std::shared_ptr<dataset> generic_sample_factory::create_sample(const std::string & path_sample) {
std::shared_ptr<dataset> sample_data(new dataset);
size_t sample_dimension = 0;
std::ifstream file_sample(path_sample);
if (file_sample.is_open()) {
std::string file_line;
while(std::getline(file_sample, file_line)) {
if (file_line.empty())
continue;
double value = 0.0;
point sample_point;
std::istringstream stream_value(file_line);
while(stream_value >> value) {
sample_point.push_back(value);
}
if (sample_dimension == 0) {
sample_dimension = sample_point.size();
}
else {
if (sample_dimension != sample_point.size()) {
throw std::runtime_error("Points from input data set should have the same dimension.");
}
}
sample_data->push_back(sample_point);
}
}
void graphene::sample()
{
ofstream file_sample("sample", ios_base::app);
for(int i=0; i<TOT; i++)
{
file_sample<<setprecision(15)<<p[i]<<endl;
}
file_sample.close();
}
bool strref_samples()
{
// separating full path into name, extension, directory
strref file_sample("c:\\windows\\system32\\autochk.exe");
strref file_ext = file_sample.after_last('.'); // file extension
strref file_name = file_sample.after_last_or_full('\\', '/'); // file name including extension
strref file_name_no_ext = file_name.before_last('.'); // file name excluding extension
strref file_dir = file_sample.before_last('\\', '/'); // file path excluding name
printf("\nfull path: " STRREF_FMT "\nextension: " STRREF_FMT "\nname: " STRREF_FMT
"\nname no ext: " STRREF_FMT "\ndirectory: " STRREF_FMT "\n\n",
STRREF_ARG(file_sample), STRREF_ARG(file_ext), STRREF_ARG(file_name),
STRREF_ARG(file_name_no_ext), STRREF_ARG(file_dir));
// fnv1a helper
unsigned int hash = file_sample.fnv1a();
printf("fnv1a hash of \"" STRREF_FMT "\" is 0x%08x\n", STRREF_ARG(file_sample), hash);
// find substrings in text
strref text("Many gnomes don't live underwater, this is a widely appreciated convenience among fish.");
// case insensitive substring find
int pos = text.find(strref("THIS"));
if (pos < 0 || text[pos] != 't')
return false;
// find by rolling hash is an option, may be a performance improvement in large text blocks
pos = text.find_rh_case(strref("Many"));
if (pos < 0 || text[pos] != 'M')
return false;
pos = text.find_rh_case(strref("widely"));
if (pos < 0 || text[pos] != 'w')
return false;
// search character by character, usually performing well
pos = text.find_case("Many");
if (pos < 0 || text[pos] != 'M')
return false;
// find the position of a secondary search
pos = text.find_after_last(' ', 'i');
if (pos < 0 || text[pos] != 'i')
return false;
// find last string match
pos = text.find_last("on");
if (pos < 0 || text[pos] != 'o')
return false;
pos = text.find(',');
if (pos < 0 || text[pos] != ',')
return false;
pos = text.find_after('i', (strl_t)pos);
if (pos < 0 || (text[pos] != 'i' && pos <= text.find('i')))
return false;
pos = (int)text.find_or_full(';', 10);
if ((strl_t)pos != text.get_len())
return false;
pos = (int)text.find_or_full('.', 10);
if (pos <= 0 || text[pos] != '.')
return false;
pos = text.find(strref("this"));
if (pos < 0 || text[pos] != 't')
return false;
pos = text.find("among");
if (pos < 0 || text[pos] != 'a')
return false;
pos = text.find_case("This");
if (pos >= 0)
return false;
pos = text.find_case("this");
if (pos < 0 || text[pos] != 't')
return false;
pos = text.find_last(strref("fi"));
if (pos < 0 || text[pos] != 'f')
return false;
pos = text.find_last("on");
if (pos < 0 || text[pos] != 'o')
return false;
return true;
}
