HIST to read-from-minibuffer

HIST to read-from-minibuffer

[Leo]

Hello,

I am trying to read a string with a hist variable. I want the default to
be the first element of the hist variable and the history to start from
the SECOND element.

I can seem to get the behaviour I want with '(hist . 1) but not '(hist .
2). However the doc string says:

 "positions are counted starting from 1 at the beginning of the list"

This looks contradictory to the behaviour I observed. Is this a mistake
on my understand or the doc string? Thanks.

Leo

Re: HIST to read-from-minibuffer

[Juri Linkov]

> I am trying to read a string with a hist variable. I want the default to
> be the first element of the hist variable and the history to start from
> the SECOND element.
>
> I can seem to get the behaviour I want with '(hist . 1) but not '(hist .
> 2). However the doc string says:
>
>  "positions are counted starting from 1 at the beginning of the list"
>
> This looks contradictory to the behaviour I observed. Is this a mistake
> on my understand or the doc string? Thanks.

Sorry, I don't understand what is contradictory?

(let ((hist '("a" "b" "c")))
  (read-from-minibuffer "Prompt: " nil nil nil '(hist . 2)))

M-p => "c"

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: HIST to read-from-minibuffer

[Leo]

On 2010-05-25 17:09 +0100, Juri Linkov wrote:
> Sorry, I don't understand what is contradictory?
>
> (let ((hist '("a" "b" "c")))
>   (read-from-minibuffer "Prompt: " nil nil nil '(hist . 2)))
>
> M-p => "c"

I think I don't understand this sentence:

"Positions are counted starting from 1 at the beginning of the list."

I thought HIST is a 1-based list, but it is still 0-based. Could you
make the doc-string a bit clearer? Thanks.

For example, 0 works just fine:

 (let ((hist '("a" "b" "c")))
   (read-from-minibuffer "Prompt: " nil nil nil '(hist . 0)))

M-p => "a"

Leo

Re: HIST to read-from-minibuffer

[Juri Linkov]

>> (let ((hist '("a" "b" "c")))
>>   (read-from-minibuffer "Prompt: " nil nil nil '(hist . 2)))
>>
>> M-p => "c"
>
> I think I don't understand this sentence:
>
> "Positions are counted starting from 1 at the beginning of the list."
>
> I thought HIST is a 1-based list, but it is still 0-based. Could you
> make the doc-string a bit clearer? Thanks.

It is 1-based.

> For example, 0 works just fine:
>
>  (let ((hist '("a" "b" "c")))
>    (read-from-minibuffer "Prompt: " nil nil nil '(hist . 0)))
>
> M-p => "a"

"a" is the 1st element of HIST.  M-p moves to the 1st history element here.

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: HIST to read-from-minibuffer

[Leo]

On 2010-05-25 18:54 +0100, Juri Linkov wrote:
>>> (let ((hist '("a" "b" "c")))
>>>   (read-from-minibuffer "Prompt: " nil nil nil '(hist . 2)))
>>>
>
>>> M-p => "c"
>>
>> I think I don't understand this sentence:
>>
>> "Positions are counted starting from 1 at the beginning of the list."
>>
>> I thought HIST is a 1-based list, but it is still 0-based. Could you
>> make the doc-string a bit clearer? Thanks.
>
> It is 1-based.
>
>> For example, 0 works just fine:
>>
>>  (let ((hist '("a" "b" "c")))
>>    (read-from-minibuffer "Prompt: " nil nil nil '(hist . 0)))
>>
>> M-p => "a"
>
> "a" is the 1st element of HIST.  M-p moves to the 1st history element here.

This bit is what confuses me:

(hist . 0) 
        ^
        +----- why '0' here to mean first element?

Leo

Re: HIST to read-from-minibuffer

[Leo]

On 2010-05-25 19:20 +0100, Leo wrote:
> This bit is what confuses me:
>
> (hist . 0) 
>         ^
>         +----- why '0' here to mean first element?

The doc-string says: positions are counting from '1' while '0' means the
first element. That '1' and '0' is rather confusing.

Leo

Re: HIST to read-from-minibuffer

[Juri Linkov]

>>>> (let ((hist '("a" "b" "c")))
>>>>   (read-from-minibuffer "Prompt: " nil nil nil '(hist . 2)))
>>>>
>>
>>>> M-p => "c"
>>>
>>> I think I don't understand this sentence:
>>>
>>> "Positions are counted starting from 1 at the beginning of the list."
>>>
>>> I thought HIST is a 1-based list, but it is still 0-based. Could you
>>> make the doc-string a bit clearer? Thanks.
>>
>> It is 1-based.
>>
>>> For example, 0 works just fine:
>>>
>>>  (let ((hist '("a" "b" "c")))
>>>    (read-from-minibuffer "Prompt: " nil nil nil '(hist . 0)))
>>>
>>> M-p => "a"
>>
>> "a" is the 1st element of HIST.  M-p moves to the 1st history element here.
>
> This bit is what confuses me:
>
> (hist . 0)
>         ^
>         +----- why '0' here to mean first element?

'(hist . 0) is the same as 'hist and means no history is used by default.
'(hist . 1) means the history is moved to the first history element.

Please pay attention to this text in the docstring of `read-from-minibuffer':

  For consistency, you should also specify that
  element of the history as the value of INITIAL-CONTENTS.

This means you have to insert the selected history element
as initial contents:

  (let ((hist '("a" "b" "c")))
    (read-from-minibuffer "Prompt: " (nth 1 hist) nil nil '(hist . 2)))

Here you can see the difference - the same history element is
first in INITIAL-CONTENTS, but second in HIST.

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: HIST to read-from-minibuffer

[Leo]

>> This bit is what confuses me:
>>
>> (hist . 0)
>>         ^
>>         +----- why '0' here to mean first element?
>
> '(hist . 0) is the same as 'hist and means no history is used by default.
> '(hist . 1) means the history is moved to the first history element.
>
> Please pay attention to this text in the docstring of `read-from-minibuffer':
>
>  For consistency, you should also specify that
>  element of the history as the value of INITIAL-CONTENTS.
>
> This means you have to insert the selected history element
> as initial contents:
>
>  (let ((hist '("a" "b" "c")))
>    (read-from-minibuffer "Prompt: " (nth 1 hist) nil nil '(hist . 2)))
>
> Here you can see the difference - the same history element is
> first in INITIAL-CONTENTS, but second in HIST.

The elisp info is very clear on explaining the HIST argument. The doc
string intends to say if you specify a cons cell you'd better have a
HISTPOS >= 1; but then it uses "Positions are counted starting from 1
at the beginning of the list". When I first looked at it, I thought
the HIST has a special index scheme than normal lists which it
doesn't. This is where I am confused. But if others understand it
fine, just ignore this.

Best,

Leo