Aespeakup should not drop all non-ascii characters.

This fixes an issue with non-english languages.
Thanks to Samuel Thibault for the patch.
This commit is contained in:
William Hubbs 2009-04-05 19:24:36 -05:00
commit 3dbdcb21cb

View file

@ -110,7 +110,7 @@ static void process_buffer(struct synth_t *s, char *buf, ssize_t length)
start = 0;
end = 0;
while (start < length) {
while (isprint(buf[end]) && end < length)
while ((buf[end] < 0 || buf[end] >= ' ') && end < length)
end++;
if (end != start) {
txtLen = end - start;