Closed fryguybob closed 11 years ago
What we ought to do is pass the list of diagram names extracted in the first pass into the second pass (along with the code blocks), so that URLs like these can be properly ignored.
Good point. But we don't want to miss a warning for a legitimate missing definition.
On Tue, Mar 5, 2013 at 11:02 AM, Brent Yorgey notifications@github.comwrote:
What we ought to do is pass the list of diagram names extracted in the first pass into the second pass (along with the code blocks), so that URLs like these can be properly ignored.
— Reply to this email directly or view it on GitHubhttps://github.com/diagrams/diagrams-haddock/issues/7#issuecomment-14447642 .
Right, to be clear: "the list" will be generated just by looking at names referenced from URLs in the first pass. In the second pass, if we see a URL that's not on the list, we ignore it (since that means it got CPP'ed away in the first pass or something. If it is on the list, we try to compile it -- which may result in an error (which we would display) if the thing it refers to is not in scope.
On Tue, Mar 5, 2013 at 11:08 AM, Ryan Yates notifications@github.comwrote:
Good point. But we don't want to miss a warning for a legitimate missing definition.
On Tue, Mar 5, 2013 at 11:02 AM, Brent Yorgey notifications@github.comwrote:
What we ought to do is pass the list of diagram names extracted in the first pass into the second pass (along with the code blocks), so that URLs like these can be properly ignored.
— Reply to this email directly or view it on GitHub< https://github.com/diagrams/diagrams-haddock/issues/7#issuecomment-14447642>
.
— Reply to this email directly or view it on GitHubhttps://github.com/diagrams/diagrams-haddock/issues/7#issuecomment-14448004 .
If you have a diagram url inside a skipped #ifdef it will correctly ignore the code in comments for that section, but incorrectly find that url. This issues a warning, but is unlikely to cause harm.