xenia/kakoune
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Simplify Quantifier logic in regex parsing

Browse Source 
Remove redundant type enum
This commit is contained in:
Maxime Coste 2024-03-12 20:20:12 +11:00
parent e06acd3dc8
commit 6264b049d9
1 changed files with 16 additions and 30 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

46
src/regex_impl.cc
View File

@ -46,31 +46,15 @@ struct ParsedRegex
struct Quantifier struct Quantifier
{ {
enum Type : char static constexpr int16_t infinite = std::numeric_limits<int16_t>::max();
{
One, int16_t min = 0, max = 0;
Optional,
RepeatZeroOrMore,
RepeatOneOrMore,
RepeatMinMax,
};
Type type = One;
bool greedy = true; bool greedy = true;
int16_t min = -1, max = -1;
bool allows_none() const bool allows_none() const { return min == 0; }
{ bool allows_infinite_repeat() const { return max == infinite; };
return type == Quantifier::Optional or
type == Quantifier::RepeatZeroOrMore or
(type == Quantifier::RepeatMinMax and min <= 0);
}
bool allows_infinite_repeat() const friend bool operator==(Quantifier, Quantifier) = default;
{
return type == Quantifier::RepeatZeroOrMore or
type == Quantifier::RepeatOneOrMore or
(type == Quantifier::RepeatMinMax and max < 0);
};
}; };
using NodeIndex = int16_t; using NodeIndex = int16_t;
@ -554,7 +538,7 @@ private:
ParsedRegex::Quantifier quantifier() ParsedRegex::Quantifier quantifier()
{ {
if (at_end()) if (at_end())
return {ParsedRegex::Quantifier::One}; return {1, 1};
constexpr int max_repeat = 1000; constexpr int max_repeat = 1000;
auto read_bound = [&]() { auto read_bound = [&]() {
@ -580,9 +564,9 @@ private:
switch (*m_pos) switch (*m_pos)
{ {
case '*': ++m_pos; return {ParsedRegex::Quantifier::RepeatZeroOrMore, check_greedy()}; case '*': ++m_pos; return {0, ParsedRegex::Quantifier::infinite, check_greedy()};
case '+': ++m_pos; return {ParsedRegex::Quantifier::RepeatOneOrMore, check_greedy()}; case '+': ++m_pos; return {1, ParsedRegex::Quantifier::infinite, check_greedy()};
case '?': ++m_pos; return {ParsedRegex::Quantifier::Optional, check_greedy()}; case '?': ++m_pos; return {0, 1, check_greedy()};
case '{': case '{':
{ {
++m_pos; ++m_pos;
@ -592,16 +576,18 @@ private:
{ {
++m_pos; ++m_pos;
max = read_bound(); max = read_bound();
if (max == -1)
max = ParsedRegex::Quantifier::infinite;
} }
if (*m_pos++ != '}') if (*m_pos++ != '}')
parse_error("expected closing bracket"); parse_error("expected closing bracket");
return {ParsedRegex::Quantifier::RepeatMinMax, check_greedy(), min, max}; return {min, max, check_greedy()};
} }
default: return {ParsedRegex::Quantifier::One}; default: return {1, 1};
} }
} }
NodeIndex add_node(ParsedRegex::Op op, Codepoint value = -1, ParsedRegex::Quantifier quantifier = {ParsedRegex::Quantifier::One}) NodeIndex add_node(ParsedRegex::Op op, Codepoint value = -1, ParsedRegex::Quantifier quantifier = {1, 1})
{ {
constexpr auto max_nodes = std::numeric_limits<int16_t>::max(); constexpr auto max_nodes = std::numeric_limits<int16_t>::max();
const NodeIndex res = m_parsed_regex.nodes.size(); const NodeIndex res = m_parsed_regex.nodes.size();
@ -641,7 +627,7 @@ private:
to_underlying(Lookaround::OpBegin) <= child.value and to_underlying(Lookaround::OpBegin) <= child.value and
child.value < to_underlying(Lookaround::OpEnd)) child.value < to_underlying(Lookaround::OpEnd))
parse_error("Lookaround does not support literals codepoint between 0xF0000 and 0xFFFFD"); parse_error("Lookaround does not support literals codepoint between 0xF0000 and 0xFFFFD");
if (child.quantifier.type != ParsedRegex::Quantifier::One) if (child.quantifier != ParsedRegex::Quantifier{1, 1})
parse_error("Quantifiers cannot be used in lookarounds"); parse_error("Quantifiers cannot be used in lookarounds");
} }
} }