New chunk option: cache.rebuild
#995
Conversation
|
This looks perfect. Could you squash the four commits commits into one? https://github.com/yihui/knitr/blob/master/CONTRIBUTING.md#pull-requests Thanks! |
|
BTW, please also add a news item to https://github.com/yihui/knitr/blob/master/NEWS.md and yourself to https://github.com/yihui/knitr/blob/master/DESCRIPTION. |
|
Due to the large amount of changes this weekend, it might be confusing for you to resolve the conflicts, so I just went ahead and merged the PR. Thanks! |
|
Thanks! I have been away from my computer for a few days. I forgot to look Zach On Sunday, April 19, 2015, Yihui Xie notifications@github.com wrote:
Zachary S.L. Foster, PhD Student |
From documentation (pull request #996):
cache.rebuild: (FALSE) ifTRUE, the chunk is evaluated even ifcache=TRUEand no changes are made. This is useful when given an R code argument.
For example:
cache.rebuild=!file.exists("path/to/file.ext")(see #238).
I had a bit of trouble understanding everything that was going on in the code, but I think I got the gist of it. I basically used the idea that when
cache.rebuild = TRUE, the code for loading and running the chunk would proceed as if the cache did not exist, but the parts of the code that saves the cache would remain unaffected. I tested this a couple different ways and it seems to accomplish the goal of #238, but I cant guarantee its free of side effects without a greater understanding of the code. I can do some more rigorous testing, but it might help if you could point me towards some things it might break. Thanks!