rock (version 0.8.1)

recode_split: Split a code into multiple codes

Description

This function conditionally splits a code into multiple codes. Note that you may want to use recode_addChildCodes() instead to not lose the original coding.

Usage

recode_split(
  input,
  codes,
  splitToCodes,
  filter = TRUE,
  output = NULL,
  filenameRegex = ".*",
  outputPrefix = "",
  outputSuffix = "_recoded",
  decisionLabel = NULL,
  justification = NULL,
  justificationFile = NULL,
  preventOverwriting = rock::opts$get("preventOverwriting"),
  encoding = rock::opts$get("encoding"),
  silent = rock::opts$get("silent")
)

Value

Invisibly, the changed source(s) or source(s) object.

Arguments

input

One of 1) a character string specifying the path to a file with a source; 2) an object with a loaded source as produced by a call to load_source(); 3) a character string specifying the path to a directory containing one or more sources; 4) or an object with a list of loaded sources as produced by a call to load_sources().

codes

A single character value with the code to split.

splitToCodes

A named list with specifying when to split to which new code. Each element of this list is a filtering criterion that will be passed on to get_source_filter() to create the actual filter that will be applied. The name of each element is the code that will be applied to utterances matching that filter. When calling recode_split() for a single source, instead of passing the filtering criterion, it is also possible to pass a filter (i.e. the result of the call to get_source_filter()), which allows more finegrained control. Note that these split filters and the corresponding codes are processed sequentially in the order specified in splitToCodes. This means that once an utterance that was coded with codes has been matched to one of these 'split filters' (and so, recoded with the corresponding 'split code', i.e., with the name of that split filter in splitToCodes), it will not be recoded again even if it also matches with other split filters down the line. Any utterances coded with the code to split up (i.e. specified in codes) that do not match with any of the split filters specified as the splitToCodes elements will not be recoded and so remain coded with codes. To create a catch-all ('else') category, pass ".*" or TRUE as a filter (see the example).

filter

Optionally, a filter to apply to specify a subset of the source(s) to process (see get_source_filter()).

output

If specified, the recoded source(s) will be written here.

filenameRegex

Only process files matching this regular expression.

outputPrefix, outputSuffix

The prefix and suffix to add to the filenames when writing the processed files to disk, in case multiple sources are passed as input.

decisionLabel

A description of the (recoding) decision that was taken.

justification

The justification for this action.

justificationFile

If specified, the justification is appended to this file. If not, it is saved to the justifier::workspace(). This can then be saved or displayed at the end of the R Markdown file or R script using justifier::save_workspace().

preventOverwriting

Whether to prevent overwriting existing files when writing the files to output.

encoding

The encoding to use.

silent

Whether to be chatty or quiet.

Examples

### Get path to example source
examplePath <-
  system.file("extdata", package="rock");

### Get a path to one example file
exampleFile <-
  file.path(examplePath, "example-1.rock");

### Load example source
loadedExample <- rock::load_source(exampleFile);

### Split a code into two codes, showing progress
recoded_source <-
  rock::recode_split(
    loadedExample,
    codes="childCode1",
    splitToCodes = list(
      and_REPLACED = " and ",
      book_REPLACED = "book",
      else_REPLACED = TRUE
    ),
    silent=FALSE
  );