fixes #252: Allow plain-code-blocks in RMarkdown #349
Conversation
Added a test to the end of test.Rmd that included a plain-code-block like this: ``` This shouldn't cause a problem ``` The code-block reproduced the `Malformed File` error described in #252 Plain-blocks match the knitr `chunk.end` pattern at both the start and the end of the block. So when plain-blocks are present there was 2k _more_ chunk ends than starts. A small function `filter_chunk_end_positions` was added to filter the positions of the chunk-ends based on the chunk-starts
|
not the most beautiful way to find the blocks, but I beleive it would work. |
|
Thanks! Could you add a example like in #322 as well to the Rmd. Also please add a note to NEWS.md explaining the change and referencing the issue number and your GitHub username. |
|
Is it OK if I work on the @randy3k - if you have a suggestion for a more idiomatic way to match the start/end positions I'm happy to modify the code; my first attempt wasn't even this beautiful ... |
|
Have updated |
|
For consistency with the other helper functions in |
Sure that makes sense, thanks! |
Main function `extract_r_source` now precedes all it's helper functions (get_knitr_pattern, filter_chunk_end_positions, replace_prefix)
Added a test to the end of test.Rmd that included a plain-code-block like this:
That is, a code-block where the interpreter is not specified (these are allowable in Rmarkdown according to section 2.5.2 of Yihui's Rmarkdown book).
The code-block reproduced the
Malformed Fileerror described in issue #252Plain-blocks match the knitr
chunk.endpattern at both the start and the endof the chunk. So when plain-blocks are present there was an extra 2k chunk ends
than starts.
A small function
filter_chunk_end_positionswas added to filter the positionsof the chunk-ends based on the chunk-starts and called from
extract_r_source