youmuliu/stgsharp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

add more ir opt code, clearified what to opt, so current opt code may need rewrite

Browse Source
This commit is contained in:
Youmu Liu
2026-06-24 18:09:05 +08:00
parent 2b05d626c7
commit 45f9b04b6d
4 changed files with 121 additions and 54 deletions
Download Patch File Download Diff File Expand all files Collapse all files
StgSharp.RegularAnalysis/Text/RegexInterpreter.IROptimize.cs
+96 -21
View File
@@ -28,6 +28,7 @@
using System;
using System.Collections.Generic;
using System.Linq;
using System.Runtime.InteropServices;
using System.Text;
using System.Threading.Tasks;
@@ -36,19 +37,22 @@ namespace StgSharp.RegularAnalysis.Text
public partial class RegexAnalyzer
{
internal static RegexInfo ScanAndOptimizeIR(
List<RegexIR> baseIR,
out List<RegexIR> optIR
// this array is not thread safe so opt process must be done in single thread.
private RegexIR[] _opt_buffer = new RegexIR[16];
internal RegexInfo ScanAndOptimizeIR(
List<RegexIR> baseIR,
out List<RegexIR> optIR
)
{
int index = 0;
Span<RegexIR> origin_span = CollectionsMarshal.AsSpan(baseIR);
List<RegexIR> optimizedIR = new(baseIR.Count);
List<RegexIR> segment = [];
RegexInfo info = new();
do
{
// TODO scan and opt here
OptimizeIR(baseIR, ref index, segment, info);
OptimizeIR(origin_span, ref index, segment, info);
optimizedIR.AddRange(segment);
index++;
} while (index < baseIR.Count);
@@ -56,37 +60,106 @@ namespace StgSharp.RegularAnalysis.Text
return info;
}
private static void OptimizeDot(
RegexCountSetIR root,
List<RegexIR> ir,
ref int index,
List<RegexIR> optimizedIR,
RegexInfo info
private int ExpandCount(
string pattern,
int min,
int max,
bool is_greedy,
in Span<RegexIR> span
)
{
// expand count operation to less cycles.
const int fast_seq_length = 32;
int fast_count = fast_seq_length / pattern.Length;
fast_count = int.Min(fast_count, min);
if (fast_count > 0)
{
string fast_pattern = string.Concat(Enumerable.Repeat(pattern, fast_count));
int fast_min = min / fast_count;
int consumed_count = fast_min * fast_count;
int remain_min = min - consumed_count;
int remain_max = max - consumed_count;
remain_max = int.Max(-1, remain_max); //for negative stands for infinitive
if (remain_max == 0)
{
// a{2}
RegexCountSeqIR fast = new(fast_min, fast_min, is_greedy, fast_pattern);
span[0] = fast;
return 1;
} else
{
// a{2,5}
RegexCountSeqIR fast = new(fast_min, fast_min, is_greedy, fast_pattern);
RegexCountSeqIR tail = new(remain_min, remain_max, is_greedy, pattern);
span[0] = fast;
span[1] = tail;
return 2;
}
} else if (fast_count == 0)
{
// least count is 0
span[0] = new RegexCountSeqIR(0, max, is_greedy, pattern);
return 1;
} else
{
// least count is infinitive
span[0] = new RegexCountSeqIR(-1, -1, is_greedy, pattern);
return 1;
}
}
private void OptimizeDot(
RegexCountSetIR root,
Span<RegexIR> ir,
ref int index,
List<RegexIR> optimizedIR,
RegexInfo info
)
{
int next = index + 1;
if (next >= ir.Count)
optimizedIR.Clear();
if (next >= ir.Length)
{
optimizedIR.Clear();
optimizedIR.Add(root);
return;
}
RegexIR nextIR = ir[next];
if (nextIR.Command == RegexIRCommand.Match_Seq)
if (nextIR is RegexMatchSeqIR match_seq)
{
// opt .*abc style
RegexFindSeqIR find = new((nextIR as RegexMatchSeqIR)!.Pattern);
RegexFindSeqIR find = new(match_seq.Pattern);
optimizedIR.Add(find);
info.MinLength += find.Pattern.Length;
return;
} else { }
} else if (nextIR is RegexCountSeqIR count_seq)
{
// opt .* (abc)* or .*(abc){2,5} style
string pattern = count_seq.Sequence;
int min = count_seq.Min;
int max = count_seq.Max;
switch ((min, max))
{
case (0, _):
break;
default:
RegexFindSeqIR find = new(pattern);
// RegexCountSeqIR
break;
}
}
}
private static void OptimizeIR(
List<RegexIR> ir,
ref int index,
List<RegexIR> optimizedIR,
RegexInfo info
private void OptimizeIR(
Span<RegexIR> ir,
ref int index,
List<RegexIR> optimizedIR,
RegexInfo info
)
{
optimizedIR.Clear();
@@ -120,6 +193,8 @@ namespace StgSharp.RegularAnalysis.Text
public int TryDepth { get; set; }
public int GroupCount { get; set; }
}
}