You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@asterixdb.apache.org by Steven Jacobs <sj...@ucr.edu> on 2017/12/07 20:17:01 UTC

Metadata.Dataset vs asterix.Dataset

Hi all,
Right now when we compile an AQL FLWOR vs an SQL SELECT we get two
different function signatures for "Dataset"

SQL calls "asterix.Dataset" while AQL calls "Metadata.Dataset"

Does anyone know why these are distinct?

To see an example, look at the rewrittenResult in
QueryTranslator.rewriteCompileQuery for the following:

  for $x in dataset TwitterUsers
  return $x;


vs

  select * from TwitterUsers;


Steven

Re: Metadata.Dataset vs asterix.Dataset

Posted by Steven Jacobs <sj...@ucr.edu>.
Got it. Thanks! Are we still deprecating AQL?
Steven

On Thu, Dec 7, 2017 at 2:38 PM, Dmitry Lychagin <
dmitry.lychagin@couchbase.com> wrote:

> Steven,
>
> I recently merged a change that fixes namespaces for all function
> signatures that SQLPP compiler generates to match namespaces defined in
> BuiltinFunctions and AlgebricksBuiltinFunctions (so ‘asterix’ /
> ‘algebricks’). I have not reviewed AQL code at that point, so it might
> still generate them in the ‘Metadata’ namespace.
>
> Thanks,
> -- Dmitry
>
>
> On 12/7/17, 12:17 PM, "Steven Jacobs" <sj...@ucr.edu> wrote:
>
>     Hi all,
>     Right now when we compile an AQL FLWOR vs an SQL SELECT we get two
>     different function signatures for "Dataset"
>
>     SQL calls "asterix.Dataset" while AQL calls "Metadata.Dataset"
>
>     Does anyone know why these are distinct?
>
>     To see an example, look at the rewrittenResult in
>     QueryTranslator.rewriteCompileQuery for the following:
>
>       for $x in dataset TwitterUsers
>       return $x;
>
>
>     vs
>
>       select * from TwitterUsers;
>
>
>     Steven
>
>
>

Re: Metadata.Dataset vs asterix.Dataset

Posted by Mike Carey <dt...@gmail.com>.
(And we should be phasing out AQL....)


On 12/7/17 2:38 PM, Dmitry Lychagin wrote:
> Steven,
>
> I recently merged a change that fixes namespaces for all function signatures that SQLPP compiler generates to match namespaces defined in BuiltinFunctions and AlgebricksBuiltinFunctions (so ‘asterix’ / ‘algebricks’). I have not reviewed AQL code at that point, so it might still generate them in the ‘Metadata’ namespace.
>
> Thanks,
> -- Dmitry
>   
>
> On 12/7/17, 12:17 PM, "Steven Jacobs" <sj...@ucr.edu> wrote:
>
>      Hi all,
>      Right now when we compile an AQL FLWOR vs an SQL SELECT we get two
>      different function signatures for "Dataset"
>      
>      SQL calls "asterix.Dataset" while AQL calls "Metadata.Dataset"
>      
>      Does anyone know why these are distinct?
>      
>      To see an example, look at the rewrittenResult in
>      QueryTranslator.rewriteCompileQuery for the following:
>      
>        for $x in dataset TwitterUsers
>        return $x;
>      
>      
>      vs
>      
>        select * from TwitterUsers;
>      
>      
>      Steven
>      
>

Re: Metadata.Dataset vs asterix.Dataset

Posted by Dmitry Lychagin <dm...@couchbase.com>.
Steven, 

I recently merged a change that fixes namespaces for all function signatures that SQLPP compiler generates to match namespaces defined in BuiltinFunctions and AlgebricksBuiltinFunctions (so ‘asterix’ / ‘algebricks’). I have not reviewed AQL code at that point, so it might still generate them in the ‘Metadata’ namespace. 

Thanks,
-- Dmitry
 

On 12/7/17, 12:17 PM, "Steven Jacobs" <sj...@ucr.edu> wrote:

    Hi all,
    Right now when we compile an AQL FLWOR vs an SQL SELECT we get two
    different function signatures for "Dataset"
    
    SQL calls "asterix.Dataset" while AQL calls "Metadata.Dataset"
    
    Does anyone know why these are distinct?
    
    To see an example, look at the rewrittenResult in
    QueryTranslator.rewriteCompileQuery for the following:
    
      for $x in dataset TwitterUsers
      return $x;
    
    
    vs
    
      select * from TwitterUsers;
    
    
    Steven