Enforce use of UTF-8 encoding when writing strings.

This avoids incorrect saving of multi-byte characters in non-UTF-8 locales.

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: alabaster.base
 Title: Save Bioconductor Objects To File
-Version: 1.5.8
-Date: 2024-09-13
+Version: 1.5.9
+Date: 2024-09-22
 Authors@R: person("Aaron", "Lun", role=c("aut", "cre"), email="[email protected]")
 License: MIT + file LICENSE
 Description: 

R/hdf5.R

@@ -81,6 +81,9 @@ h5_create_vector <- function(handle, name, len, type, compress=6, chunks=NULL, s
 #' @export
 h5_write_vector <- function(handle, name, x, type=NULL, compress=6, chunks=NULL, scalar=FALSE, emit=FALSE) {
     if (is.null(type)) {
+        if (is.character(x)) {
+            x <- enc2utf8(x) # avoid mis-encoding multi-byte characters from Latin-1.
+        }
         type <- .choose_type(x)
     }
 
@@ -101,6 +104,9 @@ h5_write_vector <- function(handle, name, x, type=NULL, compress=6, chunks=NULL,
 #' @export
 h5_write_attribute <- function(handle, name, x, type=NULL, scalar=FALSE) {
     if (is.null(type)) {
+        if (is.character(x)) {
+            x <- enc2utf8(x) # avoid mis-encoding multi-byte characters from Latin-1.
+        }
         type <- .choose_type(x)
     }
 

R/transformVectorForHdf5.R

@@ -36,6 +36,7 @@ transformVectorForHdf5 <- function(x, .version=3) {
         }
 
     } else if (is.character(x)) {
+        x <- enc2utf8(x) # avoid mis-encoding multi-byte characters from Latin-1.
         if (anyNA(x)) {
             placeholder <- chooseMissingPlaceholderForHdf5(x)
             x[is.na(x)] <- placeholder