Vectorise replacement function (#462)

This is a breaking change but radically improves performance so I think it's worth it.

NEWS.md

@@ -1,5 +1,9 @@
 # stringr (development version)
 
+* In `str_replace_all()`, a `replacement` function now receives all values in
+  a single vector. This radically improves performance at the cost of breaking
+  some existing uses (#462).
+
 # stringr 1.5.1
 
 * Some minor documentation improvements.

R/replace.R

@@ -23,9 +23,11 @@
 #'   References of the form `\1`, `\2`, etc will be replaced with
 #'   the contents of the respective matched group (created by `()`).
 #'
-#'   Alternatively, supply a function, which will be called once for each
-#'   match (from right to left) and its return value will be used to replace
-#'   the match.
+#'   Alternatively, supply a function (or formula): it will be passed a single
+#'   character vector and should return a character vector of the same length.
+#'
+#'   To replace the complete string with `NA`, use
+#'   `replacement = NA_character_`.
 #' @return A character vector the same length as
 #'   `string`/`pattern`/`replacement`.
 #' @seealso [str_replace_na()] to turn missing values into "NA";
@@ -56,7 +58,7 @@
 #' colours <- str_c("\\b", colors(), "\\b", collapse="|")
 #' col2hex <- function(col) {
 #'   rgb <- col2rgb(col)
-#'   rgb(rgb["red", ], rgb["green", ], rgb["blue", ], max = 255)
+#'   rgb(rgb["red", ], rgb["green", ], rgb["blue", ], maxColorValue = 255)
 #' }
 #'
 #' x <- c(
@@ -180,18 +182,62 @@ str_replace_na <- function(string, replacement = "NA") {
 
 str_transform <- function(string, pattern, replacement) {
   loc <- str_locate(string, pattern)
-  str_sub(string, loc, omit_na = TRUE) <- replacement(str_sub(string, loc))
+  new <- replacement(str_sub(string, loc))
+  str_sub(string, loc, omit_na = TRUE) <- new
   string
 }
-str_transform_all <- function(string, pattern, replacement) {
+
+str_transform_all <- function(string, pattern, replacement, error_call = caller_env()) {
   locs <- str_locate_all(string, pattern)
 
-  for (i in seq_along(string)) {
-    for (j in rev(seq_len(nrow(locs[[i]])))) {
-      loc <- locs[[i]]
-      str_sub(string[[i]], loc[j, 1], loc[j, 2]) <- replacement(str_sub(string[[i]], loc[j, 1], loc[j, 2]))
-    }
+  old <- str_sub_all(string, locs)
+
+  # unchop list into a vector, apply replacement(), and then rechop back into
+  # a list
+  old_flat <- vctrs::list_unchop(old)
+  if (length(old_flat) == 0) {
+    # minor optimisation to avoid problems with the many replacement
+    # functions that use paste
+    new_flat <- character()
+  } else {
+    withCallingHandlers(
+      new_flat <- replacement(old_flat),
+      error = function(cnd) {
+        cli::cli_abort(
+          c(
+            "Failed to apply {.arg replacement} function.",
+            i = "It must accept a character vector of any length."
+          ),
+          parent = cnd,
+          call = error_call
+        )
+      }
+    )
   }
 
+  if (!is.character(new_flat)) {
+    cli::cli_abort(
+      "Function {.arg replacement} must return a character vector, not {.obj_type_friendly {new_flat}}.",
+      call = error_call
+    )
+  }
+  if (length(new_flat) != length(old_flat)) {
+    cli::cli_abort(
+      "Function {.arg replacement} must return a vector the same length as the input ({length(old_flat)}), not length {length(new_flat)}.",
+      call = error_call
+    )
+  }
+
+  idx <- chop_index(old)
+  new <- vctrs::vec_chop(new_flat, idx)
+
+  stringi::stri_sub_all(string, locs) <- new
   string
 }
+
+chop_index <- function(x) {
+  ls <- lengths(x)
+  start <- cumsum(c(1L, ls[-length(ls)]))
+  end <- start + ls - 1L
+  lapply(seq_along(ls), function(i) seq2(start[[i]], end[[i]]))
+}

R/view.R

@@ -107,7 +107,7 @@ str_view_filter <- function(x, pattern, match) {
 
 str_view_highlighter <- function(html = TRUE) {
   if (html) {
-    function(x) paste0("<span class='match'>", x, "</span>")
+    function(x) str_c("<span class='match'>", x, "</span>")
   } else {
     function(x) {
       out <- cli::col_cyan("<", x, ">")
@@ -123,9 +123,15 @@ str_view_highlighter <- function(html = TRUE) {
 
 str_view_special <- function(x, html = TRUE) {
   if (html) {
-    replace <- function(x) paste0("<span class='special'>", x, "</span>")
+    replace <- function(x) str_c("<span class='special'>", x, "</span>")
   } else {
-    replace <- function(x) cli::col_cyan("{", stri_escape_unicode(x), "}")
+    replace <- function(x) {
+      if (length(x) == 0) {
+        return(character())
+      }
+
+      cli::col_cyan("{", stri_escape_unicode(x), "}")
+    }
   }
 
   # Highlight any non-standard whitespace characters

revdep/README.md

@@ -1,48 +1,53 @@
 # Revdeps
 
-## Failed to check (24)
+## Failed to check (34)
 
-|package       |version |error |warning |note |
-|:-------------|:-------|:-----|:-------|:----|
-|admiral       |0.8.4   |1     |        |     |
-|admiralonco   |0.1.0   |1     |        |     |
-|NA            |?       |      |        |     |
-|NA            |?       |      |        |     |
-|NA            |?       |      |        |     |
-|NA            |?       |      |        |     |
-|NA            |?       |      |        |     |
-|genekitr      |?       |      |        |     |
-|ggPMX         |?       |      |        |     |
-|NA            |?       |      |        |     |
-|NA            |?       |      |        |     |
-|NA            |?       |      |        |     |
-|MARVEL        |?       |      |        |     |
-|numbat        |?       |      |        |     |
-|OlinkAnalyze  |?       |      |        |     |
-|Platypus      |?       |      |        |     |
-|RVA           |?       |      |        |     |
-|NA            |?       |      |        |     |
-|NA            |?       |      |        |     |
-|tidySEM       |?       |      |        |     |
-|NA            |?       |      |        |     |
-|tinyarray     |?       |      |        |     |
-|NA            |?       |      |        |     |
-|xpose.nlmixr2 |?       |      |        |     |
+|package        |version    |error |warning |note |
+|:--------------|:----------|:-----|:-------|:----|
+|AovBay         |0.1.0      |1     |        |     |
+|apollo         |0.3.3      |1     |        |     |
+|arealDB        |0.6.3      |1     |        |     |
+|BayesFactor    |0.9.12-4.7 |1     |        |     |
+|CGPfunctions   |0.6.3      |1     |        |     |
+|covid19india   |0.1.4      |1     |        |     |
+|CSCNet         |0.1.2      |1     |        |     |
+|dataone        |2.2.2      |1     |        |     |
+|FAIRmaterials  |0.4.2.1    |1     |        |     |
+|gmoTree        |?          |      |        |     |
+|Greg           |2.0.2      |1     |        |     |
+|interactionRCS |0.1.1      |1     |        |     |
+|iNZightPlots   |2.15.3     |1     |        |     |
+|metajam        |0.3.0      |1     |        |     |
+|miceafter      |0.5.0      |1     |        |     |
+|Multiaovbay    |0.1.0      |1     |        |     |
+|ontologics     |0.7.0      |1     |        |     |
+|popstudy       |1.0.1      |1     |        |     |
+|pre            |1.0.7      |1     |        |     |
+|psfmi          |1.4.0      |1     |        |     |
+|qris           |1.1.1      |1     |        |     |
+|QTOCen         |0.1.1      |1     |        |     |
+|quantoptr      |0.1.3      |1     |        |     |
+|quid           |0.0.1      |1     |        |     |
+|rlme           |0.5        |1     |        |     |
+|scaper         |0.1.0      |1     |        |     |
+|scCustomize    |2.1.2      |1     |        |1    |
+|scpi           |2.2.5      |1     |        |     |
+|scRNAstat      |0.1.1      |1     |        |     |
+|SEERaBomb      |2019.2     |1     |        |     |
+|seqimpute      |2.0.0      |1     |        |     |
+|SimplyAgree    |0.2.0      |1     |        |     |
+|tidyseurat     |0.8.0      |1     |        |     |
+|weightQuant    |1.0.1      |1     |        |     |
 
-## New problems (12)
+## New problems (7)
 
-|package       |version  |error  |warning |note |
-|:-------------|:--------|:------|:-------|:----|
-|[cmcR](problems.md#cmcr)|0.1.9    |__+1__ |        |     |
-|[crispRdesignR](problems.md#crisprdesignr)|1.1.6    |__+1__ |        |2    |
-|[cspp](problems.md#cspp)|0.3.2    |__+1__ |__+1__  |     |
-|[flair](problems.md#flair)|0.0.2    |__+2__ |__+1__  |2    |
-|[GetLattesData](problems.md#getlattesdata)|1.4.1    |__+2__ |__+1__  |     |
-|[glmmPen](problems.md#glmmpen)|1.5.1.10 |__+1__ |        |2    |
-|[mpwR](problems.md#mpwr)|0.1.0    |__+2__ |__+1__  |     |
-|[postpack](problems.md#postpack)|0.5.3    |__+1__ |__+1__  |1    |
-|[repr](problems.md#repr)|1.1.4    |__+1__ |        |2    |
-|[tidyfst](problems.md#tidyfst)|1.7.5    |__+1__ |        |1    |
-|[tidyft](problems.md#tidyft)|0.4.5    |__+1__ |        |2    |
-|[zipangu](problems.md#zipangu)|0.3.1    |__+2__ |        |1    |
+|package     |version |error    |warning |note   |
+|:-----------|:-------|:--------|:-------|:------|
+|[huxtable](problems.md#huxtable)|5.5.6   |1 __+1__ |        |2      |
+|[latex2exp](problems.md#latex2exp)|0.9.6   |__+2__   |        |__+1__ |
+|[phenofit](problems.md#phenofit)|0.3.9   |__+2__   |        |1      |
+|[priceR](problems.md#pricer)|1.0.1   |__+1__   |        |1      |
+|[PubChemR](problems.md#pubchemr)|2.0     |__+2__   |        |       |
+|[rtiddlywiki](problems.md#rtiddlywiki)|0.1.0   |__+1__   |        |       |
+|[salty](problems.md#salty)|0.1.0   |__+2__   |        |1      |
 

revdep/cran.md

@@ -1,71 +1,73 @@
 ## revdepcheck results
 
-We checked 1675 reverse dependencies (1663 from CRAN + 12 from Bioconductor), comparing R CMD check results across CRAN and dev versions of this package.
+We checked 2023 reverse dependencies (2022 from CRAN + 1 from Bioconductor), comparing R CMD check results across CRAN and dev versions of this package.
 
- * We saw 12 new problems
- * We failed to check 12 packages
+ * We saw 7 new problems
+ * We failed to check 33 packages
 
 Issues with CRAN packages are summarised below.
 
 ### New problems
 (This reports the first line of each new failure)
 
-* cmcR
-  checking tests ... ERROR
-
-* crispRdesignR
-  checking examples ... ERROR
-
-* cspp
-  checking examples ... ERROR
-  checking re-building of vignette outputs ... WARNING
-
-* flair
-  checking examples ... ERROR
-  checking tests ... ERROR
-  checking re-building of vignette outputs ... WARNING
-
-* GetLattesData
+* huxtable
   checking examples ... ERROR
-  checking tests ... ERROR
-  checking re-building of vignette outputs ... WARNING
 
-* glmmPen
+* latex2exp
   checking tests ... ERROR
+  checking running R code from vignettes ... ERROR
+  checking re-building of vignette outputs ... NOTE
 
-* mpwR
+* phenofit
   checking examples ... ERROR
   checking tests ... ERROR
-  checking re-building of vignette outputs ... WARNING
 
-* postpack
+* priceR
   checking examples ... ERROR
-  checking re-building of vignette outputs ... WARNING
 
-* repr
+* PubChemR
   checking tests ... ERROR
+  checking running R code from vignettes ... ERROR
 
-* tidyfst
-  checking examples ... ERROR
-
-* tidyft
-  checking examples ... ERROR
+* rtiddlywiki
+  checking tests ... ERROR
 
-* zipangu
+* salty
   checking examples ... ERROR
   checking tests ... ERROR
 
 ### Failed to check
 
-* admiral       (NA)
-* admiralonco   (NA)
-* genekitr      (NA)
-* ggPMX         (NA)
-* MARVEL        (NA)
-* numbat        (NA)
-* OlinkAnalyze  (NA)
-* Platypus      (NA)
-* RVA           (NA)
-* tidySEM       (NA)
-* tinyarray     (NA)
-* xpose.nlmixr2 (NA)
+* AovBay         (NA)
+* apollo         (NA)
+* arealDB        (NA)
+* BayesFactor    (NA)
+* CGPfunctions   (NA)
+* covid19india   (NA)
+* CSCNet         (NA)
+* dataone        (NA)
+* FAIRmaterials  (NA)
+* Greg           (NA)
+* interactionRCS (NA)
+* iNZightPlots   (NA)
+* metajam        (NA)
+* miceafter      (NA)
+* Multiaovbay    (NA)
+* ontologics     (NA)
+* popstudy       (NA)
+* pre            (NA)
+* psfmi          (NA)
+* qris           (NA)
+* QTOCen         (NA)
+* quantoptr      (NA)
+* quid           (NA)
+* rlme           (NA)
+* scaper         (NA)
+* scCustomize    (NA)
+* scpi           (NA)
+* scRNAstat      (NA)
+* SEERaBomb      (NA)
+* seqimpute      (NA)
+* SimplyAgree    (NA)
+* tidyseurat     (NA)
+* weightQuant    (NA)