New option for Nexus indexer CLI: unpack

[Alin Dreghiciu]

Hi,

I'm using a tool named Luke to explore Lucene indexes created by Nexus indexer. But this tool are not able to work with the new indexing format (nexus-maven-repository-index.gz) as it is. So, I finded useful to implement a new option in CLI to allow unpacking of a known Lucene index directory format.

To use it you do:

<cli> -u - if the nexus-maven-repository-index.gz is in the current directory and you want the unpacked index to be placed in the current directory 

or you can add -d to specify the directory and/or -i to specify the directory containing the index file.

If you find it useful, let me know so I can create a jira issue for it and attach a patch (or commit ;)

--
Alin Dreghiciu

[Tamás Cservenák]

+1 sounds fine with me ;)

~t~

On Mon, Jul 20, 2009 at 8:07 PM, Alin Dreghiciu <adreghiciu@...> wrote:

Hi,

I'm using a tool named Luke to explore Lucene indexes created by Nexus indexer. But this tool are not able to work with the new indexing format (nexus-maven-repository-index.gz) as it is. So, I finded useful to implement a new option in CLI to allow unpacking of a known Lucene index directory format.

To use it you do:

<cli> -u - if the nexus-maven-repository-index.gz is in the current directory and you want the unpacked index to be placed in the current directory 

or you can add -d to specify the directory and/or -i to specify the directory containing the index file.

If you find it useful, let me know so I can create a jira issue for it and attach a patch (or commit ;)

--
Alin Dreghiciu

[Brian Fox]

Alin, the new .gz is a proprietary binary format; that is it's
strictly not lucene. This was intentional as lucene doesn't preserve
backwards compatability and basically locked us into a single version
of 2.3 if we wanted other clients to be able to read the index we
transport. The new format is read in and imported internally into a
lucene index, which could be 2.3 or 2.4 etc.

2009/7/20 Tamás Cservenák <tamas@...>:

> +1 sounds fine with me ;)
> ~t~
>
> On Mon, Jul 20, 2009 at 8:07 PM, Alin Dreghiciu <adreghiciu@...> wrote:
>>
>> Hi,
>> I'm using a tool named Luke to explore Lucene indexes created by Nexus indexer. But this tool are not able to work with the new indexing format (nexus-maven-repository-index.gz) as it is. So, I finded useful to implement a new option in CLI to allow unpacking of a known Lucene index directory format.
>> To use it you do:
>> <cli> -u - if the nexus-maven-repository-index.gz is in the current directory and you want the unpacked index to be placed in the current directory
>> or you can add -d to specify the directory and/or -i to specify the directory containing the index file.
>> If you find it useful, let me know so I can create a jira issue for it and attach a patch (or commit ;)
>> --
>> Alin Dreghiciu
>>
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: nexus-dev-unsubscribe@...
For additional commands, e-mail: nexus-dev-help@...

[Brian Fox]

Ok, rereading your email, perhaps you knew that already ;-) based on
my new understanding, yeah that sounds usefull

2009/7/20 Brian Fox <brianf@...>:

> Alin, the new .gz is a proprietary binary format; that is it's
> strictly not lucene. This was intentional as lucene doesn't preserve
> backwards compatability and basically locked us into a single version
> of 2.3 if we wanted other clients to be able to read the index we
> transport. The new format is read in and imported internally into a
> lucene index, which could be 2.3 or 2.4 etc.
>
> 2009/7/20 Tamás Cservenák <tamas@...>:
>> +1 sounds fine with me ;)
>> ~t~
>>
>> On Mon, Jul 20, 2009 at 8:07 PM, Alin Dreghiciu <adreghiciu@...> wrote:
>>>
>>> Hi,
>>> I'm using a tool named Luke to explore Lucene indexes created by Nexus indexer. But this tool are not able to work with the new indexing format (nexus-maven-repository-index.gz) as it is. So, I finded useful to implement a new option in CLI to allow unpacking of a known Lucene index directory format.
>>> To use it you do:
>>> <cli> -u - if the nexus-maven-repository-index.gz is in the current directory and you want the unpacked index to be placed in the current directory
>>> or you can add -d to specify the directory and/or -i to specify the directory containing the index file.
>>> If you find it useful, let me know so I can create a jira issue for it and attach a patch (or commit ;)
>>> --
>>> Alin Dreghiciu
>>>
>>
>>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: nexus-dev-unsubscribe@...
For additional commands, e-mail: nexus-dev-help@...

[Alin Dreghiciu]

Yes I figured out that. Thanx for info anyway.

I actually already implement it and committed including unit testing.

What I missed is where is the "official" guide for indexer CLI so I can document it.

Alin 

On Tue, Jul 21, 2009 at 6:32 AM, Brian Fox <brianf@...> wrote:

Ok, rereading your email, perhaps you knew that already ;-) based on
my new understanding, yeah that sounds usefull

2009/7/20 Brian Fox <brianf@...>:

> Alin, the new .gz is a proprietary binary format; that is it's
> strictly not lucene. This was intentional as lucene doesn't preserve
> backwards compatability and basically locked us into a single version
> of 2.3 if we wanted other clients to be able to read the index we
> transport. The new format is read in and imported internally into a
> lucene index, which could be 2.3 or 2.4 etc.
>
> 2009/7/20 Tamás Cservenák <tamas@...>:
>> +1 sounds fine with me ;)
>> ~t~
>>
>> On Mon, Jul 20, 2009 at 8:07 PM, Alin Dreghiciu <adreghiciu@...> wrote:
>>>
>>> Hi,
>>> I'm using a tool named Luke to explore Lucene indexes created by Nexus indexer. But this tool are not able to work with the new indexing format (nexus-maven-repository-index.gz) as it is. So, I finded useful to implement a new option in CLI to allow unpacking of a known Lucene index directory format.
>>> To use it you do:
>>> <cli> -u - if the nexus-maven-repository-index.gz is in the current directory and you want the unpacked index to be placed in the current directory
>>> or you can add -d to specify the directory and/or -i to specify the directory containing the index file.
>>> If you find it useful, let me know so I can create a jira issue for it and attach a patch (or commit ;)
>>> --
>>> Alin Dreghiciu
>>>
>>
>>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: nexus-dev-unsubscribe@...
For additional commands, e-mail: nexus-dev-help@...

--
Alin Dreghiciu
Software Developer - Looking for new projects!
My profile: http://www.linkedin.com/in/alindreghiciu
My blog: http://adreghiciu.blogspot.com
http://www.ops4j.org - New Energy for OSS Communities - Open Participation Software.
http://www.qi4j.org - New Energy for Java - Domain Driven Development.