ExcludeTags is a list of tags to exclude + StopOnParsingError option

[pleveill]

• transform ExcludeTags from unique tags to a list of tags to exclude
• Add option StopOnParsingError on gerkins parsing error (due to syntax error ...) to all runners. Default values is false.

These two new feature are needed to generate doc on a big tree of feature (from a subcontractor) that is progressively included into a csproj each time a new feature is ok (syntax ...)

[dirkrombauts]

Thanks for taking the time to contribute to Pickles. I will not merge this pull request, for a couple of reasons:

• It does too much. You work on ExcludeTags and you add StopOnParsingError. Please send separate pull requests for separate features. Git branches will help you to keep the work separate.
• ExcludeTags should support expressions, as described in Idea: Exclude Scenarios or Features by tags #18
• I am not convinced of the need of StopOnParsingError. Why do you need it?

[pleveill]

Thanks for your answer. About exclude tags the best should be a filter string like used by vstest/nunit/… runner that understand and, or, () … This is a huge job and the best should be to get/start from existing c# code found somewhere. Currently, this is a to big activity for my company into a documentation activity scope. Then I have coded only the “highway case” with an implicit « and » between all tags. It not perfect but definitively better that a unique tag. For the StopOnParsingError, we not specially need it. What we need is that doc generation not stop on the first error but continue and just report error (like any compiler … tools). This was not the case, then I updated the code (2 files 2 lines). Then to be able to use the tool with the same behaviors than as previously, I have also added the StopOnParsingError option (lot of files are updated due to lot of data structure duplication, especially into xaml code). * Then, if you are interested, I can push only the “not stop on the first error but continue” code update that is very small (2 files 2 lines) and mandatory for us. I will try to explain why we needs this: It is mandatory for us because our tree of feature files contains a lot of mal formed feature files written by high level tester and manager before our test GUI platform was available. Or cause these files are not included into our test *csproj files but are into git ! There are corrected progressively but we talk about more than 1000 wrong files into a tree of more than 10000 feature files. This is not a good practice to push malformed file into git but it has been done and now it is too late. We can just progressively clean this. Then for us it is mandatory that the doc generator not stop at the first malformed file and may be more important, that it report list of malformed files as we can list them and plan their correction. De : Dirk Rombauts [mailto:[email protected]] Envoyé : mardi 7 mars 2017 15:00 À : picklesdoc/pickles <[email protected]> Cc : Leveille, Pascal (Nokia - FR) <[email protected]>; Author <[email protected]> Objet : Re: [picklesdoc/pickles] ExcludeTags is a list of tags to exclude + StopOnParsingError option (#441) Thanks for taking the time to contribute to Pickles. I will not merge this pull request, for a couple of reasons: * It does too much. You work on ExcludeTags and you add StopOnParsingError. Please send separate pull requests for separate features. Git branches will help you to keep the work separate. * ExcludeTags should support expressions, as described in #18<#18> * I am not convinced of the need of StopOnParsingError. Why do you need it? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<#441 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AWS-wwB58GWFrjgxmOzRZf7LgdWlpGU0ks5rjWL2gaJpZM4MT_kV>.

[dirkrombauts]

I agree that it would be better to have a report about all malformed .feature files instead of stopping at the first "bad" one. But Pickles should still return with an error code at the end of the run, so that people who rely on pickles returning an error in their scripts can continue to do so. If you can send me a pull request that does that, I will review it.

About the Exclude tags: yes, it's a big task. But how about this: you basically wrote code that takes a couple of tags, separated by ; and treat them as OR. So what if you adapt your code to use OR as a separator instead of ;? That would bring us one step forward.

You also did a bit of code to include ExcludeTags in the MSBuild version, right? could you send that code as a separate pull request please?

[pleveill]

I will do it. De : Dirk Rombauts [mailto:[email protected]] Envoyé : mercredi 8 mars 2017 13:48 À : picklesdoc/pickles <[email protected]> Cc : Leveille, Pascal (Nokia - FR) <[email protected]>; Author <[email protected]> Objet : Re: [picklesdoc/pickles] ExcludeTags is a list of tags to exclude + StopOnParsingError option (#441) I agree that it would be better to have a report about all malformed .feature files instead of stopping at the first "bad" one. But Pickles should still return with an error code at the end of the run, so that people who rely on pickles returning an error in their scripts can continue to do so. If you can send me a pull request that does that, I will review it. About the Exclude tags: yes, it's a big task. But how about this: you basically wrote code that takes a couple of tags, separated by ; and treat them as OR. So what if you adapt your code to use OR as a separator instead of ;? That would bring us one step forward. You also did a bit of code to include ExcludeTags in the MSBuild version, right? could you send that code as a separate pull request please? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<#441 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AWS-w0JoR5M2YxyoB64PFZyVdiSY2q84ks5rjqNogaJpZM4MT_kV>.