hrbrmstr
/
splashr
mirror of https://git.sr.ht/~hrbrmstr/splashr
1
1
Fork 0
Code Issues Releases Wiki Activity
Tools to Work with the 'Splash' JavaScript Rendering Service in R
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
56 Commits
2 Branches
0 Tags
11 MiB
 
Tree: d6aa675ea4
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from 'd6aa675ea4'
${ noResults }
splashr/cran-comments.md

2.1 KiB
Raw Blame History

Test environments

  • local OS X install, R 3.4.1 on both 10.12 and 10.13 Beta 6
  • local ubuntu 3.4.1 and r-devel
  • ubuntu on travis-ci, R oldrel, 3.4.1 and r-devel
  • win-builder (devel and release)

R CMD check results

0 errors | 0 warnings | 1 note

  • This is a new release.

Reverse dependencies

This is a new release, so there are no reverse dependencies.

Hey folks,

The tests skip on CRAN and the examples are \dontrun{} because they require an active Splash server running for the most accurate results (mocking would not exercise the interaction with the Splash API server). The core functionality is much the same as RSelenium, but Splash is more designed for scraping than Selenium which is more geared towards web site automated testing. This will hopefully be a nice addition to the R web scraping world.

It is fully tested on Travis: https://travis-ci.org/hrbrmstr/splashr with 2 different Python configurations across all three R versions. It's an active test that spins up a local Docker container and runs on push and weekly. (It imports the 'docker' package, hence my mentioning the diverse python configs).

Neither Appveyor (and, all other free Windows CI systems) nor rhub nor WinBuilder support co-runbninbg Docker linux containers so the tests are not run there, but the package is built regularly on WinBuilder and rhub's Windows and "CRAN" environments to catch any other platform-related package errors.

There are many Imports because this works with a diverse amount of web content and also tries to play well with 'httr' and 'rvest' as well as the 'hartools' package.

There are three vignettes and I tried to reduce the image sizes for the figures in them as much as possible. I'll glady do some cropping if you would like the total size to be even smaller. I fully understand wanting to keep pacakges as small as possible.

The package is also well-documented and has 44% code coverage: https://codecov.io/gh/hrbrmstr/splashr/tree/master/R I'm going to get that up closer to 80% for the next release which will extend the "DSL" functions.

Thx for your time & efforts!

-boB

P.S. Hope you folks had a great summer!