Extend documentation mode to document tests#756
Conversation
|
The CI error seems a little bit odd, maybe roswell is using a different rove version from quicklisp? |
|
Any idea on how to solve this CI issue? Maybe don't include rove as dependency? (ping @cxxxr) |
|
I do not currently understand the fundamental cause, but it does seem like something is wrong if the production code is dependent on the test framework. |
Indeed, I moved the file to scritps/ and added an entry to the Makefile, the idea of the append version is that it can be generated with different implementations easily 👍 |
|
fukamachi/rove@82be18b |
|
I think this feature is very great. |
|
I changed the rove calls for the new version on quicklisp and it seems to work correctly 👍 |
cxxxr
left a comment
cxxxr
left a comment
There was a problem hiding this comment.
I made a few comments, but overall I think it is very good.
Minor comments: Indentation includes tabs, but I prefer to use all spaces.
Is it because it is the default setting in emacs?
| (insert-character point #\newline))))) | ||
|
|
||
| (defun generate-markdown-file (filename) | ||
| (defgeneric generate-markdown-file (filename type)) |
There was a problem hiding this comment.
The method dispatch is very attractive and I love it.
But here I felt it would be better to simply separate the functions
What do you think?
There was a problem hiding this comment.
Ummm, I was thinking to use this functionality to generate other kind of markdown, like with maybe benchmark, so I think it can be reused.
But I tend to overuse generics, so if you like I can change it back to a function 👍
There was a problem hiding this comment.
I was only thinking about statically determining which method the caller is.
I thought a function would be better in that case.
However, if we need extensibility in the future, it may be better to make it a method.
There was a problem hiding this comment.
However, if we need extensibility in the future, it may be better to make it a method.
Ummm, I do think it may be extensible in the future indeed, in any case I'll change it if you want 👍
There was a problem hiding this comment.
Well, this might be fine as it is too.
| (enough-namestring (location-file location) (asdf:system-source-directory :lem)) | ||
| (location-line-number location)))) | ||
|
|
||
| #+sbcl |
There was a problem hiding this comment.
Since separation by *features* is complicated, I think it is preferable to keep it to a minimum.
except sbcl, what about command-definition-location returning meaningless values (e.g., always set position to 0)?
There was a problem hiding this comment.
Yes, I thought it would be better to consolidate on the called side rather than the caller.
How about setting the file name to nil to indicate that it has no meaning?
|
Thank you so much! |
2 participants
The idea is to generate a markdown file with a table of each tests, this is intended to use not only with SBCL but with more implementations.