New Check: All .Rd files have \value tags

[mpadge]

This is now caught automatically on CRAN incoming checks, so would be useful to have here, via checks for roxygen2 @return tags. That in turn would require installation of package, so affects #61

[mpadge]

Re-opening to address this issue identified by @santikka ropensci/software-review#554 (comment)

[santikka]

@mpadge A similar issue arises when defining a new method and its instances for different classes which share a return value, for example in dynamite there is new method get_code for both dynamiteformula and dynamitefit objects, which are all documented together. Defining a @return tag in the particular method instances will create duplicate entries under Value in the main method documentation page.

[mpadge]

Thanks again for your help @santikka. The previous approach directly parsed the roxygen2 blocks; the above commit replaces that with direct parsing of the .Rd files via tools::parse_Rd. The whole test attempts to replicate how the CRAN check works, without having any way of knowing exactly how they implement it. Their one would, however, definitely use parse_Rd, so this updated version should come a lot closer.

The previous, roxygen2-based one gave this list from dynamite:

• as_draws.dynamitefit
• coef.dynamitefit
• dynamite
• formula.dynamitefit
• aux
• +.dynamiteformula
• print.dynamiteformula
• get_priors.dynamiteformula
• get_priors.dynamitefit
• get_code.dynamiteformula
• get_code.dynamitefit
• get_data.dynamiteformula
• get_data.dynamitefit
• fitted.dynamitefit
• mcmc_diagnostics.dynamitefit
• plot_nus
• predict.dynamitefit
• print.dynamitefit
• summary.dynamitefit

The above commit reduces that to what i hope would reflect your expectations:

• coef.dynamitefit
• dynamite
• fitted.dynamitefit
• formula.dynamitefit
• plot_nus
• +.dynamiteformula
• predict.dynamitefit
• print.dynamitefit
• print.dynamiteformula
• summary.dynamitefit

I've closed this again now, but please comment on whether you now judge the results to appropriately reflect the docs of dynamite. Thank you so much for all the help youi've given us! I'll update the "allcontributors" section of the README straight away to include you.

[mpadge]

@santikka I've updated the bot report for your submission to reduce listed funtions to the second, shorter version above.

[santikka]

@mpadge Thanks! That list looks correct, because we hadn't documented the return value for almost any of the generic methods for dynamitefit objects.