//! Converts escape sequences to the corresponding characters
void char_constants< char >::translate_escape_sequences(string_type& str)
{
using namespace std; // to make sure we can use C functions unqualified
string_type::iterator it = str.begin();
while (it != str.end())
{
it = std::find(it, str.end(), '\\');
if (std::distance(it, str.end()) >= 2)
{
it = str.erase(it);
switch (*it)
{
case 'n':
*it = '\n'; break;
case 'r':
*it = '\r'; break;
case 'a':
*it = '\a'; break;
case '\\':
++it; break;
case 't':
*it = '\t'; break;
case 'b':
*it = '\b'; break;
case 'x':
{
string_type::iterator b = it;
if (std::distance(++b, str.end()) >= 2)
{
char_type c1 = *b++, c2 = *b++;
if (isxdigit(c1) && isxdigit(c2))
{
*it++ = char_type((to_number(c1) << 4) | to_number(c2));
it = str.erase(it, b);
}
}
break;
}
default:
{
if (*it >= '0' && *it <= '7')
{
string_type::iterator b = it;
int c = (*b++) - '0';
if (*b >= '0' && *b <= '7')
c = c * 8 + (*b++) - '0';
if (*b >= '0' && *b <= '7')
c = c * 8 + (*b++) - '0';
*it++ = char_type(c);
it = str.erase(it, b);
}
break;
}
}
}
}
}
static std::deque<string_type> tokenize(string_type input, string_type br)
{
typename string_type::iterator a,b,c;
std::deque<string_type> tokens;
// std::cerr << "Tokenising string=" << input << " break=" << br << std::endl;
while ( input.length() )
{
a = input.begin();
c = input.end();
typename string_type::size_type e = input.find(br);
if ( e != string_type::npos )
{
b = a+e;
tokens.push_back(string_type(a,b));
input = string_type(b+br.length(),c);
}
else
{
tokens.push_back(input);
input = string_type();
}
// std::cerr << "Pushed token " << tokens.back() << " input now " << input << std::endl;
}
return tokens;
}
void put_string(iter_type& oi, const string_type& s1) const
{
typename string_type::const_iterator si,end;
for (si=s1.begin(), end=s1.end(); si!=end; si++, oi++) {
*oi = *si;
}
}
template<class manipulator, class parser_type> void set_and_configure(manipulator & manip, const string_type & name, parser_type parser) {
// need to parse string
string_type prev ;
bool parsing_params = false;
string_type params;
string_type stripped_str;
for ( typename string_type::const_iterator b = name.begin(), e = name.end(); b != e; ++b) {
if ( (*b == '(') && !parsing_params) {
if ( parser.has_manipulator_name() ) {
parsing_params = true;
params.clear();
}
}
else if ( (*b == ')') && parsing_params) {
BOOST_ASSERT ( parser.has_manipulator_name() );
manip.configure_inner( parser.get_manipulator_name(), params);
parser.clear();
parsing_params = false;
}
else {
if ( parsing_params)
params += *b;
else {
stripped_str += *b;
parser.add( *b);
}
}
}
manip.string( stripped_str);
}
OutputIterator format
(
OutputIterator out
, const string_type &fmt
, regex_constants::match_flag_type flags = regex_constants::format_default
) const
{
detail::results_traits<char_type> traits;
typename string_type::const_iterator cur = fmt.begin(), end = fmt.end();
if(0 != (regex_constants::format_literal & flags))
{
out = std::copy(cur, end, out);
}
else while(cur != end)
{
if(BOOST_XPR_CHAR_(char_type, '$') != *cur)
{
*out++ = *cur++;
}
else if(++cur == end)
{
*out++ = BOOST_XPR_CHAR_(char_type, '$');
}
else if(BOOST_XPR_CHAR_(char_type, '$') == *cur)
{
*out++ = *cur++;
}
else if(BOOST_XPR_CHAR_(char_type, '&') == *cur) // whole match
{
++cur;
out = std::copy((*this)[ 0 ].first, (*this)[ 0 ].second, out);
}
else if(BOOST_XPR_CHAR_(char_type, '`') == *cur) // prefix
{
++cur;
out = std::copy(this->prefix().first, this->prefix().second, out);
}
else if(BOOST_XPR_CHAR_(char_type, '\'') == *cur) // suffix
{
++cur;
out = std::copy(this->suffix().first, this->suffix().second, out);
}
else if(-1 != traits.value(*cur, 10)) // a sub-match
{
int max = static_cast<int>(this->size() - 1);
int br_nbr = detail::toi(cur, end, traits, 10, max);
detail::ensure(0 != br_nbr, regex_constants::error_subreg, "invalid back-reference");
out = std::copy((*this)[ br_nbr ].first, (*this)[ br_nbr ].second, out);
}
else
{
*out++ = BOOST_XPR_CHAR_(char_type, '$');
*out++ = *cur++;
}
}
return out;
}
user_string convert<user_string>(const string_type & source,
system::error_code & ec)
{
user_string temp;
for (string_type::const_iterator it = source.begin();
it != source.end(); ++it)
temp += *it - 1;
return temp;
}
//! Applies string replacements starting from the specified position
result_type operator() (string_type& str, typename string_type::size_type start_pos = 0) const
{
base_type::operator() (str, start_pos);
typedef typename string_type::iterator string_iterator;
for (string_iterator it = str.begin() + start_pos, end = str.end(); it != end; ++it)
{
char_type c = *it;
if (c < 0x20 || c > 0x7e)
{
char_type buf[(std::numeric_limits< char_type >::digits + 3) / 4 + 3];
std::size_t n = traits_type::print_escaped(buf, c);
std::size_t pos = it - str.begin();
str.replace(pos, 1, buf, n);
it = str.begin() + n - 1;
end = str.end();
}
}
}
//! helper function to put the various member string into stream
OutItrT put_string(OutItrT next, const string_type& str) const
{
typename string_type::const_iterator itr = str.begin();
while(itr != str.end()) {
*next = *itr;
++itr;
++next;
}
return next;
}
mime_content_type parse_content_type ( const string_type &theHeader ) {
tracer t ( __func__ );
mime_content_type retVal;
typename string_type::const_iterator first = theHeader.begin ();
mime_content_type_parser<typename string_type::const_iterator> ct_parser;
bool b = qi::parse ( first, theHeader.end (), ct_parser, retVal );
if (!b)
throw mime_parsing_error ( "Failed to parse the 'Content-Type' header" );
return retVal;
}
void parse_input_line(const string_type& s)
{
// set matches back to starting values:
for(int i = 0; i < MAX_MATCHES; ++i)
{
matches[i] = -2;
}
parse_function op;
do_test = false;
regex_grep(op, s.begin(), s.end(), parse_expression);
jm_trace("expression: " << make_narrow(expression).c_str());
jm_trace("search string: " << make_narrow(search_text).c_str());
}
/**
@brief appends a string (inserts it at the end)
*/
void append_string(const string_type & str) {
if ( m_reserve_append < (int)str.size()) {
int new_reserve_append = (int)str.size() + m_grow_size ;
resize_string( m_reserve_prepend, new_reserve_append);
}
BOOST_ASSERT(m_reserve_append >= (int)str.size() );
int start_idx = (int)m_str.size() - m_reserve_append;
std::copy(str.begin(), str.end(), m_str.begin() + start_idx);
m_reserve_append -= (int)str.size();
m_full_msg_computed = false;
}
/**
@brief appends a string (inserts it at the end)
*/
void append_string(const string_type & str) {
if ( m_reserve_append < str.size()) {
std::size_t new_reserve_append = str.size() + m_grow_size ;
resize_string( m_reserve_prepend, new_reserve_append);
}
BOOST_ASSERT(m_reserve_append >= str.size());
typename string_type::difference_type start_idx = static_cast<typename string_type::difference_type>(m_str.size() - m_reserve_append);
std::copy(str.begin(), str.end(), m_str.begin() + start_idx);
m_reserve_append -= str.size();
m_full_msg_computed = false;
}
static bool equals(const string_type& str1, const string_type& str2) {
if(str1.size() != str2.size()) {
return false;
} else if(str1.empty() && str2.empty()) {
return true;
} else {
return std::equal(
str1.begin(),
str1.end(),
str2.begin()
);
}
}
bool parse(scanner_t& scanner, skip_t& ) const
{
scanner.save();
for (typename string_type::const_iterator itor = literal.begin();
itor != literal.end(); ++itor)
{
const int val = scanner.read();
if (val == -1 ||
traits_type::to_char_type(val) != *itor)
{
scanner.rollback();
return false;
}
}
scanner.commit();
return true;
}
DecodedBBox decode_bbox(const string_type & _hash_string)
{
// Copy of the string down-cased
// Wish this was ruby, then it would be simple: _hash_string.downcase();
string_type hash_string(_hash_string);
std::transform(
_hash_string.begin(),
_hash_string.end(),
hash_string.begin(),
::tolower);
DecodedBBox output;
output.maxlat = 90;
output.maxlon = 180;
output.minlat = -90;
output.minlon = -180;
bool islon = true;
for(int i = 0, max = hash_string.length(); i < max; i++) {
int char_index = base32_codes_index_of(hash_string[i]);
for (int bits = 4; bits >= 0; --bits) {
int bit = (char_index >> bits) & 1;
if (islon) {
double mid = (output.maxlon + output.minlon) / 2;
if(bit == 1) {
output.minlon = mid;
} else {
output.maxlon = mid;
}
} else {
double mid = (output.maxlat + output.minlat) / 2;
if(bit == 1) {
output.minlat = mid;
} else {
output.maxlat = mid;
}
}
islon = !islon;
}
}
return output;
}
bool is_c(Char e) {
char_eq f(e);
return std::find_if(c_.begin(),c_.end(),f)!=c_.end();
}
bool is_escape(Char e) {
char_eq f(e);
return std::find_if(escape_.begin(),escape_.end(),f)!=escape_.end();
}
/// Returns the end of the contained sequence of the given string
static const_iterator_type end(string_type const& s)
{
return s.end();
}
void set_string(const string_type & str) {
m_str.resize( str.size() + m_reserve_prepend + m_reserve_append);
std::copy( str.begin(), str.end(), m_str.begin() + m_reserve_prepend);
m_full_msg_computed = false;
}
//! Converts escape sequences to the corresponding characters
void char_constants< wchar_t >::translate_escape_sequences(string_type& str)
{
using namespace std; // to make sure we can use C functions unqualified
string_type::iterator it = str.begin();
while (it != str.end())
{
it = std::find(it, str.end(), L'\\');
if (std::distance(it, str.end()) >= 2)
{
it = str.erase(it);
switch (*it)
{
case L'n':
*it = L'\n'; break;
case L'r':
*it = L'\r'; break;
case L'a':
*it = L'\a'; break;
case L'\\':
++it; break;
case L't':
*it = L'\t'; break;
case L'b':
*it = L'\b'; break;
case L'x':
{
string_type::iterator b = it;
if (std::distance(++b, str.end()) >= 2)
{
char_type c1 = *b++, c2 = *b++;
if (iswxdigit(c1) && iswxdigit(c2))
{
*it++ = char_type((to_number(c1) << 4) | to_number(c2));
it = str.erase(it, b);
}
}
break;
}
case L'u':
{
string_type::iterator b = it;
if (std::distance(++b, str.end()) >= 4)
{
char_type c1 = *b++, c2 = *b++, c3 = *b++, c4 = *b++;
if (iswxdigit(c1) && iswxdigit(c2) && iswxdigit(c3) && iswxdigit(c4))
{
*it++ = char_type(
(to_number(c1) << 12) |
(to_number(c2) << 8) |
(to_number(c3) << 4) |
to_number(c4));
it = str.erase(it, b);
}
}
break;
}
case L'U':
{
string_type::iterator b = it;
if (std::distance(++b, str.end()) >= 8)
{
char_type c1 = *b++, c2 = *b++, c3 = *b++, c4 = *b++;
char_type c5 = *b++, c6 = *b++, c7 = *b++, c8 = *b++;
if (iswxdigit(c1) && iswxdigit(c2) && iswxdigit(c3) && iswxdigit(c4) &&
iswxdigit(c5) && iswxdigit(c6) && iswxdigit(c7) && iswxdigit(c8))
{
*it++ = char_type(
(to_number(c1) << 28) |
(to_number(c2) << 24) |
(to_number(c3) << 20) |
(to_number(c4) << 16) |
(to_number(c5) << 12) |
(to_number(c6) << 8) |
(to_number(c7) << 4) |
to_number(c8));
it = str.erase(it, b);
}
}
break;
}
default:
{
if (*it >= L'0' && *it <= L'7')
{
string_type::iterator b = it;
int c = (*b++) - L'0';
if (*b >= L'0' && *b <= L'7')
c = c * 8 + (*b++) - L'0';
if (*b >= L'0' && *b <= L'7')
c = c * 8 + (*b++) - L'0';
*it++ = char_type(c);
it = str.erase(it, b);
}
break;
}
}
}
}
}
inline void write(const string_type & src, string_type & dest) {
dest.insert( dest.begin(), src.begin(), src.end() );
}
void set_string(const string_type & str) {
m_str.resize(str.size() + m_reserve_prepend + m_reserve_append);
std::copy( str.begin(), str.end(), m_str.begin() + static_cast<typename string_type::difference_type>(m_reserve_prepend));
m_full_msg_computed = false;
}
bool is_quote(Char e) {
char_eq f(e);
return std::find_if(quote_.begin(),quote_.end(),f)!=quote_.end();
}
/*! Take a line from the csv, turn it into a time_zone_type,
* and add it to the map. Zone_specs in csv file are expected to
* have eleven fields that describe the time zone. Returns true if
* zone_spec successfully added to database */
bool parse_string(string_type& s)
{
std::vector<string_type> result;
typedef boost::token_iterator_generator<boost::escaped_list_separator<charT>, string_type::const_iterator, string_type >::type token_iter_type;
token_iter_type i = boost::make_token_iterator<string_type>(s.begin(), s.end(),boost::escaped_list_separator<charT>());
token_iter_type end;
while (i != end) {
result.push_back(*i);
i++;
}
enum db_fields { ID, STDABBR, STDNAME, DSTABBR, DSTNAME, GMTOFFSET,
DSTADJUST, START_DATE_RULE, START_TIME, END_DATE_RULE,
END_TIME, FIELD_COUNT
};
//take a shot at fixing gcc 4.x error
const unsigned int expected_fields = static_cast<unsigned int>(FIELD_COUNT);
if (result.size() != expected_fields) {
std::stringstream msg;
msg << "Expecting " << FIELD_COUNT << " fields, got "
<< result.size() << " fields in line: " << s;
throw bad_field_count(msg.str());
}
// initializations
bool has_dst = true;
if(result[DSTABBR] == std::string()) {
has_dst = false;
}
// start building components of a time_zone
time_zone_names names(result[STDNAME], result[STDABBR],
result[DSTNAME], result[DSTABBR]);
time_duration_type utc_offset =
posix_time::duration_from_string(result[GMTOFFSET]);
dst_adjustment_offsets adjust(time_duration_type(0,0,0),
time_duration_type(0,0,0),
time_duration_type(0,0,0));
boost::shared_ptr<rule_type> rules;
if(has_dst) {
adjust = dst_adjustment_offsets(
posix_time::duration_from_string(result[DSTADJUST]),
posix_time::duration_from_string(result[START_TIME]),
posix_time::duration_from_string(result[END_TIME])
);
rules =
boost::shared_ptr<rule_type>(parse_rules(result[START_DATE_RULE],
result[END_DATE_RULE]));
}
string_type id(result[ID]);
boost::shared_ptr<time_zone_base_type> zone(new time_zone_type(names, utc_offset, adjust, rules));
return (add_record(id, zone));
}
date_type
parse_date(std::istreambuf_iterator<charT>& sitr,
std::istreambuf_iterator<charT>& stream_end,
string_type format) const
{
bool use_current_char = false;
charT current_char = *sitr;
unsigned short year(0), month(0), day(0), day_of_year(0);
const_itr itr(format.begin());
while (itr != format.end() && (sitr != stream_end)) {
if (*itr == '%') {
itr++;
if (*itr != '%') { //ignore '%%'
unsigned short i = 0;
switch(*itr) {
case 'a':
{
//this value is just throw away. It could be used for
//error checking potentially, but it isn't helpful in
//actually constructing the date - we just need to get it
//out of the stream
match_results mr = m_weekday_short_names.match(sitr, stream_end);
unsigned int wkday = mr.current_match;
if (mr.has_remaining()) {
current_char = mr.last_char();
use_current_char = true;
}
break;
}
case 'A':
{
//this value is just throw away. It could be used for
//error checking potentially, but it isn't helpful in
//actually constructing the date - we just need to get it
//out of the stream
match_results mr = m_weekday_long_names.match(sitr, stream_end);
unsigned int wkday = mr.current_match;
if (mr.has_remaining()) {
current_char = mr.last_char();
use_current_char = true;
}
break;
}
case 'b':
{
match_results mr = m_month_short_names.match(sitr, stream_end);
month = mr.current_match;
if (mr.has_remaining()) {
current_char = mr.last_char();
use_current_char = true;
}
break;
}
case 'B':
{
match_results mr = m_month_long_names.match(sitr, stream_end);
month = mr.current_match;
if (mr.has_remaining()) {
current_char = mr.last_char();
use_current_char = true;
}
break;
}
case 'd':
{
day = var_string_to_int<unsigned short, charT>(sitr, 2);
break;
}
case 'j':
{
day_of_year = fixed_string_to_int<unsigned short, charT>(sitr, 3);
break;
}
case 'm':
{
month = var_string_to_int<unsigned short, charT>(sitr, 2);
break;
}
case 'Y':
{
year = fixed_string_to_int<unsigned short, charT>(sitr, 4);
break;
}
case 'y':
{
year = fixed_string_to_int<unsigned short, charT>(sitr, 2);
year += 2000; //make 2 digit years in this century
break;
}
default:
{} //ignore those we don't understand
}//switch
}
itr++; //advance past format specifier
}
else { //skip past chars in format and in buffer
itr++;
if (use_current_char) {
use_current_char = false;
current_char = *sitr;
}
else {
sitr++;
}
}
}
if (day_of_year != 0) {
date_type d(year-1,12,31); //end of prior year
return d + duration_type(day_of_year);
}
return date_type(year, month, day);
}
