commit cd386e712b7587740115c0fb0815a43052758ae2 Quentin Rameau <quinq@fifth.space> 2021-01-24 14:53:06 +0100 mbsprint: fix character encoding errors handling We were not advancing the input string correctly in case of error. Now we do that and print in a slightly more unified way. Send remarks to Hiltjo@melikenobugs.tk
sacc.c | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/sacc.c b/sacc.c index 08bfe84..74f41cb 100644 --- a/sacc.c +++ b/sacc.c @@ -103,7 +103,8 @@ mbsprint(const char *s, size_t len) { wchar_t wc; size_t col = 0, i, slen; - int rl, w; + const char *p; + int rl, pl, w; if (!len) return col; @@ -112,20 +113,23 @@ mbsprint(const char *s, size_t len) for (i = 0; i < slen; i += rl) { rl = mbtowc(&wc, s + i, slen - i < 4 ? slen - i : 4); if (rl == -1) { - mbtowc(NULL, NULL, 0); /* reset state */ - fputs("\xef\xbf\xbd", stdout); /* replacement character */ - col++; - rl = 1; - continue; + /* reset state */ + mbtowc(NULL, NULL, 0); + p = "\xef\xbf\xbd"; /* replacement character */ + pl = 3; + rl = w = 1; + } else { + if ((w = wcwidth(wc)) == -1) + continue; + pl = rl; + p = s + i; } - if ((w = wcwidth(wc)) == -1) - continue; if (col + w > len || (col + w == len && s[i + rl])) { fputs("\xe2\x80\xa6", stdout); /* ellipsis */ col++; break; } - fwrite(s + i, 1, rl, stdout); + fwrite(p, 1, pl, stdout); col += w; } return col;