- From: <cmarchand@oxiane.com>
- Date: Tue, 28 Mar 2017 11:58:39 +0200
- To: Hank Ratzesberger <hank@werx.io>
- Cc: xproc-dev@w3c.org
Hi Hank, welcome to this list ! In Prague, Achim has presented an embedded Morgana in a web container. See https://github.com/xproc/showroom/wiki/xmlprague2017#xproc-pipelines-at-the-heart-of-webapps input and output pipeline ports must not be configured in the pipeline, only declared. So, implementations may provide solution to connect a stream as input or as output. Morgana has a recent feature to allow to pipe to unix pipelines, so it is possible to connect to System.in or System.out, and so you should connect to any [In|Out]putStream XProc is a language, and is not bound to java specifically - even if the two active implementations are written in Java. It is your responsability to embed an implementation into a servlet container, but it shouldn't be difficult. Best regards, Christophe Le 2017-03-27 17:55, Hank Ratzesberger a écrit : > Hi, > > I'm writing to this list for the first time, and breaking the first > rule, since I have not looked at the archives to see if this topic has > come up previously, but if you will indulge me... > > Last year I attempted to run JBoss Data Virtualization, which provides > a way to make SOAP and REST messages, JNDI and Hibernate sources, all > kinds of data sources look like sql database tables. Well to me it > seemed a perfect solution for 20 years ago. > > If xproc were sitting in a servlet container with some way to > configure endpoints and processing scripts it could create the kinds > of documents applications use (which is mostly json but svg pdf html > csv also) indeed if you have seen some of these Hibernate queries > it's obvious there's some backflips to create the right json for some > front end UI. > > Another use case is BPMN as a front controller to manage the logic > state. > > Well that's my salvo. Thank you > Hank.ratzesberger
Received on Tuesday, 28 March 2017 09:59:08 UTC