Typo: .substr in pmacros.texi is .substring in pmacros.scm

Typo: .substr in pmacros.texi is .substring in pmacros.scm

[Hans-Peter Nilsson]

Please consider applying this patch.  It would IMHO be very useful
if .sym-built symbols were .pmacro-expanded, as I initially thought.
For example, a .pmacro could apply .sym on an argument and create
references to other, specialized .pmacro:s.  Currently this can be
accomplished only with the apparently frowned-upon "(.eval
(pmacro-expand '(.sym arg -something)))".  If you want, I guess I
can elaborate.

2002-01-28  Hans-Peter Nilsson  <hp@axis.com>

	* doc/pmacros.texi: Fix .substr typo to .substring.
	Mention that .sym expansions are not further expanded.

Index: pmacros.texi
===================================================================
RCS file: /cvs/src/src/cgen/doc/pmacros.texi,v
retrieving revision 1.2
diff -p -c -r1.2 pmacros.texi
*** pmacros.texi	2001/03/19 03:36:24	1.2
--- pmacros.texi	2002/01/28 18:33:26
*************** Preprocessor macros provide a way of sim
*** 21,27 ****
  * Convert a number to a hex::     The @code{.hex} builtin
  * Convert a string to uppercase:: The @code{.upcase} builtin
  * Convert a string to lowercase:: The @code{.downcase} builtin
! * Getting part of a string::      The @code{.substr} builtin
  * List splicing::                 The @code{.splice} builtin
  * Number generation::             The @code{.iota} builtin
  * Mapping a macro over a list::   The @code{.map} builtin
--- 21,27 ----
  * Convert a number to a hex::     The @code{.hex} builtin
  * Convert a string to uppercase:: The @code{.upcase} builtin
  * Convert a string to lowercase:: The @code{.downcase} builtin
! * Getting part of a string::      The @code{.substring} builtin
  * List splicing::                 The @code{.splice} builtin
  * Number generation::             The @code{.iota} builtin
  * Mapping a macro over a list::   The @code{.map} builtin
*************** Acceptable arguments are symbols, string
*** 144,150 ****
  The result is a symbol with the arguments concatenated together.
  Numbers are converted to a string, base 10, and then to a symbol.
  The result must be a valid Scheme symbol with the additional restriction
! that the first character must be a letter.
  
  @node String concatenation
  @section String concatenation
--- 144,151 ----
  The result is a symbol with the arguments concatenated together.
  Numbers are converted to a string, base 10, and then to a symbol.
  The result must be a valid Scheme symbol with the additional restriction
! that the first character must be a letter.  The result is not further
! expanded, regardless of whether there's a macro by that name.
  
  @node String concatenation
  @section String concatenation
*************** Example:
*** 222,232 ****
  
  @node Getting part of a string
  @section Getting part of a string
! @cindex .substr
  
! Extract a part of a string with @code{.substr}.
  
! Syntax: @code{(.substr string start end)}
  
  where @samp{start} is the starting character, and @samp{end} is one past
  the ending character.  Character numbering begins at position 0.
--- 223,233 ----
  
  @node Getting part of a string
  @section Getting part of a string
! @cindex .substring
  
! Extract a part of a string with @code{.substring}.
  
! Syntax: @code{(.substring string start end)}
  
  where @samp{start} is the starting character, and @samp{end} is one past
  the ending character.  Character numbering begins at position 0.
*************** string is returned.
*** 236,242 ****
  Example:
  
  @smallexample
! (.substr "howzitgoineh?" 2 6) --> "wzit"
  @end smallexample
  
  @node List splicing
--- 237,243 ----
  Example:
  
  @smallexample
! (.substring "howzitgoineh?" 2 6) --> "wzit"
  @end smallexample
  
  @node List splicing

brgds, H-P

Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm

[Frank Ch. Eigler]

[-- Attachment #1: Type: text/plain, Size: 404 bytes --]

Hi -

On Mon, Jan 28, 2002 at 07:57:08PM +0100, Hans-Peter Nilsson wrote:
> Please consider applying this patch.  

Sure - feel free to commit it if you are able.


> It would IMHO be very useful if .sym-built symbols were
> .pmacro-expanded, as I initially thought.  [...]

I see this sort of recursive macro-expansion could be useful.
Does someone have an argument against it?


- FChE

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm

[Hans-Peter Nilsson]

> Date: Mon, 28 Jan 2002 14:05:29 -0500
> From: "Frank Ch. Eigler" <fche@redhat.com>

> On Mon, Jan 28, 2002 at 07:57:08PM +0100, Hans-Peter Nilsson wrote:
> > It would IMHO be very useful if .sym-built symbols were
> > .pmacro-expanded, as I initially thought.  [...]
> 
> I see this sort of recursive macro-expansion could be useful.
> Does someone have an argument against it?

I didn't see any argument against it, except for the comment in
the code.  This works for my simple command-line examples
including the one in the comment but beware of the scheme/CGEN
newbie.  (Is this an ok way to compose ChangeLog entries for
nested defines?)

Ok to commit?

2002-03-19  Hans-Peter Nilsson  <hp@axis.com>

	* pmacros.scm (-pmacro-expand,scan): If result is a symbol,
	call scan-symbol on it.

Index: pmacros.scm
===================================================================
RCS file: /cvs/src/src/cgen/pmacros.scm,v
retrieving revision 1.1.1.1
diff -c -p -r1.1.1.1 pmacros.scm
*** pmacros.scm	2000/07/28 04:11:52	1.1.1.1
--- pmacros.scm	2002/03/19 14:30:25
***************
*** 229,239 ****
  			((and (list? exp) (not (null? exp))) (scan-list exp))
  			; Not a symbol or expression, return unchanged.
  			(else exp))))
!       ; ??? We use to re-examine `result' to see if it was another pmacro
!       ; invocation.  This allowed doing things like ((.sym a b c) arg1 arg2)
!       ; where `abc' is a pmacro.  Scheme doesn't work this way, so it was
!       ; removed.  It can be put back should it ever be warranted.
!       result))
  
    (if -pmacro-trace?
        (begin
--- 229,239 ----
  			((and (list? exp) (not (null? exp))) (scan-list exp))
  			; Not a symbol or expression, return unchanged.
  			(else exp))))
!       ; Re-examining `result' to see if it is another pmacro invocation
!       ; allows doing things like ((.sym a b c) arg1 arg2)
!       ; where `abc' is a pmacro.  Scheme doesn't work this way, but then
!       ; this is CGEN.
!       (if (symbol? result) (scan-symbol result) result)))
  
    (if -pmacro-trace?
        (begin

brgds, H-P

Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm

[Frank Ch. Eigler]

[-- Attachment #1: Type: text/plain, Size: 509 bytes --]

Hi -

> > I see this sort of recursive macro-expansion could be useful.
> > Does someone have an argument against it?
> 
> I didn't see any argument against it [...]

Right.

> Ok to commit?

Sure, thanks.

> 2002-03-19  Hans-Peter Nilsson  <hp@axis.com>
> 
> 	* pmacros.scm (-pmacro-expand,scan): If result is a symbol,
> 	call scan-symbol on it.

The ChangeLog entry is of fine form, though I might add 

                         ...  , to enable recursive macro-expansion.


- FChE

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm

[Doug Evans]

Hans-Peter Nilsson writes:
 > > Date: Mon, 28 Jan 2002 14:05:29 -0500
 > > From: "Frank Ch. Eigler" <fche@redhat.com>
 > 
 > > On Mon, Jan 28, 2002 at 07:57:08PM +0100, Hans-Peter Nilsson wrote:
 > > > It would IMHO be very useful if .sym-built symbols were
 > > > .pmacro-expanded, as I initially thought.  [...]
 > > 
 > > I see this sort of recursive macro-expansion could be useful.
 > > Does someone have an argument against it?
 > 
 > I didn't see any argument against it, except for the comment in
 > the code.  This works for my simple command-line examples
 > including the one in the comment but beware of the scheme/CGEN
 > newbie.  (Is this an ok way to compose ChangeLog entries for
 > nested defines?)
 > 
 > !       ; ??? We use to re-examine `result' to see if it was another pmacro
 > !       ; invocation.  This allowed doing things like ((.sym a b c) arg1 arg2)
 > !       ; where `abc' is a pmacro.  Scheme doesn't work this way, so it was
 > !       ; removed.  It can be put back should it ever be warranted.

I'm apprehensive, but if people want to try this go ahead.

If people start getting into trouble because of this I hope they share their
experience.

Recursive expansion of pmacros (was: Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm)

[Hans-Peter Nilsson]

> From: Doug Evans <dje@transmeta.com>
> Date: Tue, 19 Mar 2002 08:14:02 -0800 (PST)

> Hans-Peter Nilsson writes:
>  > > Date: Mon, 28 Jan 2002 14:05:29 -0500
>  > > From: "Frank Ch. Eigler" <fche@redhat.com>
>  > 
>  > > On Mon, Jan 28, 2002 at 07:57:08PM +0100, Hans-Peter Nilsson wrote:
>  > > > It would IMHO be very useful if .sym-built symbols were
>  > > > .pmacro-expanded, as I initially thought.  [...]
>  > > 
>  > > I see this sort of recursive macro-expansion could be useful.
>  > > Does someone have an argument against it?
>  > 
>  > I didn't see any argument against it, except for the comment in
>  > the code.  This works for my simple command-line examples
>  > including the one in the comment but beware of the scheme/CGEN
>  > newbie.  (Is this an ok way to compose ChangeLog entries for
>  > nested defines?)
>  > 
>  > !       ; ??? We use to re-examine `result' to see if it was another pmacro
>  > !       ; invocation.  This allowed doing things like ((.sym a b c) arg1 arg2)
>  > !       ; where `abc' is a pmacro.  Scheme doesn't work this way, so it was
>  > !       ; removed.  It can be put back should it ever be warranted.
> 
> I'm apprehensive, but if people want to try this go ahead.

Hey, time has run out for protests! :-)

> If people start getting into trouble because of this I hope they share their
> experience.

And their joy!  But that would probably be a first. ;-)

> From: Doug Evans <dje@transmeta.com>
> Date: Tue, 19 Mar 2002 08:14:59 -0800 (PST)

> btw, do you have an actual use for it?
> Or is this just something that you think might be
> useful in the future?

As I wrote, I had already startet to depend upon it in a port in
progress.  I greatly simplified things (or made it hard in the
absence, I should say).

brgds, H-P

Recursive expansion of pmacros (was: Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm)

[Doug Evans]

Hans-Peter Nilsson writes:
 > As I wrote, I had already startet to depend upon it in a port in
 > progress.  I greatly simplified things (or made it hard in the
 > absence, I should say).

Note that there are otherways to do this, besides the implicit re-lookup.

Re: Recursive expansion of pmacros (was: Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm)

[Hans-Peter Nilsson]

> From: Doug Evans <dje@transmeta.com>
> Date: Tue, 19 Mar 2002 12:24:36 -0800 (PST)
> Cc: cgen@sources.redhat.com

> Hans-Peter Nilsson writes:
>  > As I wrote, I had already startet to depend upon it in a port in
>  > progress.  I greatly simplified things (or made it hard in the
>  > absence, I should say).
> 
> Note that there are otherways to do this, besides the implicit re-lookup.

I'm totally blank (what "this"?)  I generally want to map a
pmacro on a list defined as a pmacro, generating calls to other
pmacros.  How can I do that without re-lookup?

(define-pmacro x (a b c))
(define-pmacro (do-a y) (+ y 1))
(define-pmacro (do-b y) (- y 1))
(define-pmacro '(do-c y) (* y 2))
(pmacro-expand '(.map (.pmacro (z) ((.sym code- z) ((.sym do- z) w))) x))
=> ((code-a (+ w 1)) (code-b (- w 1)) (code-c (* w 2)))

Hmm, maybe that didn't come out clear.  Tell me if you feel you
need something better.

brgds, H-P

Re: Recursive expansion of pmacros (was: Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm)

[Doug Evans]

Hans-Peter Nilsson writes:
 > > Note that there are otherways to do this, besides the implicit re-lookup.
 > 
 > How can I do that without re-lookup?
                            ^
                            ^
You left out the "implicit".
e.g. what's the opposite of implicit? [hint hint]

Re: Recursive expansion of pmacros (was: Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm)

[Hans-Peter Nilsson]

> From: Doug Evans <dje@transmeta.com>
> Date: Tue, 19 Mar 2002 13:24:48 -0800 (PST)

> Hans-Peter Nilsson writes:
>  > > Note that there are otherways to do this, besides the implicit re-lookup.
>  > 
>  > How can I do that without re-lookup?
>                             ^
>                             ^
> You left out the "implicit".
> e.g. what's the opposite of implicit? [hint hint]

Ah, you mean I should use (.eval (pmacro-expand '(.sym ...)))!
But that's frowned upon (pmacros.scm) (supposedly because you
break through the CGEN description interface into scheme and
nasty things not accounted for can happen):

 ; .eval is an experiment.  Ports that consider themselves to be of beta
 ; quality or better don't use it.

Besides, I *expected* implicit re-lookup of a symbol built-up
from (.sym ...).  I think it comes naturally: what happens to
abc should happen to the expansion of (.sym a b c).  If people
don't want re-lookup, may there can be added explicit deferring,
by, say, (quote ...) ;-)

brgds, H-P

Re: Recursive expansion of pmacros (was: Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm)

[Doug Evans]

Hans-Peter Nilsson writes:
 > > You left out the "implicit".
 > > e.g. what's the opposite of implicit? [hint hint]
 > 
 > Ah, you mean I should use (.eval (pmacro-expand '(.sym ...)))!

Almost, but not quite.  I can see why that might be one's first guess.

 > But that's frowned upon (pmacros.scm) (supposedly because you
 > break through the CGEN description interface into scheme and
 > nasty things not accounted for can happen):
 > 
 >  ; .eval is an experiment.  Ports that consider themselves to be of beta
 >  ; quality or better don't use it.

.eval has connotations far beyond the problem at hand.

 > Besides, I *expected* implicit re-lookup of a symbol built-up
 > from (.sym ...).  I think it comes naturally: what happens to
 > abc should happen to the expansion of (.sym a b c).  If people
 > don't want re-lookup, may there can be added explicit deferring,
 > by, say, (quote ...) ;-)

Let's wait and see what, if any, kinds of problems this causes.

Re: Recursive expansion of pmacros (was: Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm)

[Hans-Peter Nilsson]

> From: Doug Evans <dje@transmeta.com>
> Date: Tue, 19 Mar 2002 14:47:07 -0800 (PST)

> Hans-Peter Nilsson writes:
>  > > You left out the "implicit".
>  > > e.g. what's the opposite of implicit? [hint hint]
>  > 
>  > Ah, you mean I should use (.eval (pmacro-expand '(.sym ...)))!
> 
> Almost, but not quite.  I can see why that might be one's first guess.

I'm too dense.  Please spell it out.

brgds, H-P

Re: Recursive expansion of pmacros (was: Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm)

[Doug Evans]

Hans-Peter Nilsson writes:
 > > From: Doug Evans <dje@transmeta.com>
 > > Date: Tue, 19 Mar 2002 14:47:07 -0800 (PST)
 > 
 > > Hans-Peter Nilsson writes:
 > >  > > You left out the "implicit".
 > >  > > e.g. what's the opposite of implicit? [hint hint]
 > >  > 
 > >  > Ah, you mean I should use (.eval (pmacro-expand '(.sym ...)))!
 > > 
 > > Almost, but not quite.  I can see why that might be one's first guess.
 > 
 > I'm too dense.  Please spell it out.

Later on I thought I said enough to indicate what I was getting at.
Sorry.

 > > .eval has connotations far beyond the problem at hand.

i.e. Don't call it .eval.  Call it something else.

Re: Typo: .substr in pmacros.texi is .substring in pmacros.scm

[Doug Evans]

Hans-Peter Nilsson writes:
 > I didn't see any argument against it, except for the comment in
 > the code.  This works for my simple command-line examples
 > including the one in the comment but beware of the scheme/CGEN
 > newbie.  (Is this an ok way to compose ChangeLog entries for
 > nested defines?)

btw, do you have an actual use for it?
Or is this just something that you think might be
useful in the future?