WIP: DIY pagination

[krlmlr]

Rough draft for pagination as suggested by @jennybc.

• Each response stores the original request
• Extract "page" param of "next" link in URL
• Update "page" parameter in original request, issue it

Use case:

p <- list(search.code("difftime user:cran", per_page = 100))
p <- c(p, list(next.page(p[[length(p)]])))
# rinse and repeat
items <- p %>% lapply("[[", "content") %>% lapply("[[", "items") %>% unlist(recursive = FALSE)

This is strictly in a "works for me" state. Possible extensions:

• Documentation
• Error handling
• Rename to fetch.page(page = "next"), also allow specifying a page number
• Implement fetch.all.pages() as straightforward extension (needs arg warn = FALSE in fetch.page())
• Implement get.page.count(request) so that fetch.all.pages() can show a progress bar
• Use all parameters from the link, not just "page" (needs discussion)
• Cleanup

Do we want to follow the approach sketched here?

Fixes #30.

[cscheid]

This one requires a bit more careful of a review, and it's unlikely I'll get to it before thursday or so next week. Ok?

[krlmlr]

This needs a bit of work from me, I just wanted to know if you like the design or if you prefer something completely different.

[cscheid]

Oof, this completely fell off the radar for me. Thanks for the bump. I'll be looking at it this week.

[cscheid]

@jennybc, since you're a user of rgithub and a much better R developer than me, I'd love to hear your thoughts on this PR.

[krlmlr]

Just tried it again for pull requests, still works for me.