Please ignore this message, this must be some cpan glitch (cpan croaked
about the checksum [1], but I agreed) — once I downloaded and installed
Marpa-R2-2.079_015.tar.gz manually (perl Build.pl, Build, Build test, Build
install), the above tests pass on both win32 and cygwin. Also, install by
cpan
I'd like opinions on adding a interface statement to the SLIF.
Currently, for backward compatibility reasons, some of the defaults are
not the best. I'd like to add a
interface 2
statement which, if present, will makes these the new defaults:
lexeme default = forgiving = 1
:default ::=
(1) while *::array* is ok for both lexemes and G1 rules, to let the user
see both the parse result and and how it is parsed, I'd suggest the
following defaults I've caught myself using frequently:
:default ::= action = [lhs,values]
lexeme default = action = [lhs,value] forgiving = 1
where *lhs
In my working version, the meta-grammar now runs LATM, with no issues.
This change probably won't make it into the next indexed release, but
will make it into the one after.
The difference between changing the default for the meta-grammar (and
therefore for the parsing of all SLIF DSL) and
On Sat, Feb 15, 2014 at 9:37 PM, Jeffrey Kegler
jeffreykeg...@jeffreykegler.com wrote:
In my working version, the meta-grammar now runs LATM, with no issues.
This change probably won't make it into the next indexed release, but will
make it into the one after.
I've added global forgiveness
I perhaps was too round-about in the way I stated it -- the next
developer's version will have an LATM meta-grammar for the SLIF. -- jeffrey
On 02/15/2014 12:08 PM, Ruslan Shvedov wrote:
On Sat, Feb 15, 2014 at 9:37 PM, Jeffrey Kegler
jeffreykeg...@jeffreykegler.com
Re (1) lhs name would be mildly difficult -- right now the arrays are
assembled in C/XS code (the process is blazingly fast), but that layer
does *not* know about symbol names. That could be changed, but it's
perhaps more like something for interface 3.
Re (2), I am not 100% sure I
Marpa::R2 V 2.079015
Counts: Tests: 542. Modules: 8. Passes: 8. Fails: 0
Duration: 1 minute and 37 seconds
--
You received this message because you are subscribed to the Google Groups
marpa parser group.
To unsubscribe from this group and stop receiving emails from it, send an email
to
(1) Why:
interface 2
and not:
interface = 2
Or even
:default = interface = 2
And yes, it's rather longer. But saying ':default' is a good thing (i.e. a
heads-up) for beginners...
But is this (new token) really an improvement over getting the user to
specify their defaults
Yes, good guess. I meant interface 2. -- jeffrey
On 02/15/2014 03:47 PM, Ron Savage wrote:
(7)
revision 2
I'm confused. Did you mean interface 2?
--
You received this message because you are subscribed to the Google Groups marpa
parser group.
To unsubscribe from this group and stop
My motivation came from my first attempts to write the documentation for
LATM. It means that the prospective user coming to the top-level
synopsis is plunged immediately into the SLIF's least perspicuous
syntax, and advanced issues. I could just tell the user to treat the
two lines you
This relates to one of the big points which I periodically try to make
about parsers and languages. I believe that when it comes to
programming languages we've brain-washed ourselves to like unreadable
syntax, because that's all traditional parsers could parse, and a line
with lots of
Yes, right. -- jeffrey
On 02/15/2014 03:47 PM, Ron Savage wrote:
(2) The current situation:
So:
interface = 1
(Implicitly) means:
lexeme default = forgiving = 0
Right?
--
You received this message because you are subscribed to the Google Groups marpa
parser group.
To unsubscribe from
There's an example of DSL using bless = ::lhs here
https://metacpan.org/pod/release/JKEGL/Marpa-R2-2.079_015/pod/Progress.pod#The-example.
And sample output here
https://metacpan.org/pod/release/JKEGL/Marpa-R2-2.079_015/pod/Progress.pod#Parse-value-at-error-location.
There aren't many
:default ::= action = ::array sound the best to me, because it is natural
when we read a grammar to expect as many values as there are RHS.
I quite liked to proposal of ron, although IMHO lhs better fits in
blessing, despite the fact that beginners will always certainly do
Data::Dumper(),
should read I quite liked to proposal of rns - sorry.
You can delete this message having read it -;
Le dimanche 16 février 2014 02:58:51 UTC+1, Durand Jean-Damien a écrit :
I quite liked to proposal of ron, ./..
--
You received this message because you are subscribed to the Google Groups
Re interface 2, I can go with something that has a bit more hype, like
dsl maker 2
-- jeffrey
On 02/15/2014 05:58 PM, Durand Jean-Damien wrote:
:default ::= action = ::array sound the best to me, because it is
natural when we read a grammar to expect as many values as there are RHS.
I
let's go with interface
Le dimanche 16 février 2014 03:35:22 UTC+1, Jeffrey Kegler a écrit :
Re interface 2, I can go with something that has a bit more hype,
like
dsl maker 2
-- jeffrey
On 02/15/2014 05:58 PM, Durand Jean-Damien wrote:
:default ::= action = ::array sound the
On Sun, Feb 16, 2014 at 12:42 AM, Jeffrey Kegler
jeffreykeg...@jeffreykegler.com wrote:
Re (1) lhs name would be mildly difficult -- right now the arrays are
assembled in C/XS code (the process is blazingly fast), but that layer does
*not* know about symbol names. That could be changed, but
A presentation by Rob Pike —
http://cuddle.googlecode.com/hg/talk/lex.html#landing-slide — basically,
template parser deliberately built from scratch — no regexp, lex, yacc,
etc. — made me wonder if he'd used marpa when he could.
--
You received this message because you are subscribed to the
20 matches
Mail list logo
