// TODO US-ASCII support only, no UTF-8 support
// While UTF-8 might work in some cases, we do not guarantee full functionality
template <typename StringView> inline auto decompose(const StringView &lhs, const StringView &rhs)
{
auto const lcs = longest_common_substring(lhs, rhs);
// trim spaces, transform to lower
const auto trim = [](StringView view) {
// we compare suffixes based on this value, it might break UTF chars, but as long as we are
// consistent in handling, we do not create bad results
std::string str = boost::to_lower_copy(view.to_string());
auto front = str.find_first_not_of(" ");
if (front == std::string::npos)
return str;
auto back = str.find_last_not_of(" ");
return str.substr(front, back - front + 1);
};
if (lcs.empty())
{
return std::make_tuple(trim(lhs), trim(rhs), std::string(), std::string());
}
// find the common substring in both
auto lhs_pos = lhs.find(lcs);
auto rhs_pos = rhs.find(lcs);
BOOST_ASSERT(lhs_pos + lcs.size() <= lhs.size());
BOOST_ASSERT(rhs_pos + lcs.size() <= rhs.size());
// prefixes
auto lhs_prefix = (lhs_pos > 0) ? lhs.substr(0, lhs_pos) : StringView();
auto rhs_prefix = (rhs_pos > 0) ? rhs.substr(0, rhs_pos) : StringView();
// suffices
auto lhs_suffix = lhs.substr(lhs_pos + lcs.size());
auto rhs_suffix = rhs.substr(rhs_pos + lcs.size());
return std::make_tuple(trim(lhs_prefix), trim(lhs_suffix), trim(rhs_prefix), trim(rhs_suffix));
}
void EnzymaticDigestion::digestUnmodifiedString(const StringView sequence, std::vector<StringView>& output, Size min_length, Size max_length) const
{
// initialization
output.clear();
// naive cleavage sites
std::vector<Size> pep_positions = tokenize_(sequence.getString());
Size count = pep_positions.size();
// disable max length filter by setting to maximum length
if (max_length == 0)
{
max_length = sequence.size();
}
// no cleavage sites? return full string
if (count == 0)
{
if (sequence.size() >= min_length && sequence.size() <= max_length)
{
output.push_back(sequence);
}
return;
}
for (Size i = 1; i != count; ++i)
{
// add if cleavage product larger then min length
Size l = pep_positions[i] - pep_positions[i - 1];
if (l >= min_length && l <= max_length)
{
output.push_back(sequence.substr(pep_positions[i - 1], pep_positions[i] - 1));
}
}
// add last cleavage product (need to add because end is not a cleavage site) if larger then min length
Size l = sequence.size() - pep_positions[count - 1];
if (l >= min_length && l <= max_length)
{
output.push_back(sequence.substr(pep_positions[count - 1], sequence.size() - 1));
}
// generate fragments with missed cleavages
for (Size i = 1; ((i <= missed_cleavages_) && (i < count)); ++i)
{
for (Size j = 1; j < count - i; ++j)
{
Size l = pep_positions[j + i] - pep_positions[j - 1];
if (l >= min_length && l <= max_length)
{
output.push_back(sequence.substr(pep_positions[j - 1], pep_positions[j + i] - 1));
}
}
// add last cleavage product (need to add because end is not a cleavage site)
Size l = sequence.size() - pep_positions[count - i - 1];
if (l >= min_length && l <= max_length)
{
output.push_back(sequence.substr(pep_positions[count - i - 1], sequence.size() - 1 ));
}
}
}