Hello every one, we would like to submit you a new plugin we have developed for SonarQube. Here come required information:
Description:
SonarQube i-Code CNES plugin adds the ability to import i-Code CNES results for Fortran (77 & 90) & Shell into SonarQube. i-Code CNES is an open source analysis tool which both checks rules and measure metrics on three languages : Shell, Fortran 77 & Fortran 90. This plugin is able to read i-Code’s results and import them as SonarQube’s issues and measures.
At a glance, your plugin key doesn’t conform to the requirements. Also, we’ll need to test. Can you point me to an appropriate set of projects to do that with? Even better if you already have handy the reports to import.
(Sorry, I don’t know how to type the special ‘i’.)
This works nicely and generally looks well put together, and I think this can probably go into the Marketplace as-is, but before I add it, I’d like to clarify a few points:
It looks like all the Fortran rules, for both ‘languages’, have descriptions of “Unreferenced rule”. Was that on purpose? Rule descriptions are there to help users understand what they’re doing wrong, and are particularly helpful in cases like this, where I guess a specific format is expected:
You’ve tagged a number of rules with either ‘f77’ or ‘f90’. Was it just to have some tag on each rule? If so, I’d advise against it; it’s redundant with the language (mostly) yet its presence on some rules and not others is likely to be confusing
You’ve tagged some Fortran 90 rules and one Shell rule with ‘f77’. Was this in error?
You’ve added a Nesting metric. It’s not clear to me what it’s about, but in case it’s relevant to one of the existing metric categories, you’re free to put it there instead of creating a new one.
I got the following in my analysis log, and “0.00%” for duplications in my project: “INFO: Calculating CPD for 0 files”. It might be worth taking another look at since I think how duplications are calculated is mostly language-independent and maybe not something you have to do (altho to be honest, I really don’t know).
It looks like all the Fortran rules, for both ‘languages’, have descriptions of “Unreferenced rule”. Was that on purpose? Rule descriptions are there to help users understand what they’re doing wrong, and are particularly helpful in cases like this, where I guess a specific format is expected
There are no descriptions for now for Fortran rules. We will work on adding some.
You’ve tagged a number of rules with either ‘f77’ or ‘f90’. Was it just to have some tag on each rule? If so, I’d advise against it; it’s redundant with the language (mostly) yet its presence on some rules and not others is likely to be confusing
It was to differentiate f77 & f90 rules when there was only one Fortran language. Tag can be removed.
You’ve added a Nesting metric. It’s not clear to me what it’s about, but in case it’s relevant to one of the existing metric categories, you’re free to put it there instead of creating a new one.
Nesting represents the maximum number of nesting blocks. Maybe it could be added in ‘Size’ category?
I got the following in my analysis log, and “0.00%” for duplications in my project: “INFO: Calculating CPD for 0 files”. It might be worth taking another look at since I think how duplications are calculated is mostly language-independent and maybe not something you have to do (altho to be honest, I really don’t know).
This plugin only imports i-Code issues, i-Code does not calculate duplications. If there is a SonarQube API to simply calculate them we can add CPD to the plugin.
I have edited the post and the plugin according to the first three points.
The new version integrate your feedback except for duplications that needs to be discussed for a next major version. It won’t be supported for this one.
