Hey Ramesh, I don't know whether or not we've moved the code to GitHub yet, but you can at least find a copy of it on CodePlex: http://odatavalidation.codeplex.com/. As mentioned below, I'm not averse to recreating this tool. We would want to leverage the learning we have in the tool so far (i.e., port the rules that have already been created) but if we do this in a shared way, as you mention, I would really like to specify the rules as a DSL or otherwise easily extensible toolset.
I'll risk showing my ignorance here, but this sounds like of exactly like a problem Python would be well-suited to solve (since it's pure interpreted code and is simple to understand). Does anyone know whether there's an existing framework we can start from? -----Original Message----- From: Ramesh Reddy [mailto:[email protected]] Sent: Monday, February 16, 2015 11:26 PM To: [email protected] Subject: Re: [DISCUSS] Taking part in the GSoC Mark, Conceptually that is exactly what I had it in my mind, not sure how I missed this before on the odata.org site. Can you share where the code is hosted? Ideally I would like to run this locally on my machine to verify my service during the development. As per your contribution of rules comment that is most important aspect IMO, I was thinking, somehow one creates a DSL script, or small extension code that can be easily added to the framework to be part of the test. I need to play around more to digest your comment about conformation level. Let's figure out how we can further enhance this tool. Thanks. Ramesh.. ----- Original Message ----- > Sorry for weighing in late here. How would we posture the relationship > between this proposal and http://services.odata.org/validation/? I can > state my personal opinion, which doesn't necessarily represent > Microsoft's > opinion: we built something because we saw a need. We haven't done a > good job of advertising it and to some extent it's an NP problem > (loosely > applied) in that it's very easy to prove that a service DOESN'T > conform to OData in a certain respect, very difficult to prove that it > DOES conform to OData in all respects necessary to achieve a conformance > level. > > The code that we have for the validator today is open source and so we > certainly could consider porting it to pretty much anything. An ideal > end state in my opinion would be one in which we have a framework that > can execute validation rules, and the validation rules should ideally > be easy to contribute. The ideal ideal would be something language > agnostic so that all of us OData folks can contribute regardless of > which language we program in on a daily basis. Happy to discuss in > more detail if it would be helpful, but I LOVE the idea of a GSoC project for > OData. > > Thanks, > Mark > > -----Original Message----- > From: Ramesh Reddy [mailto:[email protected]] > Sent: Monday, February 16, 2015 2:34 AM > To: [email protected] > Subject: Re: [DISCUSS] Taking part in the GSoC > > Here is the proposal > ------------------------------------------------------------------- > Project: Develop OData Conformance Checker Tool > > Brief explanation: > OData > (https://www.oasis-open.org/committees/tc_home.php?wg_abbrev=odata) > simplifes data sharing across disparate applications in enterprise, > Cloud, and mobile devices and it is OASIS standard. This protocol > defines various service conformance levels that vendors and/or > framework implementation projects expected to develop before a service > is called OData compatible service. This project proposes to develop a > test suite that can interrogate a service and provide a results as to > conformance level, adherence to the OData specification. Using this > tool a one can make sure the service is implemented in accordance to > the specification and also promotes the cross vendor usage of services. > > Expected results: > I envision some kind of visual tool that is either web based using any > of javascript libraries or it can be simple JUnit kind of code based > tool, at the end gives a report about the conformance level, and also > reports where the service is not complying with specification. You can > also think like FindBugs kind of plugin for OData. > > Knowledge Prerequisite: > Since OData is REST protocol, this can implemented using any language, > for doing web based tool familiarity with java script is necessary > other wise Java or .Net > > Mentor: Ramesh Reddy, [email protected] > > ------------------------------------------------------------------- > > I am not sure how you guys handling the organization stuff with GSoc, > but please include above proposal. > > Thanks > > Ramesh.. > > > > ----- Original Message ----- > > As this is vendor/project neutral request to create the test suite I > > am little reluctant to create a JIRA in Olingo issues, as then it > > becomes Olingo issue. May be a WIKI page to detailing all the GSoc > > projects good idea, and based on taker we can create individual > > issues where they belong. > > I am open for mentoring. I will write something up. > > > > Ramesh.. > > > > ----- Original Message ----- > > > Hi Ramesh, > > > > > > thank you very much for your feedback :) > > > > > > I did not include the JPA extension because I do not have the same > > > extensive knowledge about this extension as Chandan. This would > > > make mentoring a student hard for me. If Chandan would be willing > > > to help out as a mentor I would reconsider this and open an issue. > > > The same goes for the JavaScript library part. I do not know > > > enough about that to feel comfortable mentoring a student working on this. > > > > > > About Olingo 570: I know this is work in progress but I hope until > > > June when a student would start implementing there is a stable > > > version of the metadata document in Json. If not there is plenty > > > of other work to be done for the Json Serializer like supporting > > > odata.metadta=full. We can adjust the scope if a student actually > > > is interested in this issue. > > > > > > About Olingo 569: Yes a cars sample would be sufficient. Currently > > > the implementation is lacking quite a bit of functionality and I > > > don`t have the time to work on this. So I had hoped a student > > > would develop this further. > > > The feedback to your github idea will follow in another thread. > > > > > > Finally about your project idea: I like that very much. We always > > > had an issue in V2 with services behaving differently because they > > > interpreted the specification differently. Would you like to open > > > an issue for that in the JIRA? If yes please tag it with gsoc2015, > > > mentor and java so it appears in the global Apache ideas list. > > > > > > Best Regards, > > > Christian > > > > > > -----Original Message----- > > > From: Ramesh Reddy [mailto:[email protected]] > > > Sent: Montag, 9. Februar 2015 21:21 > > > To: [email protected] > > > Subject: Re: [DISCUSS] Taking part in the GSoC > > > > > > Christian, > > > > > > - How about the JPA extension for V4? > > > - Java Script library for V4? > > > > > > OLINGO-570 - this is still work in much progress and there was > > > recent discussions in OData TC to make use of JSON Schema but JSON > > > Schema has not been fully evolved for the OData needs. > > > OLINGO-569 - This is good. Couple comments. The CARS example > > > started is that not sufficient? Also as part of discussion > > > OLINGO-482 with Michael B, I wrote an alternative extension > > > framework for server side processing, which includes the "TripPin" > > > example developed by the Microsoft folks. You can see the example > > > and the extended framework here > > > https://github.com/rareddy/olingo-odata4/tree/olingo-server-extens > > > io n/lib/server-core-ext . There is separate discussion thread on > > > this. > > > I very much appreciate any feed back on that. > > > > > > Also, OData TC, there has been some talk on developing a library > > > of tests that can > > > - Test the conformance level > > > - Specification Adherence > > > - Cross vendor compatibility > > > > > > Using the same example service that is implemented by all the > > > framework implementors, think like TCK tests for Java. I was > > > hoping, to evolve the above TripPin service test into such a test > > > suite, but it does require lot more time than I can spare right > > > now. That can be a good project for GSoC student to develop into a > > > testing tool. > > > This could be web based or series of JUnit tests etc. > > > > > > Thanks. > > > > > > Ramesh.. > > > > > > > > > ----- Original Message ----- > > > > Hi all, > > > > > > > > The application deadline for the Google Summer of Code is this > > > > Friday 13th, 2015. I would suggest that the Olingo project takes > > > > part this year. I would volunteer as a mentor. > > > > > > > > Since the deadline is already this Friday I went ahead and > > > > created > > > > 3 project ideas already: > > > > OData $search Query Parser for Olingo > > > > https://issues.apache.org/jira/browse/OLINGO-568 > > > > Implement an OData V4 sample service using Olingo > > > > https://issues.apache.org/jira/browse/OLINGO-569 > > > > Implement OData Json Metadocument Serializer/Parser > > > > https://issues.apache.org/jira/browse/OLINGO-570 > > > > > > > > Please let me know what you think about this if you have the > > > > time > > > > :) > > > > > > > > Best Regards, > > > > Christian > > > > > > > > > > > > > >
