summaryrefslogtreecommitdiffstats
path: root/src/basic/utf8.c
diff options
context:
space:
mode:
authorZbigniew Jędrzejewski-Szmek <zbyszek@in.waw.pl>2021-11-12 10:27:13 +0100
committerZbigniew Jędrzejewski-Szmek <zbyszek@in.waw.pl>2021-11-16 13:54:38 +0100
commit9b49a3b49e78fa46ab5da45e5a176bab3df5803a (patch)
treef227c2f1c8b7f3e785a8584df275e583e35eb333 /src/basic/utf8.c
parentbasic/utf8: inline some iterator variables (diff)
downloadsystemd-9b49a3b49e78fa46ab5da45e5a176bab3df5803a.tar.xz
systemd-9b49a3b49e78fa46ab5da45e5a176bab3df5803a.zip
basic/utf8: add function to convert to ASCII
The conversion must be lossy because ASCII doesn't have enough chars.
Diffstat (limited to 'src/basic/utf8.c')
-rw-r--r--src/basic/utf8.c31
1 files changed, 31 insertions, 0 deletions
diff --git a/src/basic/utf8.c b/src/basic/utf8.c
index 2ad2151816..2532fcf81a 100644
--- a/src/basic/utf8.c
+++ b/src/basic/utf8.c
@@ -312,6 +312,37 @@ char *ascii_is_valid_n(const char *str, size_t len) {
return (char*) str;
}
+int utf8_to_ascii(const char *str, char replacement_char, char **ret) {
+ /* Convert to a string that has only ASCII chars, replacing anything that is not ASCII
+ * by replacement_char. */
+
+ _cleanup_free_ char *ans = new(char, strlen(str) + 1);
+ if (!ans)
+ return -ENOMEM;
+
+ char *q = ans;
+
+ for (const char *p = str; *p; q++) {
+ int l;
+
+ l = utf8_encoded_valid_unichar(p, SIZE_MAX);
+ if (l < 0) /* Non-UTF-8, let's not even try to propagate the garbage */
+ return l;
+
+ if (l == 1)
+ *q = *p;
+ else
+ /* non-ASCII, we need to replace it */
+ *q = replacement_char;
+
+ p += l;
+ }
+ *q = '\0';
+
+ *ret = TAKE_PTR(ans);
+ return 0;
+}
+
/**
* utf8_encode_unichar() - Encode single UCS-4 character as UTF-8
* @out_utf8: output buffer of at least 4 bytes or NULL