Add minor optimizations to get_canonical

Requiring contiguous buffers removes the need to use mxs::Buffer which
also removes the need to check for buffer boundaries.

Converted all the functions used by get_canonical into `static inline` so
that the compiler knows it can inline them. A few of them weren't `static`
which made the calls to the functions unnecessarily expensive.
This commit is contained in:
Markus Mäkelä 2019-10-09 16:53:32 +03:00 committed by Johan Wikman
parent 7f41bfa5f8
commit 8258e14bfe

View File

@ -1250,7 +1250,7 @@ mxs_pcre2_result_t modutil_mysql_wildcard_match(const char* pattern, const char*
return rval;
}
static inline bool is_next(mxs::Buffer::iterator it, mxs::Buffer::iterator end, const std::string& str)
static inline bool is_next(uint8_t* it, uint8_t* end, const std::string& str)
{
mxb_assert(it != end);
for (auto s_it = str.begin(); s_it != str.end(); ++s_it, ++it)
@ -1307,12 +1307,11 @@ static const LUT is_special([](uint8_t c) {
c) != std::string::npos;
});
static std::pair<bool, mxs::Buffer::iterator> probe_number(mxs::Buffer::iterator it,
mxs::Buffer::iterator end)
static inline std::pair<bool, uint8_t*> probe_number(uint8_t* it, uint8_t* end)
{
mxb_assert(it != end);
mxb_assert(is_digit(*it));
std::pair<bool, mxs::Buffer::iterator> rval = std::make_pair(true, it);
std::pair<bool, uint8_t*> rval = std::make_pair(true, it);
bool is_hex = *it == '0';
bool allow_hex = false;
@ -1339,7 +1338,7 @@ static std::pair<bool, mxs::Buffer::iterator> probe_number(mxs::Buffer::iterator
else if (*it == 'e')
{
// Possible scientific notation number
auto next_it = std::next(it);
auto next_it = it + 1;
if (next_it == end || (!is_digit(*next_it) && *next_it != '-'))
{
@ -1356,7 +1355,7 @@ static std::pair<bool, mxs::Buffer::iterator> probe_number(mxs::Buffer::iterator
else if (*it == '.')
{
// Possible decimal number
auto next_it = std::next(it);
auto next_it = it + 1;
if (next_it != end && !is_digit(*next_it))
{
@ -1406,7 +1405,7 @@ static inline bool is_negation(const std::string& str, int i)
return rval;
}
mxs::Buffer::iterator find_char(mxs::Buffer::iterator it, const mxs::Buffer::iterator& end, char c)
static inline uint8_t* find_char(uint8_t* it, uint8_t* end, char c)
{
for (; it != end; ++it)
{
@ -1431,13 +1430,13 @@ namespace maxscale
std::string get_canonical(GWBUF* querybuf)
{
std::string rval;
mxb_assert(GWBUF_IS_CONTIGUOUS(querybuf));
uint8_t* it = GWBUF_DATA(querybuf) + MYSQL_HEADER_LEN + 1;
uint8_t* end = GWBUF_DATA(querybuf) + gwbuf_length(querybuf);
std::string rval(end - it, 0);
int i = 0;
rval.resize(gwbuf_length(querybuf) - MYSQL_HEADER_LEN + 1);
mxs::Buffer buf(querybuf);
for (auto it = std::next(buf.begin(), MYSQL_HEADER_LEN + 1); // Skip packet header and command
it != buf.end(); ++it)
for (; it != end; ++it)
{
if (!is_special(*it))
{
@ -1449,7 +1448,7 @@ std::string get_canonical(GWBUF* querybuf)
// Jump over any escaped values
rval[i++] = *it++;
if (it != buf.end())
if (it != end)
{
rval[i++] = *it;
}
@ -1470,19 +1469,19 @@ std::string get_canonical(GWBUF* querybuf)
rval[i++] = ' ';
}
}
else if (*it == '/' && is_next(it, buf.end(), "/*"))
else if (*it == '/' && is_next(it, end, "/*"))
{
auto comment_start = std::next(it, 2);
if (comment_start == buf.end())
if (comment_start == end)
{
break;
}
else if (*comment_start != '!' && *comment_start != 'M')
{
// Non-executable comment
while (it != buf.end())
while (it != end)
{
if (is_next(it, buf.end(), "*/"))
if (is_next(it, end, "*/"))
{
// Comment end marker, return to normal parsing
++it;
@ -1491,7 +1490,7 @@ std::string get_canonical(GWBUF* querybuf)
++it;
}
if (it == buf.end())
if (it == end)
{
break;
}
@ -1503,10 +1502,10 @@ std::string get_canonical(GWBUF* querybuf)
}
}
else if ((*it == '#' || *it == '-')
&& (is_next(it, buf.end(), "# ") || is_next(it, buf.end(), "-- ")))
&& (is_next(it, end, "# ") || is_next(it, end, "-- ")))
{
// End-of-line comment, jump to the next line if one exists
while (it != buf.end())
while (it != end)
{
if (*it == '\n')
{
@ -1514,7 +1513,7 @@ std::string get_canonical(GWBUF* querybuf)
}
else if (*it == '\r')
{
if ((is_next(it, buf.end(), "\r\n")))
if ((is_next(it, end, "\r\n")))
{
++it;
}
@ -1524,14 +1523,14 @@ std::string get_canonical(GWBUF* querybuf)
++it;
}
if (it == buf.end())
if (it == end)
{
break;
}
}
else if (is_digit(*it) && (i == 0 || (!is_alnum(rval[i - 1]) && rval[i - 1] != '_')))
{
auto num_end = probe_number(it, buf.end());
auto num_end = probe_number(it, end);
if (num_end.first)
{
@ -1547,7 +1546,7 @@ std::string get_canonical(GWBUF* querybuf)
else if (*it == '\'' || *it == '"')
{
char c = *it;
if ((it = find_char(std::next(it), buf.end(), c)) == buf.end())
if ((it = find_char(it + 1, end, c)) == end)
{
break;
}
@ -1556,7 +1555,7 @@ std::string get_canonical(GWBUF* querybuf)
else if (*it == '`')
{
auto start = it;
if ((it = find_char(std::next(it), buf.end(), '`')) == buf.end())
if ((it = find_char(it + 1, end, '`')) == end)
{
break;
}
@ -1569,7 +1568,7 @@ std::string get_canonical(GWBUF* querybuf)
rval[i++] = *it;
}
mxb_assert(it != buf.end());
mxb_assert(it != end);
}
// Remove trailing whitespace
@ -1581,8 +1580,6 @@ std::string get_canonical(GWBUF* querybuf)
// Shrink the buffer so that the internal bookkeeping of std::string remains up to date
rval.resize(i);
buf.release();
return rval;
}
}