[ 
https://issues.apache.org/jira/browse/MRESOLVER-46?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16631042#comment-16631042
 ] 

Robert Scholte commented on MRESOLVER-46:
-----------------------------------------

The real issue are the checksums here. The calculation and the upload are 2 
separate steps. In case of files that's not an issue, you can read it twice. 
However, with an inputstream you must ensure that checksum and content match. 
Transformation might include adding a timestamp or just a random value.
I don't want to store this to a temporary file either, because that would open 
a backdoor for changes as well. So the only solution I could implement without 
completely changing the code is to do this in memory.


> Add support InputStream/OutputStream transformers
> -------------------------------------------------
>
>                 Key: MRESOLVER-46
>                 URL: https://issues.apache.org/jira/browse/MRESOLVER-46
>             Project: Maven Resolver
>          Issue Type: New Feature
>            Reporter: Robert Scholte
>            Assignee: Robert Scholte
>            Priority: Major
>             Fix For: 1.1.2
>
>
> There are several ideas to generate new files based on the pom or even 
> rewrite/optimize the pom file itself . At this time Maven Resolver always 
> falls back to a file. The generated file must be reliable, so there should be 
> no need for temporary storage as a file (which could be adjusted by some).
> Instead is must be possible to specify Stream Filters, so the transformation 
> is done in memory.
> Register once, picked up for both install/deploy



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to