list_rmd_chunks: Tools for working with existing chunks in Rmarkdown documents

In knitrdata: Data Language Engine for 'knitr' / 'rmarkdown'

Description

These helper functions allow one to identify all the chunks in a Rmarkdown document, split the document into pieces by a specific chunk so that one can either work with the chunk contents or remove the chunk, and remove several chunks at once.

Usage

list_rmd_chunks(
  text = readLines(file),
  file = NULL,
  chunk.start.pattern = "^```[{](.+)[}] *$",
  chunk.end.pattern = "^``` *$"
)

split_rmd_by_chunk(text = readLines(file), chunk_label, file = NULL, ...)

remove_chunks(
  text = readLines(file),
  chunk_labels,
  file = NULL,
  output.file = NULL,
  ...
)

Arguments

text	Character vector with contents of chunk, one element per line of text. If the character vector has just a single element, then an attempt will be made to split it into lines using readLines.
file	Path to file containing chunk contents. Ignored if text argument supplied. As a consequence, this means that all arguments must be named if the file argument is supplied.
chunk.start.pattern	Regular expression used to identify chunk starts. The default looks for lines beginning with three back quotes, followed by curly braces with some sort of text between them and then only spaces till the end of the line. This should generally work, but if the Rmarkdown document has chunks that have unusual headers, then this argument can be useful. In particular, if the document has chunks that begin without curly braces, these will not be recognized.
chunk.end.pattern	Regular expression used to identify the chunk end. Default should generally work.
chunk_label	Character string giving the chunk label or the chunk number (as returned by list_rmd_chunks.
...	Additional arguments to be passed to list_rmd_chunks (e.g., chunk.start.pattern).
chunk_labels	A vector of numeric or character chunk labels (as returned by list_rmd_chunks.
output.file	Name of a file where Rmd document with desired chunks removed is to be saved.

Details

list_rmd_chunks takes a Rmarkdown document and returns a data.frame listing the essential information of every chunk, including chunk type (language engine), label and start and end line numbers.

split_rmd_by_chunk takes a Rmarkdown document and a chunk label or number and returns the Rmarkdown document split into 4 pieces: the part before the chunk, the chunk header, the chunk contents, the chunk tail and the part after the chunk. These can then be used to either work with the chunk contents or remove the chunk from the Rmarkdown document.

remove_chunks removes several chunks, designated by their text or numeric labels, all at once from a Rmarkdown document.

Note that the regular expression used by default to identify chunk starts is not guaranteed to be exactly the same as that used by knitr and may not work if the Rmarkdown document has unusual chunks. In particular, each chunk must have the chunk type and chunk options enclosed in curly braces. If code chunks exist without curly braces, then these will generally be ignored, but they could potentially cause problems in unusual cases.

Functions

• list_rmd_chunks: Returns a data frame with 4 columns: the chunk type, the chunk label, the line number of the beginning of the chunk and the line number of the end of the chunk.

• split_rmd_by_chunk: Returns a list with the contents of the Rmarkdown document broken into 4 pieces: pre-chunk, chunk header, chunk contents, chunk tail, and post-chunk.

• remove_chunks: Silently returns a character vector with the contents of the Rmd file after having removed the desired chunks

Author(s)

David M. Kaplan [e-mail]

David M. Kaplan [e-mail]

David M. Kaplan [e-mail]

See Also

Other Chunk tools: create_chunk(), create_data_chunk_dialog(), insert_data_chunk_template(), remove_chunks_dialog()

Examples

# Use a temporary directory ----------------------------
owd = getwd()
td = tempdir()
setwd(td)

# Test --------------
library(knitrdata)
library(magrittr) # For pipe operator

# Create new Rmarkdown document
if (file.exists("test.create_chunks.Rmd"))
  file.remove("test.create_chunks.Rmd")
rmarkdown::draft("test.create_chunks.Rmd","github_document","rmarkdown",
                 edit=FALSE)

# List all chunks in document
chunklst = list_rmd_chunks(file="test.create_chunks.Rmd")
chunklst

# Remove the pressure chunk
xx = split_rmd_by_chunk(file="test.create_chunks.Rmd",chunk_label="pressure")
txt = c(xx$pre_chunk,xx$post_chunk)
writeLines(txt,"test.create_chunks.Rmd")

# List chunks again
chunklst = list_rmd_chunks(file="test.create_chunks.Rmd")
chunklst

# Remove all but setup chunk
remove_chunks(file="test.create_chunks.Rmd",
              chunk_labels = 2:nrow(chunklst),
              output.file="test.create_chunks.Rmd")

# List all chunks again
chunklst = list_rmd_chunks(file="test.create_chunks.Rmd")
chunklst

# Create some binary data
x = data.frame(a=1:10,b=(1:10)^2)
saveRDS(x,"test.create_chunks.RDS")

# Push chunks into Rmarkdown document
# Insert in reverse order to not have to figure out line number
txt = create_chunk(chunk_label="plot",c("x","plot(b~a,data=x)"),chunk_type="r") %>%
  insert_chunk(11,rmd.file="test.create_chunks.Rmd")
txt = data_encode("test.create_chunks.RDS","base64") %>%
  create_chunk(chunk_label="thedata",output.var="x",format="binary",loader.function=readRDS) %>%
  insert_chunk(11,txt)
txt = create_chunk(chunk_label="loadknitrdata","library(knitrdata)",chunk_type="r") %>%
  insert_chunk(11,txt)

writeLines(txt,"test.create_chunks.Rmd")

# List all chunks again
chunklst = list_rmd_chunks(file="test.create_chunks.Rmd")
chunklst

# Render document to test
if (rmarkdown::pandoc_available(version="1.12.3"))
  rmarkdown::render("test.create_chunks.Rmd")

# Clean up --------------
file.remove("test.create_chunks.Rmd","test.create_chunks.RDS",
            "test.create_chunks.md","test.create_chunks.html")
unlink("test.create_chunks_files",recursive=TRUE)

setwd(owd)

knitrdata documentation built on Dec. 8, 2020, 5:08 p.m.