bugfix #415

gagolews · Apr 7, 2021 · b413535 · b413535
1 parent 0040045
commit b413535
Show file tree

Hide file tree

Showing 4 changed files with 15 additions and 4 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: stringi
 Version: 1.5.4
-Date: 2021-03-02
+Date: 2021-04-07
 Title: Character String Processing Facilities
 Description: A multitude of character string/text/natural language
     processing tools: pattern searching (e.g., with 'Java'-like regular

diff --git a/NEWS b/NEWS
@@ -10,6 +10,9 @@
 * [INTERNAL] #414: Use `LEVELS(x)` macro instead of accessing `(x)->sxpinfo.gp`
     directly (@lukaszdaniel).
 
+* [BUGFIX] #415: `locale=''` was not equivalent to `locale=NULL`
+    in `stri_opts_collator()`.
+
 * ... (to do) ..., for example:
     #25 (locate-capture groups),
     #31 (sub_index),
@@ -24,7 +27,7 @@
 ## 1.5.3 (2020-09-04) **CRAN**
 
 * [DOCUMENTATION] stringi home page has moved to https://stringi.gagolewski.com
-  and now includes the comprehensive reference manual.
+  and now includes a comprehensive reference manual.
 
 * [NEW FEATURE] #400: `%s$%` and `%stri$%` are now binary operators
   that call base R's `sprintf()`.

diff --git a/src/stri_collator.cpp b/src/stri_collator.cpp
@@ -74,7 +74,7 @@ UCollator* stri__ucol_open(SEXP opts_collator)
 
     if (narg <= 0) { // no custom settings - use default Collator
         UErrorCode status = U_ZERO_ERROR;
-        UCollator* col = ucol_open(NULL, &status);
+        UCollator* col = ucol_open(uloc_getDefault(), &status);
         STRI__CHECKICUSTATUS_RFERROR(status, {/* do nothing special on err */}) // error() allowed here
         return col;
     }
@@ -93,7 +93,7 @@ UCollator* stri__ucol_open(SEXP opts_collator)
     UColAttributeValue  opt_STRENGTH =  UCOL_DEFAULT_STRENGTH;
     UColAttributeValue  opt_NUMERIC_COLLATION = UCOL_DEFAULT;
 //   USearchAttributeValue  opt_OVERLAP = USEARCH_OFF;
-    const char*         opt_LOCALE = NULL;
+    const char*         opt_LOCALE = uloc_getDefault();
 
     for (R_len_t i=0; i<narg; ++i) {
         if (STRING_ELT(names, i) == NA_STRING)

diff --git a/src/stri_prepare_arg.cpp b/src/stri_prepare_arg.cpp
@@ -1147,6 +1147,9 @@ const char* stri__prepare_arg_string_1_notNA(SEXP x, const char* argname)
  *
  * @version 0.5-1 (Marek Gagolewski, 2015-01-01)
  *    "@keyword=value" may use default locale from now; also, loc is trimmed
+ *
+ * @version 1.5.4 (Marek Gagolewski, 2021-04-07)
+ *    BUGFIX: locale='' is the default
  */
 const char* stri__prepare_arg_locale(SEXP loc, const char* argname, bool allowdefault, bool allowna)
 {
@@ -1159,6 +1162,11 @@ const char* stri__prepare_arg_locale(SEXP loc, const char* argname, bool allowde
             if (allowna) return NULL;
             else Rf_error(MSG__ARG_EXPECTED_NOT_NA, argname); // Rf_error allowed here
         }
+        if (strlen((const char*)CHAR(STRING_ELT(loc, 0))) == 0) {
+            UNPROTECT(1);
+            if (allowdefault) return uloc_getDefault();
+            else Rf_error(MSG__LOCALE_INCORRECT_ID); // Rf_error allowed here
+        }
 
         UErrorCode err = U_ZERO_ERROR;
         char buf[ULOC_FULLNAME_CAPACITY];