Rework logrx string handling to remove stringr/stringi dependencies

DESCRIPTION

@@ -45,9 +45,7 @@ Imports:
     rlang,
     stats,
     tidyr,
-    stringr,
     sessioninfo (>= 1.2),
-    stringi,
     tibble,
     digest,
     lifecycle,

NAMESPACE

@@ -33,16 +33,6 @@ importFrom(purrr,set_names)
 importFrom(rlang,.data)
 importFrom(sessioninfo,session_info)
 importFrom(stats,aggregate)
-importFrom(stringi,stri_wrap)
-importFrom(stringr,str_c)
-importFrom(stringr,str_count)
-importFrom(stringr,str_detect)
-importFrom(stringr,str_extract)
-importFrom(stringr,str_remove)
-importFrom(stringr,str_remove_all)
-importFrom(stringr,str_replace_all)
-importFrom(stringr,str_starts)
-importFrom(stringr,str_trim)
 importFrom(tibble,tibble)
 importFrom(tidyr,all_of)
 importFrom(tidyr,complete)

NEWS.md

@@ -10,6 +10,7 @@
   - Added comprehensive tests for both interactive and non-interactive scenarios
 
 - Removed .dcf file for old Addin (#280)
+- Reworked string handling to use base R and removed `{stringr}`/`{stringi}` dependencies. (#294)
 
 ## Documentation
 

R/interact.R

@@ -140,7 +140,6 @@ run_file <- function(file) {
 #' Safely run an R script and record results, outputs, messages, errors, warnings
 #'
 #' @importFrom purrr safely discard
-#' @importFrom stringr str_starts
 #'
 #' @param file File to run
 #'
@@ -150,7 +149,7 @@ run_file <- function(file) {
 #'
 run_safely_loudly <- function(file) {
   ret <- loudly(run_safely(file))
-  set_log_element("messages", discard(ret$messages, ~ str_starts(.x, "Error")))
+  set_log_element("messages", discard(ret$messages, ~ startsWith(.x, "Error")))
   set_log_element("output", ret$output)
   set_log_element("result", ret$result$result)
   set_log_element("warnings", ret$warnings)

R/read_log_file.R

@@ -2,11 +2,6 @@
 #'
 #' @param log_txt String vector. Object with log text lines
 #'
-#' @importFrom stringr str_detect
-#' @importFrom stringr str_count
-#' @importFrom stringr str_remove
-#' @importFrom stringr str_replace_all
-#'
 #' @return tibble that ensures formatted subsections
 #'
 #' @examples
@@ -19,21 +14,21 @@
 reformat_subsections <- function(log_txt) {
   adj_log_txt <- c()
   for (i in log_txt) {
-    adj_tf <- stringr::str_detect(
-      i,
-      "Errors:|Warnings:|Messages:|Output:|Result:"
+    adj_tf <- grepl(
+      "Errors:|Warnings:|Messages:|Output:|Result:",
+      i
     )
     if (adj_tf) {
-      nrem <- stringr::str_count(i)
-      i <- stringr::str_remove(i, ":")
+      char_count <- nchar(i)
+      i <- sub(":", "", i)
       i <-
-        paste("-", i, paste(rep("-", 54 - nrem), collapse = ""),
+        paste("-", i, paste(rep("-", 54 - char_count), collapse = ""),
           collapse = ""
         )
     }
     # replace utf8 line and double line to ascii due to cli symbol variation
-    i <- stringr::str_replace_all(i, "\u2550", "=")
-    i <- stringr::str_replace_all(i, "\u2500", "=")
+    i <- gsub("\u2550", "=", i, fixed = TRUE)
+    i <- gsub("\u2500", "=", i, fixed = TRUE)
     adj_log_txt <- c(adj_log_txt, i)
   }
   return(adj_log_txt)
@@ -43,8 +38,6 @@ reformat_subsections <- function(log_txt) {
 #'
 #' @param adj_log_txt String vector. Object with formatted log text lines
 #'
-#' @importFrom stringr str_remove_all
-#'
 #' @return list that includes nested log sections
 #'
 #' @noRd
@@ -67,8 +60,7 @@ nest_sections <- function(adj_log_txt) {
       }
     }
   }
-  sect_headers <-
-    stringr::str_remove_all(sect_headers, "-?\\s{3,}-?")
+  sect_headers <- gsub("-?\\s{3,}-?", "", sect_headers, perl = TRUE)
   names(sect_info) <- sect_headers
 
   return(sect_info)
@@ -79,26 +71,34 @@ nest_sections <- function(adj_log_txt) {
 #' @param adj_log_txt String vector. Object with formatted log text lines
 #' @param sect_info String vector. Object with nested sections
 #'
-#' @importFrom stringr str_extract
-#' @importFrom stringr str_trim
-#' @importFrom stringr str_remove_all
-#'
 #' @return list that includes nested log subsections
 #'
 #' @noRd
 #'
 nest_subsections <- function(adj_log_txt, sect_info) {
+  # Extract the first regex match from each string in `x` using `pattern`.
+  # x: character vector of strings to inspect.
+  # pattern: regular expression pattern to extract.
+  # Returns a character vector the same length as `x`, with `NA` for no match.
+  extract_match <- function(x, pattern) {
+    match_positions <- regexpr(pattern, x, perl = TRUE)
+    result <- rep(NA_character_, length(x))
+    has_match <- match_positions != -1
+    if (any(has_match)) {
+      result[has_match] <- regmatches(x, match_positions)
+    }
+    result
+  }
+
   subsect_headers <- stats::na.omit(
-    stringr::str_extract(adj_log_txt, "[\\-|\\=]\\s\\w+\\s(\\w+\\s)?[\\-|\\=]{3,70}")
+    extract_match(adj_log_txt, "[\\-|\\=]\\s\\w+\\s(\\w+\\s)?[\\-|\\=]{3,70}")
   )
   subset_sections <- function(section) {
     subsect_status <- FALSE
     subsect_info <- list()
     for (i in section) {
       if (i %in% subsect_headers) {
-        latest_subsect <- stringr::str_trim(
-          stringr::str_remove_all(i, "[\\-|\\=]")
-        )
+        latest_subsect <- trimws(gsub("[\\-|\\=]", "", i, perl = TRUE))
         subsect_status <- TRUE
       } else if (subsect_status) {
         subsect_info[[latest_subsect]] <-
@@ -135,7 +135,6 @@ nest_log <- function(adj_log_txt) {
 #'
 #' @importFrom tibble tibble
 #' @importFrom tidyr separate
-#' @importFrom stringr str_replace_all
 #' @importFrom dplyr rename_with
 #' @importFrom dplyr mutate
 #'
@@ -170,7 +169,7 @@ parse_log <- function(nested_log) {
     parsed_log$`User and File Information` <-
       nested_log$`User and File Information` %>%
       unlist() %>%
-      stringr::str_trim() %>%
+      trimws() %>%
       tibble::tibble() %>%
       tidyr::separate(".",
         sep = "\\: ",
@@ -182,24 +181,24 @@ parse_log <- function(nested_log) {
     parsed_log$`Session Information`$`Session info` <-
       nested_log$`Session Information`$`Session info` %>%
       unlist() %>%
-      stringr::str_trim() %>%
+      trimws() %>%
       tibble::tibble() %>%
       tidyr::separate(".",
         sep = "\\s",
         into = c("setting", "value"),
         extra = "merge",
         fill = "right"
       ) %>%
-      dplyr::mutate(dplyr::across(tidyselect::where(is.character), stringr::str_trim))
+      dplyr::mutate(dplyr::across(tidyselect::where(is.character), trimws))
     parsed_log$`Session Information`$`Packages` <-
       nested_log$`Session Information`$`Packages` %>%
       # remove indicator whether the package is attached to the search path
-      stringr::str_replace_all("\\*", " ") %>%
+      gsub("\\*", " ", .) %>%
       # account for loaded packages due to load_all()
-      stringr::str_replace_all(" P ", "   ") %>%
+      gsub(" P ", "   ", ., fixed = TRUE) %>%
       readr::read_table(skip = 1, col_names = FALSE)
 
-    # handle case where log is has 7 columns due to sessioninfo v1.2.2 or earlier
+    # handle case where log has 7 columns due to sessioninfo v1.2.2 or earlier
     if (ncol(parsed_log$`Session Information`$`Packages`) == 7) {
       parsed_log$`Session Information`$`Packages` <-
         parsed_log$`Session Information`$`Packages` %>%
@@ -213,39 +212,40 @@ parse_log <- function(nested_log) {
           "r_version"
         )) %>%
         dplyr::mutate(
-          lang = stringr::str_remove(lang, "\\("),
-          r_version = stringr::str_remove(r_version, "\\)")
+          lang = sub("\\(", "", lang),
+          r_version = sub("\\)", "", r_version)
         )
     } else {
+      standard_names <- c("package", "version", "date", "lib", "source")
+      actual_n <- ncol(parsed_log$`Session Information`$`Packages`)
+      final_names <- if (actual_n <= 5) {
+        standard_names[seq_len(actual_n)]
+      } else {
+        c(standard_names, paste0("extra_", seq_len(actual_n - 5)))
+      }
       parsed_log$`Session Information`$`Packages` <-
         parsed_log$`Session Information`$`Packages` %>%
-        dplyr::rename_with(~ c(
-          "package",
-          "version",
-          "date",
-          "lib",
-          "source"
-        ))
+        dplyr::rename_with(~ final_names)
     }
 
     parsed_log$`Session Information`$`External software` <-
       nested_log$`Session Information`$`External software` %>%
-      stringr::str_trim() %>%
+      trimws() %>%
       tibble::tibble() %>%
       tidyr::separate(".",
         sep = "\\s",
         into = c("setting", "value"),
         extra = "merge",
         fill = "right"
       ) %>%
-      dplyr::mutate(dplyr::across(tidyselect::where(is.character), stringr::str_trim))
+      dplyr::mutate(dplyr::across(tidyselect::where(is.character), trimws))
   }
 
   if ("Repo URLs" %in% names(nested_log)) {
     parsed_log$`Repo URLs` <-
       nested_log$`Repo URLs` %>%
       unlist() %>%
-      stringr::str_trim() %>%
+      trimws() %>%
       tibble::tibble() %>%
       tidyr::separate(".",
         sep = "\\:\\s+",
@@ -269,7 +269,7 @@ parse_log <- function(nested_log) {
         sep = "\\} ",
         into = c("library", "function_names")
       ) %>%
-      dplyr::mutate(library = stringr::str_remove(library, "\\{"))
+      dplyr::mutate(library = sub("\\{", "", library))
   }
 
   if ("Program Run Time Information" %in% names(nested_log)) {

R/writer.R

@@ -43,8 +43,6 @@ write_metadata <- function() {
 #' Format log.rx's session info attribute for writing
 #'
 #' @importFrom purrr map_chr
-#' @importFrom stringi stri_wrap
-#' @importFrom stringr str_c
 #'
 #' @return A vector of log.rx's session info attribute
 #'
@@ -59,11 +57,8 @@ write_session_info <- function() {
       .x
     )) %>%
     # wrap any other elements over 80 characters
-    map_chr(~ stri_wrap(.x,
-      width = 80, exdent = 2, simplify = FALSE, use_length = TRUE,
-      normalize = FALSE, whitespace_only = TRUE
-    ) %>%
-      map_chr(~ str_c(.x, collapse = "\n\t", character(1))))
+    map_chr(~ strwrap(.x, width = 80, exdent = 2) %>%
+      paste(collapse = "\n\t"))
 
   return(session_info)
 }
@@ -247,7 +242,7 @@ write_errors <- function() {
 
   paste0(
     "Errors:\n\t",
-    str_replace_all(errors$message, "\n", "\n\t")
+    gsub("\n", "\n\t", errors$message, fixed = TRUE)
   )
 }
 
@@ -269,15 +264,14 @@ write_warnings <- function() {
 #' Format messages attribute for writing
 #'
 #' @importFrom purrr map
-#' @importFrom stringr str_remove_all
 #'
 #' @return A formatted vector of messages
 #'
 #' @noRd
 #'
 write_messages <- function() {
   messages <- get_log_element("messages") %>%
-    map(~ str_remove_all(.x, "\n"))
+    map(~ gsub("\n", "", .x, fixed = TRUE))
 
   paste0(
     "Messages:\n\t",
@@ -287,8 +281,6 @@ write_messages <- function() {
 
 #' Format output attribute for writing
 #'
-#' @importFrom stringr str_replace_all
-#'
 #' @return A formatted vector of output
 #'
 #' @noRd
@@ -298,7 +290,7 @@ write_output <- function() {
 
   paste0(
     "Output:\n\t",
-    str_replace_all(output, "\n", "\n\t")
+    gsub("\n", "\n\t", output, fixed = TRUE)
   )
 }
 
@@ -361,12 +353,6 @@ write_extra_info <- function() {
       call. = FALSE
     )
   }
-  if (!requireNamespace("stringr", quietly = TRUE)) {
-    stop(
-      "Package \"stringr\" must be installed to use this function.",
-      call. = FALSE
-    )
-  }
   extra_info <- get_log_element("extra_info")
   results <- yaml::as.yaml(
     extra_info,
@@ -375,6 +361,9 @@ write_extra_info <- function() {
       logical = yaml::verbatim_logical
     )
   )
-  results <- stringr::str_split(results, pattern = "\n")[[1]]
+  results <- strsplit(results, split = "\n", fixed = TRUE)[[1]]
+  if (length(results) == 0 || results[length(results)] != "") {
+    results <- c(results, "")
+  }
   return(results)
 }

tests/testthat/test-parse.R

@@ -55,3 +55,24 @@ test_that("read_log_file will parse a logrx log file and create the necessary ob
   rm(scriptPath, logDir, parsedFile)
   log_remove()
 })
+
+test_that("nest_subsections handles both matched and unmatched subsection headers", {
+  matched_headers <- c("- Session info -------")
+  section_with_header <- list(
+    `Session Information` = c("- Session info -------", " setting value", " foo bar")
+  )
+
+  matched_result <- logrx:::nest_subsections(matched_headers, section_with_header)
+  expect_named(matched_result[[1]], "Session info")
+  expect_identical(
+    matched_result[[1]][["Session info"]],
+    c(" setting value", " foo bar")
+  )
+
+  no_headers <- c("plain text")
+  section_without_header <- list(`Session Information` = c("line one", "line two"))
+  unmatched_result <- logrx:::nest_subsections(no_headers, section_without_header)
+  expect_length(unmatched_result[[1]], 2)
+  expect_identical(unmatched_result[[1]][[1]], "line one")
+  expect_identical(unmatched_result[[1]][[2]], "line two")
+})

tests/testthat/test-writer.R

@@ -292,7 +292,6 @@ test_that("write_repo_urls will return a formatted log output element", {
 
 test_that("write_extra_info will return a formatted log output element", {
   skip_if_not_installed("yaml")
-  skip_if_not_installed("stringr")
   options("log.rx" = NULL)
   ref_extra_info1 <- list("some info", "more info")
   fp <- test_path("ref", "safely_loudly_test_file_result.R")