[GitHub] [arrow] ianmcook commented on a change in pull request #9878: ARROW-11513: [R] Bindings for sub/gsub [WIP]

[GitBox <gi...@apache.org>]

ianmcook commented on a change in pull request #9878:
URL: https://github.com/apache/arrow/pull/9878#discussion_r606880444



##########
File path: r/R/dplyr.R
##########
@@ -434,6 +438,75 @@ build_function_list <- function(FUN) {
   )
 }
 
+arrow_r_string_replace_function <- function(FUN, max_replacements) {
+  function(pattern, replacement, x, ignore.case = FALSE, fixed = FALSE) {
+    if (ignore.case) {
+      if (fixed) {
+        pattern <- paste0("(?i)\\Q", pattern, "\\E")
+      } else {
+        pattern <- paste0("(?i)", pattern)
+      }
+    }
+    FUN(
+      ifelse(fixed && !ignore.case, "replace_substring", "replace_substring_regex"),
+      x,
+      options = list(
+        pattern = pattern,
+        replacement = replacement,
+        max_replacements = max_replacements
+      )
+    )
+  }
+}
+
+arrow_stringr_string_replace_function <- function(FUN, max_replacements) {
+  function(string, pattern, replacement) {
+    # Assign stringr pattern modifier functions locally

Review comment:
       I considered this and concluded that there were several reasons this approach is preferable.
   
   This approach avoids requiring the user to have stringr installed or loaded. Since we do not need stringr to be installed or loaded for the `str_replace()` and `str_replace_all()` functions to work with Arrow, it would seem strange to require stringr to be installed or loaded to use these pattern modifier functions.
   
   Also, although stringr exports the pattern modifier functions and their arguments are documented and very unlikely to change without prior deprecation warnings, the structure of the objects they _return_ is an internal stringr implementation detail which we should not depend on. The stringr help entry for these  modifier functions lacks a *Value* section, so they could conceivably change anytime.




-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org