knitr::opts_chunk$set(comment = "#>", collapse = TRUE)
configr is an integrated parser package that json, ini, yaml and toml format files can now be processed. The vignette will walk you through the basics of using configr to extend existing parser in R.
Example of json, ini, yaml, toml can be used follow the instructions below.
library(configr) config.json <- system.file('extdata', 'config.json', package='configr') config.ini <- system.file('extdata', 'config.ini', package='configr') config.yaml <- system.file('extdata', 'config.yaml', package='configr') config.toml <- system.file('extdata', 'config.toml', package='configr') config.glob <- system.file('extdata', 'config.global.toml', package='configr')
is.json.file
, is.ini.file
, is.yaml.file
and is.toml.file
can be used to check the configuration file type. If input file were coincident with required, it will return TRUE. get.config.type
will using above functions and get the file type name: json, ini, yaml, toml or FALSE.
is.json.file(config.json) is.toml.file(config.toml) is.ini.file(config.ini) is.yaml.file(config.yaml) get.config.type(config.json) get.config.type(config.yaml) get.config.type(config.ini) get.config.type(config.toml)
Section names of configuration file can be get using eval.config.sections
. Python package ConfigParser sections
inspired us to add this function.
eval.config.sections(config.ini) eval.config.sections(config.toml)
read.config
can read a configuration file in R and as a list object that can pass parameter to inner read function (fromJSON/read.ini/yaml.load_file/parseToml) accordingly.
# Read in R as a list (JSON/INI/YAML/TOML be suported) # fromJSON/read.ini/readLines/yaml.load parameters can be automatch by parameter name (encoding .etc.) read.config(file = config.toml)
eval.config
return a value or a list object containing the file path, config group, filetype as the attribute.
# Get the same obj with config package, only get the # 'default or R_CONFIG_ACTIVE config sets' in config.cfg or R_CONFIGFILE_ACTIVE eval.config(file = config.yaml) # Read designated section eval.config(file = config.json, config = "comments") # Read designated section with its one value eval.config(file = config.ini, config = "comments", value = "version")
eval.config.merge
will merge multiple sections (equal to config
in eval.config
function) and reduce the layer of configuration file.
eval.config.merge(file = config.json, sections = c('default', 'comments')) eval.config.merge(file = config.toml, sections = c('default', 'comments'))
fetch.config
can parse configuration files from internet and local that merged the files and return a list.
links <- c("https://raw.githubusercontent.com/JhuangLab/BioInstaller/master/inst/extdata/config/db/db_annovar.toml", "https://raw.githubusercontent.com/JhuangLab/BioInstaller/master/inst/extdata/config/db/db_main.toml", system.file('extdata', 'config.toml', package = "configr")) x <- fetch.config(links) x[c(1:5, length(x))]
convert.config
will read a configuration file and write a configuration file with appointed file type (json. ini, yaml). Moreover, write.config
is similar to convert.config
but using the list object rather than a file.
# Convert YAML configuration file to JSON format out.json <- tempfile(fileext = ".json") convert.config(file = config.yaml, out.file = out.json, convert.to = "JSON") get.config.type(out.json) # Generate a JSON format configuration file list.test <- list(a=c(123,456)) out.fn <- sprintf("%s/test.json", tempdir()) write.config(config.dat = list.test, file.path = out.fn, write.type = "json") get.config.type(out.fn) # Generate a YAML format configuration file with defined indent write.config(config.dat = list.test, file.path = out.fn, write.type = "yaml", indent = 4) get.config.type(out.fn) # Generate a YAML format configuration file with defined indent and pointed sections #write.config(config.dat = list.test, file.path = out.fn, write.type = "yaml", sections = "a", indent = 4) #get.config.type(out.fn)
configr own several userful extra parse function, you can use the parse.extra
to finish these work for any list object. Of course, read.config
, eval.config
and eval.config.merge
can directly using parse.extra
by passing parameters to parse.extra
.
extra.list
can be used to parse the value of {{debug}}
to self
if you setted extra.list = list(debug = 'self')
other.config
can be used to parse the value of {{key:yes_flag}}
to yes
if you setted other.config = system.file('extdata', 'config.other.yaml', package='configr')
which content can be founded below.rcmd.parse
can be used to parse the value of @>@str_replace('config','g$','gr')@<@
to configr
if you setted rcmd.parse = TRUE
.bash.parse
can be used to parse the value of #>#echo bash#<#
to bash
if you setted bash.parse = TRUE
.glue.parse
can be used to paste the value of !!glue {1:5}
to ["1", "2", "3", "4", "5"]
; !!glue_numeric {1:5}
to [1, 2, 3, 4, 5] global.vars.field
can be used to parse the internal values using the given top level fields default use global_vars
(new feature in v0.3.4)Note: glue.parse
using the glue
package glue
function to do that. Just like glue('{1:5}') and be processed by unname(unlist(x)).
The !!glue
can be changed if you setted glue.flag
. It is a remarkable fact that only contain the glue.flag
character be parsed and the order of item will be changed if the glue
result were multiple values. e.g. ['{a}', '!!glue {1:5}', '{{a}}']
will be parsed to ['{a}', '1', '2', '3', '4', '5', '{{a}}']
other.config <- system.file('extdata', 'config.other.yaml', package='configr') read.config(file = other.config) config.1 <- read.config(file = config.json) config.1$default read.config(file = config.json, extra.list = list(debug = "self", debug2 = "self2"))$default sections <- c('default', 'other_config_parse') config.1[sections] read.config(file = config.json, extra.list = list(debug = "self", debug2 = "self2"), other.config = other.config)[sections] sections <- c('default', 'other_config_parse', 'rcmd_parse') # The followed two line command will return the same value config.1[sections] read.config(file = config.json, extra.list = list(debug = "self", debug2 = "self2"), other.config = other.config, rcmd.parse = T)[sections] parse.extra(config.1, extra.list = list(debug = "self", debug2 = "self2"), other.config = other.config, rcmd.parse = T)[sections] sections <- c('default', 'other_config_parse', 'rcmd_parse', 'mulitple_parse') config.1[sections] parse.extra(config.1, extra.list = list(debug = "self", debug2 = "self2", yes = "1", no = "0"), other.config = other.config, rcmd.parse = T, bash.parse = T)[sections] # glue parse raw <- c("a", "!!glue{1:5}", "c") list.raw <- list(glue = raw, nochange = 1:10) list.raw expect.parsed.1 <- c("a", "1", "2", "3", "4", "5", "c") expect.parsed.2 <- list(glue = expect.parsed.1, nochange = 1:10) parse.extra(list.raw, glue.parse = TRUE, glue.flag = "!!glue") read.config(config.glob, global.vars.field = NULL) read.config(config.glob)
configr provides a function config.help
to access external resource about various configuration format and other related materies.
Show all external urls stored in configr config.help() # Open item in browser # config.help('toml_stackoverflow_search') # Or use the row number to access # config.help(23)
config.section.del
can be used to delete a section of config, just do config$section <- NULL
.
``` {r, echo = TRUE} config <- read.config(file = config.json, extra.list = list(debug = "self", debug2 = "self2"), other.config = other.config)[sections] names(config) config <- config.sections.del(config, 'default') names(config)
`str2config` can be used to parse a string object to a configuration list ```r json_string <- '{"city" : "Crich"}\n' yaml_string <- 'foo: 123\n' json_config <- str2config(json_string) yaml_config <- str2config(yaml_string) print(json_config) print(yaml_config)
Here is the output of sessionInfo()
on the system on which this document was compiled:
sessionInfo()
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.