return of replace-match

return of replace-match

[Phillip Lord]

I've been trying out "replace-match". The elisp manual says....

 -- Function: replace-match replacement &optional fixedcase literal
          string subexp
     This function performs a replacement operation on a buffer or
     string.

     If you did the last search in a buffer, you should omit the STRING
     argument or specify `nil' for it, and make sure that the current
     buffer is the one in which you performed the last search.  Then
     this function edits the buffer, replacing the matched text with
     REPLACEMENT.  It leaves point at the end of the replacement text,
     and returns `t'.


However, if I eval the progn below (with bob where it says)

(progn
  (re-search-forward "bob")
  (replace-match "fred"))

bob



point gets moved to the end of "bob" (now "fred") but "nil" is returned.


The docstring doesn't say anything at all about return types (unless
operating over a string.

Am I doing something wrong?

Phil

Re: return of replace-match

[Michael Heerdegen]

phillip.lord@newcastle.ac.uk (Phillip Lord) writes:

> (progn
>   (re-search-forward "bob")
>   (replace-match "fred"))
>
> bob
>
> point gets moved to the end of "bob" (now "fred") but "nil" is returned.

> Am I doing something wrong?

I think the manual is just lying.  If nobody disagrees, please make a
bug report.


Regards,

Michael.

Re: return of replace-match

[Phillip Lord]

Michael Heerdegen <michael_heerdegen@web.de> writes:

> phillip.lord@newcastle.ac.uk (Phillip Lord) writes:
>
>> (progn
>>   (re-search-forward "bob")
>>   (replace-match "fred"))
>>
>> bob
>>
>> point gets moved to the end of "bob" (now "fred") but "nil" is returned.
>
>> Am I doing something wrong?
>
> I think the manual is just lying.  If nobody disagrees, please make a
> bug report.


Was my impression!

Phil

Re: return of replace-match

[Alex Kost]

Phillip Lord (2014-03-04 18:42 +0400) wrote:

> I've been trying out "replace-match". The elisp manual says....
>
>  -- Function: replace-match replacement &optional fixedcase literal
>           string subexp
>      This function performs a replacement operation on a buffer or
>      string.
>
>      If you did the last search in a buffer, you should omit the STRING
>      argument or specify `nil' for it, and make sure that the current
>      buffer is the one in which you performed the last search.  Then
>      this function edits the buffer, replacing the matched text with
>      REPLACEMENT.  It leaves point at the end of the replacement text,
>      and returns `t'.
>
>
> However, if I eval the progn below (with bob where it says)
>
> (progn
>   (re-search-forward "bob")
>   (replace-match "fred"))
>
> bob
>
>
>
> point gets moved to the end of "bob" (now "fred") but "nil" is returned.
>
>
> The docstring doesn't say anything at all about return types (unless
> operating over a string.
>
> Am I doing something wrong?
>
> Phil

It looks correct to me.  When the docstring tells nothing about the
returning value, it may be anything and you shouldn't rely on it.

`replace-match' returns a string when you operate on a string, not a
text in a buffer.

Re: return of replace-match

[Bastien]

Alex Kost <alezost@gmail.com> writes:

> It looks correct to me.  When the docstring tells nothing about the
> returning value, it may be anything and you shouldn't rely on it.

IIUC, the problem here is that the docstring says it will return `t'
and return `nil' instead.  I guess that's just a documentation bug.

-- 
 Bastien

Re: return of replace-match

[Phillip Lord]

Bastien <bzg@altern.org> writes:

> Alex Kost <alezost@gmail.com> writes:
>
>> It looks correct to me.  When the docstring tells nothing about the
>> returning value, it may be anything and you shouldn't rely on it.
>
> IIUC, the problem here is that the docstring says it will return `t'
> and return `nil' instead.  I guess that's just a documentation bug.

The docstring doesn't say anything. The elisp manual says "returns `t`".

I'll send a bug report.

Re: return of replace-match

[Nicolas Richard]

Bastien <bzg@altern.org> writes:
> Alex Kost <alezost@gmail.com> writes:
>> It looks correct to me.  When the docstring tells nothing about the
>> returning value, it may be anything and you shouldn't rely on it.
>
> IIUC, the problem here is that the docstring says it will return `t'
> and return `nil' instead.

s/docstring/info manual/
See (info "(elisp) Replacing Match")

> I guess that's just a documentation bug.

Agreed. "return Qnil" has been in the code forever as far as I can see
(which is what C-x v g can see).

-- 
Nico.