kakoune/src/selectors.cc

523 lines
15 KiB
C++
Raw Normal View History

#include "selectors.hh"
#include "string.hh"
2012-10-08 19:12:09 +02:00
#include "utf8_iterator.hh"
#include <algorithm>
namespace Kakoune
{
using Utf8Iterator = utf8::utf8_iterator<BufferIterator, utf8::InvalidBytePolicy::Pass>;
2012-10-08 19:12:09 +02:00
2012-03-07 20:20:32 +01:00
namespace
{
template<bool punctuation_is_word = false>
2012-10-08 19:12:09 +02:00
bool is_word(Codepoint c)
{
2012-09-30 16:23:18 +02:00
return Kakoune::is_word(c);
}
2012-03-07 20:20:32 +01:00
template<>
2012-10-08 19:12:09 +02:00
bool is_word<true>(Codepoint c)
2012-03-07 20:20:32 +01:00
{
return !is_blank(c) and !is_eol(c);
}
2012-10-08 19:12:09 +02:00
static bool is_punctuation(Codepoint c)
{
return not (is_word(c) or is_blank(c) or is_eol(c));
}
enum class CharCategories
{
Blank,
EndOfLine,
Word,
Punctuation,
};
2012-03-07 20:20:32 +01:00
template<bool punctuation_is_word = false>
2012-10-08 19:12:09 +02:00
CharCategories categorize(Codepoint c)
{
if (is_word(c))
return CharCategories::Word;
if (is_eol(c))
return CharCategories::EndOfLine;
if (is_blank(c))
return CharCategories::Blank;
2012-03-07 20:20:32 +01:00
return punctuation_is_word ? CharCategories::Word
: CharCategories::Punctuation;
}
2012-10-08 19:12:09 +02:00
bool is_begin(const BufferIterator& it) { return it.is_begin(); }
bool is_end(const BufferIterator& it) { return it.is_end(); }
bool is_begin(const Utf8Iterator& it) { return it.underlying_iterator().is_begin(); }
bool is_end(const Utf8Iterator& it) { return it.underlying_iterator().is_end(); }
template<typename Iterator, typename T>
2013-01-03 14:30:14 +01:00
void skip_while(Iterator& it, T condition)
{
2012-10-08 19:12:09 +02:00
while (not is_end(it) and condition(*it))
++it;
}
2012-10-08 19:12:09 +02:00
template<typename Iterator, typename T>
2013-01-03 14:30:14 +01:00
void skip_while_reverse(Iterator& it, T condition)
{
2012-10-08 19:12:09 +02:00
while (not is_begin(it) and condition(*it))
--it;
2012-10-08 19:12:09 +02:00
}
Selection utf8_selection(const Utf8Iterator& first, const Utf8Iterator& last)
{
return Selection(first.underlying_iterator(), last.underlying_iterator());
}
2012-03-07 20:20:32 +01:00
}
typedef boost::regex_iterator<BufferIterator> RegexIterator;
2012-03-07 20:20:32 +01:00
template<bool punctuation_is_word>
Selection select_to_next_word(const Selection& selection)
{
2012-10-08 19:12:09 +02:00
Utf8Iterator begin = selection.last();
2012-03-07 20:20:32 +01:00
if (categorize<punctuation_is_word>(*begin) !=
categorize<punctuation_is_word>(*(begin+1)))
++begin;
skip_while(begin, is_eol);
2012-10-08 19:12:09 +02:00
Utf8Iterator end = begin+1;
2012-03-07 20:20:32 +01:00
if (not punctuation_is_word and is_punctuation(*begin))
skip_while(end, is_punctuation);
2012-03-07 20:20:32 +01:00
else if (is_word<punctuation_is_word>(*begin))
skip_while(end, is_word<punctuation_is_word>);
2013-01-03 14:30:14 +01:00
skip_while(end, is_blank);
2013-01-03 14:30:14 +01:00
return utf8_selection(begin, end-1);
}
template Selection select_to_next_word<false>(const Selection&);
template Selection select_to_next_word<true>(const Selection&);
2012-03-07 20:20:32 +01:00
template<bool punctuation_is_word>
Selection select_to_next_word_end(const Selection& selection)
{
2012-10-08 19:12:09 +02:00
Utf8Iterator begin = selection.last();
2012-03-07 20:20:32 +01:00
if (categorize<punctuation_is_word>(*begin) !=
categorize<punctuation_is_word>(*(begin+1)))
++begin;
skip_while(begin, is_eol);
2012-10-08 19:12:09 +02:00
Utf8Iterator end = begin;
skip_while(end, is_blank);
2012-03-07 20:20:32 +01:00
if (not punctuation_is_word and is_punctuation(*end))
2013-01-03 14:30:14 +01:00
skip_while(end, is_punctuation);
2012-03-07 20:20:32 +01:00
else if (is_word<punctuation_is_word>(*end))
2013-01-03 14:30:14 +01:00
skip_while(end, is_word<punctuation_is_word>);
2013-01-03 14:30:14 +01:00
return utf8_selection(begin, end-1);
}
template Selection select_to_next_word_end<false>(const Selection&);
template Selection select_to_next_word_end<true>(const Selection&);
2012-03-07 20:20:32 +01:00
template<bool punctuation_is_word>
Selection select_to_previous_word(const Selection& selection)
{
2012-10-08 19:12:09 +02:00
Utf8Iterator begin = selection.last();
2012-03-07 20:20:32 +01:00
if (categorize<punctuation_is_word>(*begin) !=
categorize<punctuation_is_word>(*(begin-1)))
--begin;
skip_while_reverse(begin, is_eol);
2012-10-08 19:12:09 +02:00
Utf8Iterator end = begin;
skip_while_reverse(end, is_blank);
bool with_end = false;
2012-03-07 20:20:32 +01:00
if (not punctuation_is_word and is_punctuation(*end))
2013-01-03 14:30:14 +01:00
{
skip_while_reverse(end, is_punctuation);
with_end = is_punctuation(*end);
}
2012-03-07 20:20:32 +01:00
else if (is_word<punctuation_is_word>(*end))
2013-01-03 14:30:14 +01:00
{
skip_while_reverse(end, is_word<punctuation_is_word>);
with_end = is_word<punctuation_is_word>(*end);
}
2012-10-08 19:12:09 +02:00
return utf8_selection(begin, with_end ? end : end+1);
}
template Selection select_to_previous_word<false>(const Selection&);
template Selection select_to_previous_word<true>(const Selection&);
Selection select_line(const Selection& selection)
{
2012-10-08 19:12:09 +02:00
Utf8Iterator first = selection.last();
if (*first == '\n' and not is_end(first + 1))
++first;
2012-10-08 19:12:09 +02:00
while (not is_begin(first) and *(first - 1) != '\n')
--first;
2012-10-08 19:12:09 +02:00
Utf8Iterator last = first;
while (not is_end(last + 1) and *last != '\n')
++last;
2012-10-08 19:12:09 +02:00
return utf8_selection(first, last);
}
Selection select_matching(const Selection& selection)
{
2012-10-11 01:17:29 +02:00
std::vector<Codepoint> matching_pairs = { '(', ')', '{', '}', '[', ']', '<', '>' };
2012-10-08 19:12:09 +02:00
Utf8Iterator it = selection.last();
2012-10-11 01:17:29 +02:00
std::vector<Codepoint>::iterator match = matching_pairs.end();
while (not is_eol(*it))
{
match = std::find(matching_pairs.begin(), matching_pairs.end(), *it);
if (match != matching_pairs.end())
break;
++it;
}
if (match == matching_pairs.end())
return selection;
2012-10-08 19:12:09 +02:00
Utf8Iterator begin = it;
if (((match - matching_pairs.begin()) % 2) == 0)
{
int level = 0;
2012-10-11 01:17:29 +02:00
const Codepoint opening = *match;
const Codepoint closing = *(match+1);
2012-10-08 19:12:09 +02:00
while (not is_end(it))
{
if (*it == opening)
++level;
else if (*it == closing and --level == 0)
2012-10-08 19:12:09 +02:00
return utf8_selection(begin, it);
++it;
}
}
else
{
int level = 0;
2012-10-11 01:17:29 +02:00
const Codepoint opening = *(match-1);
const Codepoint closing = *match;
while (true)
{
if (*it == closing)
++level;
else if (*it == opening and --level == 0)
2012-10-08 19:12:09 +02:00
return utf8_selection(begin, it);
if (is_begin(it))
break;
--it;
}
}
return selection;
}
Selection select_surrounding(const Selection& selection,
const CodepointPair& matching,
SurroundFlags flags)
{
const bool to_begin = flags & SurroundFlags::ToBegin;
const bool to_end = flags & SurroundFlags::ToEnd;
const bool nestable = matching.first != matching.second;
2012-10-08 19:12:09 +02:00
Utf8Iterator first = selection.last();
if (to_begin)
{
int level = 0;
while (not is_begin(first))
{
if (nestable and first != selection.last() and *first == matching.second)
++level;
else if (*first == matching.first)
{
if (level == 0)
break;
else
--level;
}
--first;
}
if (level != 0 or *first != matching.first)
return selection;
}
Utf8Iterator last = selection.last();
if (to_end)
{
int level = 0;
last = first + 1;
while (not is_end(last))
{
if (nestable and *last == matching.first)
++level;
else if (*last == matching.second)
{
if (level == 0)
break;
else
--level;
}
++last;
}
if (level != 0 or *last != matching.second)
return selection;
}
if (flags & SurroundFlags::Inner)
{
if (to_begin)
++first;
if (to_end and first != last)
--last;
}
return to_end ? utf8_selection(first, last) : utf8_selection(last, first);
}
Selection select_to(const Selection& selection,
Codepoint c, int count, bool inclusive)
{
2012-10-08 19:12:09 +02:00
Utf8Iterator begin = selection.last();
Utf8Iterator end = begin;
do
{
++end;
2012-10-08 19:12:09 +02:00
skip_while(end, [c](Codepoint cur) { return not is_eol(cur) and cur != c; });
if (is_end(end) or is_eol(*end))
return selection;
}
while (--count > 0);
2012-10-08 19:12:09 +02:00
return utf8_selection(begin, inclusive ? end : end-1);
}
Selection select_to_reverse(const Selection& selection,
Codepoint c, int count, bool inclusive)
{
2012-10-08 19:12:09 +02:00
Utf8Iterator begin = selection.last();
Utf8Iterator end = begin;
do
{
--end;
2012-10-08 19:12:09 +02:00
skip_while_reverse(end, [c](Codepoint cur) { return not is_eol(cur) and cur != c; });
if (is_begin(end) or is_eol(*end))
return selection;
}
while (--count > 0);
2012-10-08 19:12:09 +02:00
return utf8_selection(begin, inclusive ? end : end+1);
}
Selection select_to_eol(const Selection& selection)
{
2012-10-08 19:12:09 +02:00
Utf8Iterator begin = selection.last();
Utf8Iterator end = begin + 1;
skip_while(end, [](Codepoint cur) { return not is_eol(cur); });
return utf8_selection(begin, end-1);
}
Selection select_to_eol_reverse(const Selection& selection)
{
2012-10-08 19:12:09 +02:00
Utf8Iterator begin = selection.last();
Utf8Iterator end = begin - 1;
skip_while_reverse(end, [](Codepoint cur) { return not is_eol(cur); });
return utf8_selection(begin, is_begin(end) ? end : end+1);
}
template<bool punctuation_is_word>
Selection select_whole_word(const Selection& selection, bool inner)
{
2012-10-08 19:12:09 +02:00
Utf8Iterator first = selection.last();
Utf8Iterator last = first;
if (is_word(*first))
{
2013-01-03 14:30:14 +01:00
skip_while_reverse(first, is_word<punctuation_is_word>);
if (not is_word<punctuation_is_word>(*first))
++first;
skip_while(last, is_word<punctuation_is_word>);
if (not inner)
skip_while(last, is_blank);
}
else if (not inner)
{
2013-01-03 14:30:14 +01:00
skip_while_reverse(first, is_blank);
if (is_blank(*first))
++first;
skip_while(last, is_blank);
if (not is_word<punctuation_is_word>(*last))
return selection;
skip_while(last, is_word<punctuation_is_word>);
}
--last;
2012-10-08 19:12:09 +02:00
return utf8_selection(first, last);
}
template Selection select_whole_word<false>(const Selection&, bool);
template Selection select_whole_word<true>(const Selection&, bool);
Selection select_whole_lines(const Selection& selection)
{
2013-01-03 14:30:14 +01:00
// no need to be utf8 aware for is_eol as we only use \n as line seperator
BufferIterator first = selection.first();
BufferIterator last = selection.last();
BufferIterator& to_line_start = first <= last ? first : last;
BufferIterator& to_line_end = first <= last ? last : first;
2013-01-03 14:30:14 +01:00
--to_line_start;
skip_while_reverse(to_line_start, [](char cur) { return not is_eol(cur); });
if (is_eol(*to_line_start))
++to_line_start;
2011-11-22 19:58:05 +01:00
2013-01-03 14:30:14 +01:00
skip_while(to_line_end, [](char cur) { return not is_eol(cur); });
2013-01-03 14:30:14 +01:00
return Selection(first, last);
}
Selection select_whole_buffer(const Selection& selection)
{
const Buffer& buffer = selection.first().buffer();
2012-10-08 19:12:09 +02:00
return Selection(buffer.begin(), utf8::previous(buffer.end()));
}
using MatchResults = boost::match_results<BufferIterator>;
static bool find_last_match(BufferIterator begin, const BufferIterator& end,
MatchResults& res, const Regex& regex)
{
MatchResults matches;
while (boost::regex_search(begin, end, matches, regex))
{
if (begin == matches[0].second)
break;
begin = matches[0].second;
res.swap(matches);
}
return not res.empty();
}
template<bool forward>
bool find_match_in_buffer(const BufferIterator pos, MatchResults& matches,
const Regex& ex)
{
auto bufbeg = pos.buffer().begin();
auto bufend = pos.buffer().end();
if (forward)
return (boost::regex_search(pos, bufend, matches, ex) or
boost::regex_search(bufbeg, pos, matches, ex));
else
return (find_last_match(bufbeg, pos, matches, ex) or
find_last_match(pos, bufend, matches, ex));
}
template<bool forward>
Selection select_next_match(const Selection& selection, const String& regex)
{
try
{
2012-10-08 19:12:09 +02:00
// regex matching do not use Utf8Iterator as boost::regex handle utf8
// decoding itself
BufferIterator begin = selection.last();
BufferIterator end = begin;
CaptureList captures;
Regex ex{regex.begin(), regex.end()};
MatchResults matches;
if (find_match_in_buffer<forward>(utf8::next(begin), matches, ex))
{
begin = matches[0].first;
end = matches[0].second;
for (auto& match : matches)
captures.push_back(String(match.first, match.second));
}
else
throw runtime_error("'" + regex + "': no matches found");
if (begin == end)
++end;
end = utf8::previous(end);
if (not forward)
std::swap(begin, end);
return Selection{begin, end, std::move(captures)};
}
catch (boost::regex_error& err)
{
throw runtime_error(String("regex error: ") + err.what());
}
}
template Selection select_next_match<true>(const Selection&, const String&);
template Selection select_next_match<false>(const Selection&, const String&);
SelectionList select_all_matches(const Selection& selection, const String& regex)
{
try
{
Regex ex(regex.begin(), regex.end());
RegexIterator re_it(selection.begin(), selection.end(), ex);
RegexIterator re_end;
SelectionList result;
for (; re_it != re_end; ++re_it)
{
BufferIterator begin = (*re_it)[0].first;
BufferIterator end = (*re_it)[0].second;
if (begin == selection.end())
continue;
CaptureList captures;
for (auto& match : *re_it)
captures.push_back(String(match.first, match.second));
result.push_back(Selection(begin, begin == end ? end : utf8::previous(end),
std::move(captures)));
}
return result;
}
catch (boost::regex_error& err)
{
throw runtime_error(String("regex error: ") + err.what());
}
}
SelectionList split_selection(const Selection& selection,
const String& separator_regex)
{
try
{
Regex ex(separator_regex.begin(), separator_regex.end());
RegexIterator re_it(selection.begin(), selection.end(), ex,
boost::regex_constants::match_nosubs);
RegexIterator re_end;
SelectionList result;
BufferIterator begin = selection.begin();
for (; re_it != re_end; ++re_it)
{
BufferIterator end = (*re_it)[0].first;
2012-10-08 19:12:09 +02:00
result.push_back(Selection(begin, (begin == end) ? end : utf8::previous(end)));
begin = (*re_it)[0].second;
}
2012-12-27 18:56:06 +01:00
result.push_back(Selection(begin, std::max(selection.first(),
selection.last())));
return result;
}
catch (boost::regex_error& err)
{
throw runtime_error(String("regex error: ") + err.what());
}
}
}