static void Main(string[] args)
{
try
{
var iterations = 5;
var builder = new StringBuilder();
var haybytes = System.IO.File.ReadAllBytes(@"..\..\mtent12.txt");
var watch = new Stopwatch();
watch.Start();
var haystring = Encoding.ASCII.GetString(haybytes);
var encodetime = watch.Elapsed;
watch.Reset();
Console.WriteLine("Text length: " + haystring.Length);
Console.WriteLine("Encoding time: " + encodetime);
Console.WriteLine();
var testcases = new TestCase[16] {
new TestCase("Twain"),
new TestCase("^Twain"),
new TestCase("Twain$"),
new TestCase("Huck[a-zA-Z]+|Finn[a-zA-Z]+"),
new TestCase("a[^x]{20}b"),
new TestCase("Tom|Sawyer|Huckleberry|Finn"),
new TestCase(".{0,3}(Tom|Sawyer|Huckleberry|Finn)"),
new TestCase("[a-zA-Z]+ing"),
new TestCase("^[a-zA-Z]{0,4}ing[^a-zA-Z]"),
new TestCase("[a-zA-Z]+ing$"),
new TestCase("^[a-zA-Z ]{5,}$"),
new TestCase("^.{16,20}$"),
new TestCase("([a-f](.[d-m].){0,2}[h-n]){2}"),
new TestCase("([A-Za-z]awyer|[A-Za-z]inn)[^a-zA-Z]"),
new TestCase(@"""[^""]{0,30}[?!\.]"""),
new TestCase("Tom.{10,25}river|river.{10,25}Tom")
};
Console.Write("Running 'First Match' test...");
for(int i = 0; i < iterations; i++)
foreach(var testcase in testcases)
{
var re2b = new rr.Regex(testcase.Pattern, rr.RegexOptions.Multiline | rr.RegexOptions.Latin1);
var re2s = new rr.Regex(testcase.Pattern, rr.RegexOptions.Multiline);
var nets = new nn.Regex(testcase.Pattern, nn.RegexOptions.Multiline);
watch.Start();
var re2ByteMatch = re2b.Match(haybytes);
testcase.AddRe2ByteResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
watch.Start();
var re2StringMatch = re2s.Match(haystring);
testcase.AddRe2StringResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
watch.Start();
var netMatch = nets.Match(haystring);
testcase.AddNETResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
if(re2ByteMatch.Value != re2StringMatch.Value)
{
Console.WriteLine();
Console.WriteLine("Match.Value: RE2 bytes failed to match RE2 string for pattern " + re2b.Pattern);
Console.WriteLine("This is not necessarily an error and may be due to accent characters.");
Console.WriteLine("RE2 bytes value: " + re2ByteMatch.Value);
Console.WriteLine("RE2 string value: " + re2StringMatch.Value);
Console.WriteLine();
}
if(re2StringMatch.Value != netMatch.Value)
{
Console.WriteLine();
Console.WriteLine("Match.Value: RE2 string failed to match .NET string for pattern " + re2b.Pattern);
Console.WriteLine("This is not necessarily an error and may be due to accent characters.");
Console.WriteLine("RE2 string value: " + re2StringMatch.Value);
Console.WriteLine(".NET string value: " + netMatch.Value);
Console.WriteLine();
}
Assert.AreEqual(re2StringMatch.Index, netMatch.Index, "Match.Index: RE2 string, NET : " + re2b.Pattern);
Assert.AreEqual(re2StringMatch.Length, netMatch.Length, "Match.Length: RE2 string, NET : " + re2b.Pattern);
}
Console.WriteLine("\n\nResults:\n\n");
PrintByteVsStringResults(testcases);
Console.WriteLine("\n");
PrintStringVsStringResults(testcases);
foreach(var testcase in testcases)
testcase.Reset();
Console.Write("\n\nRunning 'All Matches' test...");
for(int i = 0; i < iterations; i++)
foreach(var testcase in testcases)
{
var re2b = new rr.Regex(testcase.Pattern, rr.RegexOptions.Multiline | rr.RegexOptions.Latin1);
var re2s = new rr.Regex(testcase.Pattern, rr.RegexOptions.Multiline);
var nets = new nn.Regex(testcase.Pattern, nn.RegexOptions.Multiline);
watch.Start();
var re2ByteMatches = re2b.Matches(haybytes);
// Matches() methods are lazily evaluated.
testcase.Re2ByteMatchCount = re2ByteMatches.Count;
testcase.AddRe2ByteResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
watch.Start();
var re2StringMatches = re2s.Matches(haystring);
// Matches() methods are lazily evaluated.
testcase.Re2StringMatchCount = re2StringMatches.Count;
testcase.AddRe2StringResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
watch.Start();
var netMatches = nets.Matches(haystring);
// Matches() methods are lazily evaluated.
testcase.NETMatchCount = netMatches.Count;
testcase.AddNETResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
Assert.AreEqual(re2ByteMatches.Count, re2StringMatches.Count, "Match.Count: RE2 bytes, RE2 string : " + re2b.Pattern);
Assert.AreEqual(re2ByteMatches.Count, netMatches.Count, "Match.Count: RE2 bytes, NET : " + re2b.Pattern);
for(int j = 0; j < re2ByteMatches.Count; j++)
{
if(re2ByteMatches[j].Value != re2StringMatches[j].Value)
{
Console.WriteLine();
Console.WriteLine("Match.Value: RE2 bytes failed to match RE2 string for pattern " + re2b.Pattern);
Console.WriteLine("This is not necessarily an error and may be due to accent characters.");
Console.WriteLine("RE2 bytes value: " + re2ByteMatches[j].Value);
Console.WriteLine("RE2 string value: " + re2StringMatches[j].Value);
Console.WriteLine();
}
if(re2StringMatches[j].Value != netMatches[j].Value)
{
Console.WriteLine();
Console.WriteLine("Match.Value: RE2 string failed to match .NET string for pattern " + re2b.Pattern);
Console.WriteLine("This is not necessarily an error and may be due to accent characters.");
Console.WriteLine("RE2 string value: " + re2StringMatches[j].Value);
Console.WriteLine(".NET string value: " + netMatches[j].Value);
Console.WriteLine();
}
}
}
Console.WriteLine("\n\nResults:\n\n");
PrintByteVsStringResults(testcases);
Console.WriteLine("\n");
PrintStringVsStringResults(testcases);
}
catch(Exception ex)
{
Console.WriteLine(ex.Message + ex.StackTrace);
}
GC.Collect();
Console.WriteLine("Done.");
Console.ReadLine();
}
static void Main(string[] args)
{
try
{
{
Console.WriteLine("Running issue #1 test ...");
string source = "red car white car";
string pattern = @"(\w+)\s+(car)";
MatchCollection matches = Regex.Matches(source, pattern);
Debug.Assert(matches.Count == 2);
Debug.Assert(matches[0].Value == "red car");
Debug.Assert(matches[1].Value == "white car");
Debug.Assert(matches[0].Index == 0);
Debug.Assert(matches[1].Index == 8);
Console.WriteLine("\t... Success.\n");
Console.WriteLine("Running issue #2 test ...");
var r = new rr.Regex(@"\d*");
Debug.Assert(r.Match("123", 1).Index == 1);
Debug.Assert(r.Match("123", 1).Length == 2);
Debug.Assert(r.Match("123", 1).Value == "23");
Console.WriteLine("\t... Success.\n");
Console.WriteLine("Running issue #3 test ...");
int success = 0;
for (uint i = 0; i < 100; ++i)
{
using (var re2 = new Regex("a"))
{
Match match = re2.Match("a");
if (match.Success)
{
success++;
}
}
}
Debug.Assert(success == 100);
Console.WriteLine("\t... Success.\n");
}
{
Console.WriteLine("Running simple tests ...");
// Five matches, each with a value of "".
Debug.Assert(Regex.Matches("xxxx", "").Count == 5);
// Accessing an array (technically, the default Item property) with a negative index? Yup!
Debug.Assert(Regex.Matches("xxxx", "")[0].Groups[-1].Value == "");
// The returned Match containes a Group collection, and the first item in the Group collection is the Match.
Match match = Regex.Match("abcd", "abcd");
Debug.Assert((Group)match == match.Groups[0]);
// .NET allows invalid UTF-16 strings (including freestanding codepoints).
Debug.Assert(Regex.Match("\xD800", "\xD800").Length == 1);
// The translation isn't consistent for invalid strings, though.
Debug.Assert(Regex.Match(Encoding.UTF8.GetBytes("\xD800"), "\xD800") == Match.Empty);
// 2-byte UTF-16 to 3-byte UTF-8.
Debug.Assert(Regex.Match("水DŽ", "水DŽ").Length == 2);
Debug.Assert(Regex.Match(Encoding.UTF8.GetBytes("水DŽ"), "水DŽ").Length == 5);
// To the BMP ... and beyond!
Debug.Assert(Regex.Match("xx𠜎𠜱𠝹𠱓", "𠜎𠜱𠝹𠱓").Value == "𠜎𠜱𠝹𠱓");
var r = new rr.Regex("𠝹𠱓");
// Disallow beginning searches in the middle of a UTF-16 surrogate pair.
bool exception = false;
try
{
r.Match("𠜎𠜱𠝹𠱓", 1);
}
catch (ArgumentException)
{
exception = true;
}
// Indices and lengths are reported as UTF-16 code units.
Debug.Assert(r.Match("𠜎水𠜱𠝹𠱓", 5).Index == 5);
Debug.Assert(r.Match("𠜎水𠜱𠝹𠱓", 5).Length == 4);
Debug.Assert(exception);
Console.WriteLine("\t... Success.\n");
}
{
Console.WriteLine("Running performance tests ...\n");
var builder = new StringBuilder();
var haybytes = System.IO.File.ReadAllBytes(@"..\..\mtent12.txt");
var watch = new Stopwatch();
watch.Start();
var haystring = Encoding.ASCII.GetString(haybytes);
var encodetime = watch.Elapsed;
watch.Reset();
Console.WriteLine("\tText length: " + haystring.Length);
Console.WriteLine("\tEncoding time: " + encodetime);
Console.WriteLine();
var testcases = new TestCase[16] {
new TestCase("Twain"),
new TestCase("^Twain"),
new TestCase("Twain$"),
new TestCase("Huck[a-zA-Z]+|Finn[a-zA-Z]+"),
new TestCase("a[^x]{20}b"),
new TestCase("Tom|Sawyer|Huckleberry|Finn"),
new TestCase(".{0,3}(Tom|Sawyer|Huckleberry|Finn)"),
new TestCase("[a-zA-Z]+ing"),
new TestCase("^[a-zA-Z]{0,4}ing[^a-zA-Z]"),
new TestCase("[a-zA-Z]+ing$"),
new TestCase("^[a-zA-Z ]{5,}$"),
new TestCase("^.{16,20}$"),
new TestCase("([a-f](.[d-m].){0,2}[h-n]){2}"),
new TestCase("([A-Za-z]awyer|[A-Za-z]inn)[^a-zA-Z]"),
new TestCase(@"""[^""]{0,30}[?!\.]"""),
new TestCase("Tom.{10,25}river|river.{10,25}Tom")
};
Console.Write("\tRunning 'First Match' test...");
foreach (var testcase in testcases)
{
var re2b = new rr.Regex(testcase.Pattern, rr.RegexOptions.Multiline | rr.RegexOptions.Latin1);
var re2s = new rr.Regex(testcase.Pattern, rr.RegexOptions.Multiline);
var nets = new nn.Regex(testcase.Pattern, nn.RegexOptions.Multiline);
watch.Start();
var re2ByteMatch = re2b.Match(haybytes);
testcase.AddRe2ByteResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
watch.Start();
var re2StringMatch = re2s.Match(haystring);
testcase.AddRe2StringResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
watch.Start();
var netMatch = nets.Match(haystring);
testcase.AddNETResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
if (re2ByteMatch.Value != re2StringMatch.Value)
{
Console.WriteLine();
Console.WriteLine("\tMatch.Value: RE2 bytes failed to match RE2 string for pattern " + re2b.Pattern);
Console.WriteLine("\tThis is not necessarily an error and may be due to accent characters.");
Console.WriteLine("\tRE2 bytes value: " + re2ByteMatch.Value);
Console.WriteLine("\tRE2 string value: " + re2StringMatch.Value);
Console.WriteLine();
}
if (re2StringMatch.Value != netMatch.Value)
{
Console.WriteLine();
Console.WriteLine("\tMatch.Value: RE2 string failed to match .NET string for pattern " + re2b.Pattern);
Console.WriteLine("\tThis is not necessarily an error and may be due to accent characters.");
Console.WriteLine("\tRE2 string value: " + re2StringMatch.Value);
Console.WriteLine("\t.NET string value: " + netMatch.Value);
Console.WriteLine();
}
Debug.Assert(re2StringMatch.Index == netMatch.Index);
Debug.Assert(re2StringMatch.Length == netMatch.Length);
}
Console.WriteLine("\n\nResults:\n\n");
PrintByteVsStringResults(testcases);
Console.WriteLine("\n");
PrintStringVsStringResults(testcases);
Console.WriteLine("\n\t... Success.\n");
foreach (var testcase in testcases)
{
testcase.Reset();
}
Console.Write("\n\n\tRunning 'All Matches' test...");
foreach (var testcase in testcases)
{
var re2b = new rr.Regex(testcase.Pattern, rr.RegexOptions.Multiline | rr.RegexOptions.Latin1);
var re2s = new rr.Regex(testcase.Pattern, rr.RegexOptions.Multiline);
var nets = new nn.Regex(testcase.Pattern, nn.RegexOptions.Multiline);
watch.Start();
var re2ByteMatches = re2b.Matches(haybytes);
// Matches() methods are lazily evaluated.
testcase.Re2ByteMatchCount = re2ByteMatches.Count;
testcase.AddRe2ByteResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
watch.Start();
var re2StringMatches = re2s.Matches(haystring);
// Matches() methods are lazily evaluated.
testcase.Re2StringMatchCount = re2StringMatches.Count;
testcase.AddRe2StringResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
watch.Start();
var netMatches = nets.Matches(haystring);
// Matches() methods are lazily evaluated.
testcase.NETMatchCount = netMatches.Count;
testcase.AddNETResult(TimerTicksToMilliseconds(watch.ElapsedTicks));
watch.Reset();
Debug.Assert(re2ByteMatches.Count == re2StringMatches.Count);
Debug.Assert(re2ByteMatches.Count == netMatches.Count);
for (int j = 0; j < re2ByteMatches.Count; j++)
{
if (re2ByteMatches[j].Value != re2StringMatches[j].Value)
{
Console.WriteLine();
Console.WriteLine("\tMatch.Value: RE2 bytes failed to match RE2 string for pattern " + re2b.Pattern);
Console.WriteLine("\tThis is not necessarily an error and may be due to accent characters.");
Console.WriteLine("\tRE2 bytes value: " + re2ByteMatches[j].Value);
Console.WriteLine("\tRE2 string value: " + re2StringMatches[j].Value);
Console.WriteLine();
}
if (re2StringMatches[j].Value != netMatches[j].Value)
{
Console.WriteLine();
Console.WriteLine("\tMatch.Value: RE2 string failed to match .NET string for pattern " + re2b.Pattern);
Console.WriteLine("\tThis is not necessarily an error and may be due to accent characters.");
Console.WriteLine("\tRE2 string value: " + re2StringMatches[j].Value);
Console.WriteLine("\t.NET string value: " + netMatches[j].Value);
Console.WriteLine();
}
}
}
Console.WriteLine("\n\nResults:\n\n");
PrintByteVsStringResults(testcases);
Console.WriteLine("\n");
PrintStringVsStringResults(testcases);
Console.WriteLine("\n\t... Success.\n");
}
}
catch (Exception ex)
{
Console.WriteLine(ex.Message + ex.StackTrace);
}
GC.Collect();
Console.WriteLine();
Console.WriteLine("Done.");
Console.ReadLine();
}