Zengtudor/ccls
1
0
Fork 0
mirror of https://github.com/MaskRay/ccls.git synced 2024-11-27 18:11:57 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

indexer: decode invalid UTF-8 byte sequence in comments as ?

Browse Source
This commit is contained in:
Fangrui Song 2019-09-16 23:51:00 -07:00
parent ff4ee614b9
commit 71fb3f4fe9
1 changed files with 23 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
src/indexer.cc
View File

@ -423,22 +423,23 @@ public:
unsigned start_column = sm.getLineNumber(bInfo.first, bInfo.second); unsigned start_column = sm.getLineNumber(bInfo.first, bInfo.second);
std::string ret; std::string ret;
int pad = -1; int pad = -1;
for (const char *p = raw.data(), *e = raw.end(); p < e;) { for (const uint8_t *p = raw.bytes_begin(), *e = raw.bytes_end(); p < e;) {
// The first line starts with a comment marker, but the rest needs // The first line starts with a comment marker, but the rest needs
// un-indenting. // un-indenting.
unsigned skip = start_column - 1; unsigned skip = start_column - 1;
for (; skip > 0 && p < e && (*p == ' ' || *p == '\t'); p++) for (; skip > 0 && p < e && (*p == ' ' || *p == '\t'); p++)
skip--; skip--;
const char *q = p; bool high = false;
const uint8_t *q = p;
while (q < e && *q != '\n') while (q < e && *q != '\n')
q++; high |= *q++ >= 0x80;
if (q < e) if (q < e)
q++; q++;
// A minimalist approach to skip Doxygen comment markers. // A minimalist approach to skip Doxygen comment markers.
// See https://www.stack.nl/~dimitri/doxygen/manual/docblocks.html // See https://www.stack.nl/~dimitri/doxygen/manual/docblocks.html
if (pad < 0) { if (pad < 0) {
// First line, detect the length of comment marker and put into |pad| // First line, detect the length of comment marker and put into |pad|
const char *begin = p; const uint8_t *begin = p;
while (p < e && (*p == '/' || *p == '*' || *p == '-' || *p == '=')) while (p < e && (*p == '/' || *p == '*' || *p == '-' || *p == '='))
p++; p++;
if (p < e && (*p == '<' || *p == '!')) if (p < e && (*p == '<' || *p == '!'))
@ -456,7 +457,24 @@ public:
(*p == ' ' || *p == '/' || *p == '*' || *p == '<' || *p == '!')) (*p == ' ' || *p == '/' || *p == '*' || *p == '<' || *p == '!'))
prefix--, p++; prefix--, p++;
} }
ret.insert(ret.end(), p, q); if (high) {
while (p < q) {
int i = 0, c = *p < 0x80 ? 0
: *p < 0xc0 || *p >= 0xf8
? -1
: *p >= 0xf0 ? 3 : *p >= 0xe0 ? 2 : 1;
const uint8_t *r = p + 1;
for (; i < c && r < q && *r >= 0x80; i++, r++)
;
if (i == c)
ret.insert(ret.end(), (const char *)p, (const char *)r);
else
ret += '?';
p = r;
}
} else {
ret.insert(ret.end(), (const char *)p, (const char *)q);
}
p = q; p = q;
} }
while (ret.size() && isspace(ret.back())) while (ret.size() && isspace(ret.back()))