Re: listing currently defined macros

[Gary V. Vaughan]

Hi Eric,

On 19 Feb 2008, at 02:19, Eric Blake wrote:
> According to Gary V. Vaughan on 2/17/2008 2:07 AM:
>
> Hi Gary, and thanks for a review of my ideas, plus some cool ones of  
> your own.

What a team! Glad you like :-)

> | However (breaking my own rule about not introducing additional  
> macros if
> | at all possible), I think list sorting and filtering are both  
> useful enough
> | idioms that we should break them out into new builtins so that our  
> users
> | needn't resort to the hackish things you mention below... which  
> merely
> | highlight deficiencys in the current language implementation.  Also,
> | m4symbols proper will be faster in the case where the order of  
> returned
> | symbols is not important.  So:
> |
> |  -- Builtin (gnu): m4symbols([NAMES...])
> |      With no arguments, `m4symbols' expands to a unordered, comma- 
> separated
> |      list of the names of all currently defined macro.  Each macro  
> name is
> |      quoted.
>
> In other m4 implementations (I tested Solaris and OpenBSD), dumpdef  
> output
> is also unordered - it is only ordered on GNU.  But I would be okay  
> with a
> documented change to dumpdef behavior to be unordered, particularly  
> since
> we could then do:
>
> define(`dumpdef', `builtin(`dumpdef', sort(m4symbols))')
>
> to recover the former behavior.

That's interesting... yes, most definitely.

> |  -- Builtin (list): sort(NAME, [...])
> |      Expands to a sorted, comma-separated list of quoted NAME  
> arguments.
> |      This macro was added in M4 2.0.
>
> At which point, the user can easily create:
>
> define(`cmp', `ifelse(`$1', `$2', `0',
> ~                      (sort(`$1', `$2')), `($1,$2)', `-1',
> ~                      `1')')

Or, we can define it for them in a directory module of m4 with  
supporting
dsos :-)

> |  -- Builtin (list): uniq(NAME, [...])
> |      Expands to an unordered, comma-separated list of quoted NAME
> | arguments,
> |      with only the first occurrence of duplicate consecutive  
> parameters
> |      listed in the expansion.  This macro was added in M4 2.0.
> |
> |      uniq(m4symbols(`ifndef', `ifdef', `define', `define', `ifdef',
> | `undef'))
> |      =>ifdef,define,ifdef
>
> Don't you mean
> ~ =>ifndef,define,ifdef

No no. There is no ifndef builtin, so m4symbols expands to:

   `ifdef',`define',`define',`ifdef'

But (for ease of implementation) uniq only discards consecutive  
duplicates,
so the two ifdefs are both kept.

> |  -- Builtin (list): filter(PATTERN, LIST, [PATSYNTAX])
> |      Expands to an unordered, comma-separated list of quoted LIST  
> elements
> |      that match PATTERN, either according to the current RESYNTAX  
> (*note
> |      Changeresyntax::) or PATSYNTAX if passed.
> |
> |      define(`dquote', ``$@'')
> |      =>
> |      sort(filter(`*def*', dquote(m4symbols), `GLOB'))
> |      =>define,defn,dumpdef,ifdef,popdef,pushdef,undefine
>
> Hmm.  This would be the first builtin that takes a quoted list of
> elements.  Maybe it would be worth also making dquote builtin:
>
> - -- Builtin (list): dquote(ELEMENT, ...)
> ~  Expand to a quoted, comma-separated list of quoted list ELEMENTs.
>
> so that you could use dquote regardless of the current quoting  
> characters
> (in your example, it would only work with `').  Then again, dquote  
> is easy
> to implement efficiently without needing another builtin, unlike  
> filter
> (which can be implemented, but requires a foreach loop).

I agree.  Eventually we should try to be kind to users passing around  
quoted
lists, and having a quote agnostic dquote is a good plan.

Although, perhaps (as part of the plan to level the playing field  
between
C builtins and m4 macros) we should provide some means to write a not- 
yet-
expanded-quote character in the definition of a macro, which turns into
whatever is a valid quote at expansion time.  I haven't thought about  
this
much yet, but maybe a reverse syntaxtable lookup builtin:

 -- Builtin (gnu): syntax(SYNTAX-CATEGORY)
     Expand to a quoted, comma-separated list of quoted syntax table
     entries that have SYNTAX-CATEGORY.  SYNTAX-CATEGORY is one of the
     letters used to change syntax with the `changesyntax' builtin
     (*note Changesyntax::).

Ofcourse the problem with this idea is how we deal with unbalanced quote
characters.  Otherwise, I like the introspection it provides.

> | I'm sure we can implement several of the m4sugar style list  
> mangling macros
> | eventually, hence I've put these new builtins into a separate list  
> module.
>
> Sounds like a good idea to me.  I'll see if I can get some time to  
> write
> up such a module.

Excellent! :-D

> |> So maybe you could argue that we already have
> |> globbing if you can just translate the pattern correctly).
> |
> | I think this proposal has that covered.
>
> I like the idea of using 'GLOB' as another regex syntax, although it
> implies some intelligence in the underlying regex builtins (either we
> would do glob->regex translation ourselves, or we would have to decide
> between re_search vs. fnmatch).

I think the latter will better build on existing proven code, and can
probably be factored somewhat to prevent duplication among builtin
implementations.

Cheers,
        Gary
--
  ())_.              Email me: address@hidden
  ( '/           Read my blog: http://blog.azazil.net
  / )=         ...and my book: http://sources.redhat.com/autobook
`(_~)_

PGP.sig
Description: This is a digitally signed message part