regexpre-style

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: yulab.utils
 Title: Supporting Functions for Packages Maintained by 'YuLab-SMU'
-Version: 0.1.1.001
+Version: 0.1.1.002
 Authors@R: c(person("Guangchuang", "Yu", email = "[email protected]", role = c("aut", "cre"), comment = c(ORCID = "0000-0002-6485-8781")))
 Description: Miscellaneous functions commonly used by 'YuLab-SMU'.
 Imports:

NAMESPACE

@@ -4,6 +4,7 @@ S3method(print,exec)
 export(Biocpkg)
 export(CRANpkg)
 export(Githubpkg)
+export(auto_set_regexpr_style)
 export(baidu_translate)
 export(check_pkg)
 export(cn2en)
@@ -26,10 +27,14 @@ export(rbindlist)
 export(read.cb)
 export(scale_range)
 export(scihub_dl)
+export(set_PCRE)
+export(set_TRE)
+export(set_regexpr_style)
 export(show_in_excel)
 export(str_ends)
 export(str_starts)
 export(str_wrap)
+export(use_perl)
 export(yread)
 export(yread_tsv)
 importFrom(cli,cli_h2)

NEWS.md

@@ -1,5 +1,7 @@
 
-# yulab.utils 0.1.1.001
+# yulab.utils 0.1.1.002
+
++ tools to switch from PCRE or TRE in regular expression (2023-12-13, Wed)
 + `cn2en()` (2023-12-12, Tue)
 
 # yulab.utils 0.1.1

R/regexpr.R

@@ -0,0 +1,68 @@
+##' @rdname regexpr-style
+##' @export
+set_PCRE <- function() {
+    options(regexpr_use_perl = TRUE)
+}
+
+##' @rdname regexpr-style
+##' @export
+set_TRE <- function() {
+    options(regexpr_use_perl = FALSE)
+}
+
+##' @rdname regexpr-style
+##' @export
+use_perl <- function() {
+    res <- getOption("regexpr_use_perl", default = auto_set_regexpr_style())
+    return(res)
+}
+
+
+##' switch regular expression style (PCRE vs TRE)
+##'
+##' The `set_regexpr_style()` allows user to specify which style to be used, 
+##' while the `auto_set_regexpr_style()` automatically set the style depdending on 
+##' the operating system (TRE for Windows and PCRE for other OSs (Linux and Mac)).
+##' 
+##' `set_PCRE()` force to use PCRE style while `set_TRE()` force to use TRE. 
+##' 
+##' Note that all these functions are not change the behavior of `gsub()` and `regexpr()`. 
+##' The functions are just set a global option to store the user's choice of whether using `perl = TRUE`.
+##'
+##' Users can access the option via `use_perl()` and pass the return value to `gusb()` or `regexpr()` to specify the style in use.
+##' 
+##' @rdname regexpr-style
+##' @param style one of 'PCRE' or 'TRE'
+##' @return logical value of whether use perl
+##' @reference <https://stackoverflow.com/questions/47240375/regular-expressions-in-base-r-perl-true-vs-the-default-pcre-vs-tre>
+##' @export
+##' @author Guangchuang Yu
+set_regexpr_style <- function(style) {
+    if (missing(style)) {
+        message("style is not specific, set automatically.")
+        auto_set_regexpr_style()
+    } else {
+        style <- match.arg(style, c("PCRE", "TRE"))
+        if (style == "PCRE") {
+            set_PCRE()
+        } else {
+            set_TRE()
+        }
+    }
+    res <- getOption("regexpr_use_perl")
+    invisible(res)
+}
+
+##' @rdname regexpr-style
+##' @export
+auto_set_regexpr_style <- function() {
+    os <- Sys.info()[1]
+    if (os == "Windows") {
+        set_TRE()
+        res <- FALSE
+    } else {
+        set_PCRE()
+        res <- TRUE
+    }
+    invisible(res)
+}