight.")
 
 (defun octave-syntax-propertize-function (start end)
+  (funcall (syntax-propertize-rules
+            ("\\\\" (0 (when (eq (nth 3 (save-excursion
+                                          (syntax-ppss (match-beginning 0))))
+                                 ?\")
+                         (string-to-syntax "\\")))))
+           start end)
   (goto-char start)
   (octave-syntax-propertize-sqs end)
   (funcall (syntax-propertize-rules
@@ -242,7 +248,7 @@ (defvar octave-mode-syntax-table
     (modify-syntax-entry ?& "."   table)
     (modify-syntax-entry ?| "."   table)
     (modify-syntax-entry ?! "."   table)
-    (modify-syntax-entry ?\\ "\\" table)
+    (modify-syntax-entry ?\\ "."  table)
     (modify-syntax-entry ?\' "."  table)
     ;; Was "w" for abbrevs, but now that it's not necessary any more,
     (modify-syntax-entry ?\` "."  table)





^ permalink raw reply related	[flat|nested] 12+ messages in thread

* bug#14332: 24.3; Octave indentation bug
  2013-05-06  8:26             ` Leo Liu
@ 2013-05-06 14:48               ` Stefan Monnier
  2013-05-06 15:57                 ` Leo Liu
  0 siblings, 1 reply; 12+ messages in thread
From: Stefan Monnier @ 2013-05-06 14:48 UTC (permalink / raw)
  To: Leo Liu; +Cc: 14332

> +            ("\\\\" (0 (when (eq (nth 3 (save-excursion
> +                                          (syntax-ppss (match-beginning 0))))
> +                                 ?\")
> +                         (string-to-syntax "\\")))))

The rule looks OK, but why not put it inside the subsequent
`syntax-propertize-rules'?


        Stefan





^ permalink raw reply	[flat|nested] 12+ messages in thread

* bug#14332: 24.3; Octave indentation bug
  2013-05-06 14:48               ` Stefan Monnier
@ 2013-05-06 15:57                 ` Leo Liu
  2013-05-06 21:37                   ` Stefan Monnier
  0 siblings, 1 reply; 12+ messages in thread
From: Leo Liu @ 2013-05-06 15:57 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: 14332

On 2013-05-06 22:48 +0800, Stefan Monnier wrote:
> The rule looks OK, but why not put it inside the subsequent
> `syntax-propertize-rules'?

Wouldn't this (octave-syntax-propertize-sqs end) before the
syntax-propertize-rules form causes some \'s to be missed?

Leo





^ permalink raw reply	[flat|nested] 12+ messages in thread

* bug#14332: 24.3; Octave indentation bug
  2013-05-06 15:57                 ` Leo Liu
@ 2013-05-06 21:37                   ` Stefan Monnier
  2013-05-06 23:58                     ` Leo Liu
  0 siblings, 1 reply; 12+ messages in thread
From: Stefan Monnier @ 2013-05-06 21:37 UTC (permalink / raw)
  To: Leo Liu; +Cc: 14332

>> The rule looks OK, but why not put it inside the subsequent
>> `syntax-propertize-rules'?
> Wouldn't this (octave-syntax-propertize-sqs end) before the
> syntax-propertize-rules form causes some \'s to be missed?

No, octave-syntax-propertize-sqs just skips over the contents of
single-quoted strings and from what you say you only want to handle
backslash in double-quoted strings.


        Stefan





^ permalink raw reply	[flat|nested] 12+ messages in thread

* bug#14332: 24.3; Octave indentation bug
  2013-05-06 21:37                   ` Stefan Monnier
@ 2013-05-06 23:58                     ` Leo Liu
  0 siblings, 0 replies; 12+ messages in thread
From: Leo Liu @ 2013-05-06 23:58 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: 14332-done

Fixed in trunk.

On 2013-05-07 05:37 +0800, Stefan Monnier wrote:
> No, octave-syntax-propertize-sqs just skips over the contents of
> single-quoted strings and from what you say you only want to handle
> backslash in double-quoted strings.

Thank you, Stefan, for the help. Much appreciated.

Leo





^ permalink raw reply	[flat|nested] 12+ messages in thread

end of thread, other threads:[~2013-05-06 23:58 UTC | newest]

Thread overview: 12+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-05-02  4:04 bug#14332: 24.3; Octave indentation bug Leo Liu
2013-05-05  5:29 ` Stefan Monnier
2013-05-05  5:35   ` Leo Liu
2013-05-05  7:17     ` Stefan Monnier
2013-05-05  7:28       ` Leo Liu
2013-05-06  1:04         ` Stefan Monnier
2013-05-06  3:29           ` Leo Liu
2013-05-06  8:26             ` Leo Liu
2013-05-06 14:48               ` Stefan Monnier
2013-05-06 15:57                 ` Leo Liu
2013-05-06 21:37                   ` Stefan Monnier
2013-05-06 23:58                     ` Leo Liu

Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/emacs.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).