do some source gen work, findseq, matchseq, matchseq framework done, on find set

This commit is contained in:
Youmu Liu
2026-07-16 18:30:50 +08:00
parent ee7cbd869c
commit 0d41058632
5 changed files with 428 additions and 85 deletions
@@ -5,6 +5,7 @@
// SPDX-License-Identifier: MIT
// -----------------------------------------------------------------------------
using Microsoft.CodeAnalysis.CSharp.Syntax;
using StgSharp.RegularAnalysis.Abstraction;
using System;
using System.Collections.Generic;
@@ -20,6 +21,7 @@ namespace StgSharp.RegularAnalysis.Text
{
private const string current_ptr = "current_ptr";
private const string offset = "offset";
private const string orig_text = "text";
internal static SequenceEmitter<string> GenerateSource(
@@ -28,7 +30,7 @@ namespace StgSharp.RegularAnalysis.Text
{
SequenceEmitter<string> se = new();
List<SequenceEmitter<string>> sub_match_list = [];
VarNameGenerator ng = new();
SourceGenContext ng = new();
RegexInfo info = source.Info;
List<RegexIR> ir = source.IRs;
@@ -36,6 +38,28 @@ namespace StgSharp.RegularAnalysis.Text
RegexIR[] window = new RegexIR[8];
Span<RegexIR> scan_win = window;
SequenceEmitter<string> match_e = new();
int index = 0;
do
{
int result;
int count = int.Min(8, ir_sp.Length - index);
ir_sp.Slice(index, count).CopyTo(scan_win);
if (TryGenerateMatchSeq(scan_win, match_e, ng) > 0)
{
count += 1;
} else if (TryGenerateMatchSet(scan_win, match_e, ng) > 0)
{
count++;
} else if (TryGenerateFindSeq(scan_win, match_e, ng) > 0)
{
count++;
} else { }
index += count;
} while (index < ir.Count);
_ = se.AppendLine($@"public override MatchResult Match({ROS_char} text)")//
.AppendLine(@"{");
@@ -47,24 +71,13 @@ namespace StgSharp.RegularAnalysis.Text
.AppendLine($@"bool region_{i}_success;");
}
_ = se.AppendLine($@"int {current_ptr} = 0;");
_ = se.AppendLine($@"int {current_ptr} = 0;")
.AppendLine($@"int {offset} = 0;")
.AppendLine();
int index = 0;
do
{
int result;
_ = ng.GenVarDefine(se);
int count = int.Min(8, ir_sp.Length - index);
ir_sp.Slice(index, count).CopyTo(scan_win);
if (TryGenerateMatchSeq(scan_win, se, ng) > 0)
{
count++;
} else if (TryGenerateMatchSet(scan_win, se, ng) > 0)
{
count++;
} else { }
index += count;
} while (index < ir.Count);
_ = se.Append(match_e);
foreach (SequenceEmitter<string> _sub in sub_match_list)
{
@@ -77,66 +90,171 @@ namespace StgSharp.RegularAnalysis.Text
return se;
}
#region var naming
private class VarNameGenerator
{
private int char_counter;
public string GenerateCharName()
{
string name = $"__char_{char_counter}";
char_counter++;
return name;
}
}
#endregion
#region sub generator
private static int TryGenerateMatchSeq(
private static int TryGenerateFindSeq(
Span<RegexIR> scan_win,
SequenceEmitter<string> se,
VarNameGenerator ng
SourceGenContext ng
)
{
if (scan_win[0] is not RegexMatchSeqIR seq) {
if (scan_win[0] is not RegexFindSeqIR seq) {
return -1;
}
string pattern = seq.Pattern;
int length = pattern.Length;
string index_caller = seq.IsFirst ? "IndexOf" : "LastIndexOf";
_ = se.AppendLine()
.AppendLine($@"if(text[{current_ptr}..({current_ptr} + {length})] == {pattern})")
.AppendLine($@"if(({offset} = [{current_ptr}..].{index_caller}({pattern})) != -1)")
.AppendLine($@"{{")
.AppendLine($@"{current_ptr} += {length};")
.AppendLine($@"{current_ptr} += {offset} + {length};")
.AppendLine($@"}}");
return length;
}
private static int TryGenerateMatchSeq(
Span<RegexIR> scan_win,
SequenceEmitter<string> se,
SourceGenContext ng
)
{
if (scan_win[0] is not RegexFindSeqIR seq) {
return -1;
}
string pattern = seq.Pattern;
int length = pattern.Length;
_ = se.AppendLine()
.AppendLine($@"if([{current_ptr}..({current_ptr} + {length})] == {pattern})")
.AppendLine($@"{{")
.AppendLine($@"{current_ptr} += {offset} + {length};")
.AppendLine($@"}}");
return length;
}
private static int TryGenerateFindSet(
Span<RegexIR> scan_win,
SequenceEmitter<string> se,
SourceGenContext ng
)
{
if (scan_win[0] is not RegexFindSetIR seq) {
return -1;
}
string pattern = seq.Charset;
RegexCharSet[] set_pattern = FigureCharSetType(pattern, out bool accept);
bool first = seq.IsFirst;
string remain = ng.CurRemain;
string offset = ng.FindResult;
_ = se.AppendLine($@"// match charset of {pattern}")
.AppendLine($@"{remain} = {orig_text}[{current_ptr}..];")
.AppendLine();
_ = se.AppendLine("if(");
RegexCharSet p = set_pattern[0];
string _case = accept ? GenerateAcceptCase(p) : GenerateDenyCase(p);
_ = se.AppendLine($@"if ({_case}||");
for (int i = 1; i < set_pattern.Length - 1; i++)
{
p = set_pattern[i];
_case = accept ? GenerateAcceptCase(p) : GenerateDenyCase(p);
if (string.IsNullOrEmpty(_case))
{
continue;
}
_ = se.AppendLine($@" {_case} ||");
}
p = set_pattern[^1];
_case = accept ? GenerateAcceptCase(p) : GenerateDenyCase(p);
_ = se.AppendLine($@"{_case})")
.AppendLine("{")
.AppendLine($@"{current_ptr} += 1;")
.AppendLine("}");
return 1;
string GenerateAcceptCase(
RegexCharSet c
)
{
string caller;
switch (c.Type)
{
// Handle single character
case CharSetType.Single:
caller = first ? "IndexOfAny" : "LastIndexOfAny";
return $@"({offset} = {remain}.{caller}(""{c.Value}"") != -1)";
// Handle range of characters
case CharSetType.Range:
caller = first ? "IndexOfAnyInRange" : "LastIndexOfAnyInRange";
return $@"({offset} = {remain}.{caller}('{c.Value[0]}', '{c.Value[1]}') != -1)";
// Handle set of characters
case CharSetType.Set:
switch (c.Value)
{
case "s":
caller = first ? "IndexOfAnyWhiteSpace" : "LastIndexOfAnyWhiteSpace";
return $@"({offset} = {remain}.{caller}() != -1)";
case "w":
default:
break;
}
case CharSetType.None:
return string.Empty;
default:
return string.Empty;
}
}
string GenerateDenyCase(
RegexCharSet c
)
{
switch (c.Type)
{
// Handle single character
case CharSetType.Single:
// Handle range of characters
case CharSetType.Range:
// Handle set of characters
case CharSetType.Set:
case CharSetType.None:
return string.Empty;
default:
return string.Empty;
}
}
}
private static int TryGenerateMatchSet(
Span<RegexIR> scan_win,
SequenceEmitter<string> se,
VarNameGenerator ng
SourceGenContext ng
)
{
if (scan_win[0] is not RegexMatchSeqIR seq) {
return -1;
}
string pattern = seq.Pattern;
RegexCharSet[] set_pattern = FigureCharSetType(pattern);
RegexCharSet[] set_pattern = FigureCharSetType(pattern, out bool accept);
string var_char = ng.GenerateCharName();
string var_char = ng.CurChar;
_ = se.AppendLine($@"// match charset of {pattern}")
.AppendLine($@"char {var_char} = {orig_text}[{current_ptr}];")
.AppendLine();
bool accept = set_pattern[0].Accept;
_ = se.AppendLine("if(");
RegexCharSet p = set_pattern[0];
string _case = GenerateCase(p);
@@ -164,29 +282,63 @@ namespace StgSharp.RegularAnalysis.Text
return 1;
string GenerateCase(
RegexCharSet c
)
{
return c.Type switch
switch (c.Type)
{
// Handle single character
CharSetType.Single => $@"{var_char} == '{c.Value}'",
case CharSetType.Single:
string set = c.Value;
if (set.Length == 1)
{
return $@"{var_char} == '{c.Value}'";
} else if (set.Length < 5)
{
StringBuilder sb = new();
_ = sb.Append('(');
foreach (char item in set) {
_ = sb.Append(CultureInfo.InvariantCulture, $@"item == '{item}' or ");
}
_ = sb.Append(')');
return sb.ToString();
} else
{
return $@"""{c.Value}"".AsSpan().Contains({var_char})";
}
// Handle range of characters
CharSetType.Range => $@"({var_char} >= '{c.Value[0]}' && {var_char} <= '{c.Value[1]}')",
case CharSetType.Range:
return $@"({var_char} >= '{c.Value[0]}' && {var_char} <= '{c.Value[1]}')";
// Handle set of characters
CharSetType.Set => c.Value switch
{
"s" => $@"char.IsWhiteSpace({var_char})",
"w" => $@"char.IsLetterOrDigit({var_char}) || {var_char} == '_'",
"d" => $@"char.IsDigit({var_char})",
_ => throw new InvalidCastException(),
},
CharSetType.None => string.Empty,
_ => string.Empty,
};
case CharSetType.Set:
if (c.Accept)
{
return c.Value switch
{
"s" => $@"char.IsWhiteSpace({var_char})",
"w" => $@"char.IsLetterOrDigit({var_char}) || {var_char} == '_'",
"d" => $@"char.IsDigit({var_char})",
_ => throw new InvalidCastException(),
};
} else
{
return c.Value switch
{
"s" => $@"!char.IsWhiteSpace({var_char})",
"w" => $@"!char.IsLetterOrDigit({var_char}) && {var_char} != '_'",
"d" => $@"!char.IsDigit({var_char})",
_ => throw new InvalidCastException(),
};
}
case CharSetType.None:
return string.Empty;
default:
return string.Empty;
}
}
}
@@ -218,12 +370,12 @@ namespace StgSharp.RegularAnalysis.Text
}
public static RegexCharSet[] FigureCharSetType(
ReadOnlySpan<char> source
private static RegexCharSet[] FigureCharSetType(
ReadOnlySpan<char> source,
out bool accept
)
{
// range or single
bool accept;
if (source[0] == '[')
{
int idx;
@@ -232,7 +384,7 @@ namespace StgSharp.RegularAnalysis.Text
ReadOnlySpan<char> seq = source.Slice(idx, source.Length - idx - 1);
if (seq.IndexOf('-') == -1)
{
return [new RegexCharSet(seq.ToString(), CharSetType.Single, accept)];
return [ new RegexCharSet(seq.ToString(), CharSetType.Single, true) ];
} else
{
List<RegexCharSet> list = [];
@@ -241,17 +393,18 @@ namespace StgSharp.RegularAnalysis.Text
int pos = seq.IndexOf('-');
if (pos == -1)
{
RegexCharSet single = new(seq.ToString(), CharSetType.Single, accept);
list.Add(single);
// has no range
SplitSet(seq, list);
break;
} else
if (pos > 1) {
list.Add(new(seq[0..(pos - 1)].ToString(), CharSetType.Single, accept));
} else if (pos > 1)
{
// has other sequence before range
SplitSet(seq[0..(pos - 1)], list);
}
list.Add(new($"{seq[pos - 1]}{seq[pos + 1]}", CharSetType.Range, accept));
list.Add(new($"{seq[pos - 1]}{seq[pos + 1]}", CharSetType.Range, true));
seq = seq[(pos + 2)..];
}
return [.. list];
return [ .. list ];
}
} else if (source[0] == '\\')
{
@@ -260,9 +413,9 @@ namespace StgSharp.RegularAnalysis.Text
return char.ToLower(source[0], CultureInfo.InvariantCulture)
switch
{
's' => [new("s", CharSetType.Set, accept)],
'w' => [new("w", CharSetType.Set, accept)],
'd' => [new("d", CharSetType.Set, accept)],
's' => [ new("s", CharSetType.Set, accept) ],
'w' => [ new("w", CharSetType.Set, accept) ],
'd' => [ new("d", CharSetType.Set, accept) ],
_ => throw new NotSupportedException()
};
} else
@@ -270,6 +423,40 @@ namespace StgSharp.RegularAnalysis.Text
// unknown, throw
throw new InvalidCastException($"Unknown char set: {source}");
}
static void SplitSet(
ReadOnlySpan<char> seq,
List<RegexCharSet> sets
)
{
if (seq.Length == 0 || sets == null) {
return;
}
int pos;
ReadOnlySpan<char> remain = seq;
while ((pos = remain.IndexOf('\\')) != -1 || remain.Length > 0)
{
if (pos != 0)
{
ReadOnlySpan<char> singles = remain[0..pos];
sets.Add(new RegexCharSet(singles.ToString(), CharSetType.Single, true));
}
bool accept = char.IsAsciiLetterUpper(remain[pos + 1]);
RegexCharSet set_set = char.ToLower(remain[pos + 1], CultureInfo.InvariantCulture)
switch
{
's' => new("s", CharSetType.Set, accept),
'w' => new("w", CharSetType.Set, accept),
'd' => new("d", CharSetType.Set, accept),
_ => throw new NotSupportedException()
};
sets.Add(set_set);
remain = remain[(pos + 2)..];
}
if (remain.Length > 0) {
sets.Add(new RegexCharSet(remain.ToString(), CharSetType.Single, true));
}
}
}
#endregion