Indexing Status

[Preston Carman <pr...@apache.org>]

Hi Steven,

What is the status of the indexing code merge? Also, could you describe the
set up? As I understand it, a new expression (such as "create-index")
imports a collection of XML documents into the index. The index has key
value pairs based on the full path (key) to a element node (value) for each
XML document. Is this correct? How are the key and value stored? Are these
strings or XDM instances? Once the index is created, you can use a new
collection function based on the index (such as "collection-using-index").
Please fill in the gaps.

Aloha,
Preston

Re: Indexing Status

[Steven Jacobs <sj...@ucr.edu>]

Fixed! Something about my setup was making the vxq script decide to choose
a new JAVA_HOME, which was pointing to 1.6

Steven

On Thu, Sep 10, 2015 at 12:03 PM, Till Westmann <ti...@apache.org> wrote:

> Can’t see any output … what format should it be in?
>
> > On Sep 10, 2015, at 11:55 AM, Steven Jacobs <sj...@ucr.edu> wrote:
> >
> > Here is a full output of my problem, in case anyone has new advice:
> >
> > ​
> >
> > On Thu, Sep 10, 2015 at 11:09 AM, Steven Jacobs <sjaco002@ucr.edu
> <ma...@ucr.edu>> wrote:
> > Unfortunately my big blocker is that VXQuery will not run on my machine.
> It seems that no matter what I try, Java still disagrees with me. I have
> merged my code with current master, but I can't test it. I also didn't
> document it well originally, so getting it running is a bit of a precursor
> to getting good documentation on how exactly it works.
> > Steven
> >
> > On Thu, Sep 10, 2015 at 10:41 AM, Till Westmann <tillw@apache.org
> <ma...@apache.org>> wrote:
> > Yes, some description would be nice. Reading this e-mail I’m
> specifically interested in the lifecycle of the data and the indexes.
> > So far all lifecycle management happens outside of the query processor
> and the query processor cannot handle side-effecting operations (like
> "create-index”). So I’m wondering how we should do that.
> >
> > Cheers,
> > Till
> >
> > > On Sep 10, 2015, at 10:11 AM, Preston Carman <prestonc@apache.org
> <ma...@apache.org>> wrote:
> > >
> > > Hi Steven,
> > >
> > > What is the status of the indexing code merge? Also, could you
> describe the
> > > set up? As I understand it, a new expression (such as "create-index")
> > > imports a collection of XML documents into the index. The index has key
> > > value pairs based on the full path (key) to a element node (value) for
> each
> > > XML document. Is this correct? How are the key and value stored? Are
> these
> > > strings or XDM instances? Once the index is created, you can use a new
> > > collection function based on the index (such as
> "collection-using-index").
> > > Please fill in the gaps.
> > >
> > > Aloha,
> > > Preston
> >
> >
> >
>
>

Re: Indexing Status

[Till Westmann <ti...@apache.org>]

Can’t see any output … what format should it be in?

> On Sep 10, 2015, at 11:55 AM, Steven Jacobs <sj...@ucr.edu> wrote:
> 
> Here is a full output of my problem, in case anyone has new advice: 
> 
> ​
> 
> On Thu, Sep 10, 2015 at 11:09 AM, Steven Jacobs <sjaco002@ucr.edu <ma...@ucr.edu>> wrote:
> Unfortunately my big blocker is that VXQuery will not run on my machine. It seems that no matter what I try, Java still disagrees with me. I have merged my code with current master, but I can't test it. I also didn't document it well originally, so getting it running is a bit of a precursor to getting good documentation on how exactly it works.
> Steven
> 
> On Thu, Sep 10, 2015 at 10:41 AM, Till Westmann <tillw@apache.org <ma...@apache.org>> wrote:
> Yes, some description would be nice. Reading this e-mail I’m specifically interested in the lifecycle of the data and the indexes.
> So far all lifecycle management happens outside of the query processor and the query processor cannot handle side-effecting operations (like "create-index”). So I’m wondering how we should do that.
> 
> Cheers,
> Till
> 
> > On Sep 10, 2015, at 10:11 AM, Preston Carman <prestonc@apache.org <ma...@apache.org>> wrote:
> >
> > Hi Steven,
> >
> > What is the status of the indexing code merge? Also, could you describe the
> > set up? As I understand it, a new expression (such as "create-index")
> > imports a collection of XML documents into the index. The index has key
> > value pairs based on the full path (key) to a element node (value) for each
> > XML document. Is this correct? How are the key and value stored? Are these
> > strings or XDM instances? Once the index is created, you can use a new
> > collection function based on the index (such as "collection-using-index").
> > Please fill in the gaps.
> >
> > Aloha,
> > Preston
> 
> 
> 

Re: Indexing Status

[Steven Jacobs <sj...@ucr.edu>]

Here is a full output of my problem, in case anyone has new advice:

​

On Thu, Sep 10, 2015 at 11:09 AM, Steven Jacobs <sj...@ucr.edu> wrote:

> Unfortunately my big blocker is that VXQuery will not run on my machine.
> It seems that no matter what I try, Java still disagrees with me. I have
> merged my code with current master, but I can't test it. I also didn't
> document it well originally, so getting it running is a bit of a precursor
> to getting good documentation on how exactly it works.
> Steven
>
> On Thu, Sep 10, 2015 at 10:41 AM, Till Westmann <ti...@apache.org> wrote:
>
>> Yes, some description would be nice. Reading this e-mail I’m specifically
>> interested in the lifecycle of the data and the indexes.
>> So far all lifecycle management happens outside of the query processor
>> and the query processor cannot handle side-effecting operations (like
>> "create-index”). So I’m wondering how we should do that.
>>
>> Cheers,
>> Till
>>
>> > On Sep 10, 2015, at 10:11 AM, Preston Carman <pr...@apache.org>
>> wrote:
>> >
>> > Hi Steven,
>> >
>> > What is the status of the indexing code merge? Also, could you describe
>> the
>> > set up? As I understand it, a new expression (such as "create-index")
>> > imports a collection of XML documents into the index. The index has key
>> > value pairs based on the full path (key) to a element node (value) for
>> each
>> > XML document. Is this correct? How are the key and value stored? Are
>> these
>> > strings or XDM instances? Once the index is created, you can use a new
>> > collection function based on the index (such as
>> "collection-using-index").
>> > Please fill in the gaps.
>> >
>> > Aloha,
>> > Preston
>>
>>
>

Re: Indexing Status

[Steven Jacobs <sj...@ucr.edu>]

Unfortunately my big blocker is that VXQuery will not run on my machine. It
seems that no matter what I try, Java still disagrees with me. I have
merged my code with current master, but I can't test it. I also didn't
document it well originally, so getting it running is a bit of a precursor
to getting good documentation on how exactly it works.
Steven

On Thu, Sep 10, 2015 at 10:41 AM, Till Westmann <ti...@apache.org> wrote:

> Yes, some description would be nice. Reading this e-mail I’m specifically
> interested in the lifecycle of the data and the indexes.
> So far all lifecycle management happens outside of the query processor and
> the query processor cannot handle side-effecting operations (like
> "create-index”). So I’m wondering how we should do that.
>
> Cheers,
> Till
>
> > On Sep 10, 2015, at 10:11 AM, Preston Carman <pr...@apache.org>
> wrote:
> >
> > Hi Steven,
> >
> > What is the status of the indexing code merge? Also, could you describe
> the
> > set up? As I understand it, a new expression (such as "create-index")
> > imports a collection of XML documents into the index. The index has key
> > value pairs based on the full path (key) to a element node (value) for
> each
> > XML document. Is this correct? How are the key and value stored? Are
> these
> > strings or XDM instances? Once the index is created, you can use a new
> > collection function based on the index (such as
> "collection-using-index").
> > Please fill in the gaps.
> >
> > Aloha,
> > Preston
>
>

Re: Indexing Status

[Till Westmann <ti...@apache.org>]

Yes, some description would be nice. Reading this e-mail I’m specifically interested in the lifecycle of the data and the indexes.
So far all lifecycle management happens outside of the query processor and the query processor cannot handle side-effecting operations (like "create-index”). So I’m wondering how we should do that.

Cheers,
Till

> On Sep 10, 2015, at 10:11 AM, Preston Carman <pr...@apache.org> wrote:
> 
> Hi Steven,
> 
> What is the status of the indexing code merge? Also, could you describe the
> set up? As I understand it, a new expression (such as "create-index")
> imports a collection of XML documents into the index. The index has key
> value pairs based on the full path (key) to a element node (value) for each
> XML document. Is this correct? How are the key and value stored? Are these
> strings or XDM instances? Once the index is created, you can use a new
> collection function based on the index (such as "collection-using-index").
> Please fill in the gaps.
> 
> Aloha,
> Preston