public async Task Count_Timeout_ThrowsAfterTooLongExecution(RegexEngine engine)
{
if (RegexHelpers.IsNonBacktracking(engine))
{
// Test relies on backtracking taking a long time
return;
}
const string Pattern = @"^(\w+\s?)*$";
const string Input = "An input string that takes a very very very very very very very very very very very long time!";
Regex r = await RegexHelpers.GetRegexAsync(engine, Pattern, RegexOptions.None, TimeSpan.FromMilliseconds(1));
Stopwatch sw = Stopwatch.StartNew();
Assert.Throws <RegexMatchTimeoutException>(() => r.Count(Input));
Assert.Throws <RegexMatchTimeoutException>(() => r.Count(Input.AsSpan()));
Assert.InRange(sw.Elapsed.TotalSeconds, 0, 10); // arbitrary upper bound that should be well above what's needed with a 1ms timeout
switch (engine)
{
case RegexEngine.Interpreter:
case RegexEngine.Compiled:
sw = Stopwatch.StartNew();
Assert.Throws <RegexMatchTimeoutException>(() => Regex.Count(Input, Pattern, RegexHelpers.OptionsFromEngine(engine), TimeSpan.FromMilliseconds(1)));
Assert.Throws <RegexMatchTimeoutException>(() => Regex.Count(Input.AsSpan(), Pattern, RegexHelpers.OptionsFromEngine(engine), TimeSpan.FromMilliseconds(1)));
Assert.InRange(sw.Elapsed.TotalSeconds, 0, 10); // arbitrary upper bound that should be well above what's needed with a 1ms timeout
break;
}
}
public async Task CharactersComparedOneByOne_Invariant(RegexEngine engine, RegexOptions options)
{
// Regex compares characters one by one. If that changes, it could impact the behavior of
// a case like this, where these characters are not the same, but the strings compare
// as equal with the invariant culture (and some other cultures as well).
const string S1 = "\u00D6\u200D";
const string S2 = "\u004F\u0308";
// Validate the chosen strings to make sure they compare the way we want to test via Regex
Assert.False(S1[0] == S2[0]);
Assert.False(S1[1] == S2[1]);
Assert.StartsWith(S1, S2, StringComparison.InvariantCulture);
Assert.True(S1.Equals(S2, StringComparison.InvariantCulture));
// Test varying lengths of strings to validate codegen changes that kick in at longer lengths
foreach (int multiple in new[] { 1, 10, 100 })
{
string pattern = string.Concat(Enumerable.Repeat(S1, multiple));
string input = string.Concat(Enumerable.Repeat(S2, multiple));
Regex r;
// Validate when the string is at the beginning of the pattern, as it impacts prefix matching.
r = await RegexHelpers.GetRegexAsync(engine, pattern, options);
Assert.False(r.IsMatch(input));
Assert.True(r.IsMatch(pattern));
// Validate when it's not at the beginning of the pattern, as it impacts "multi" matching.
r = await RegexHelpers.GetRegexAsync(engine, "[abc]" + pattern, options);
Assert.False(r.IsMatch("a" + input));
Assert.True(r.IsMatch("a" + pattern));
}
}
public async Task Replace(RegexEngine engine, string pattern, string input, string replacement, RegexOptions options, int count, int start, string expected)
{
// A few tests exceed the 1000 limit, they reach 6003
RegexHelpers.SetSafeSizeThreshold(6005);
Regex r;
try
{
r = await RegexHelpers.GetRegexAsync(engine, pattern, options);
}
finally
{
RegexHelpers.RestoreSafeSizeThresholdToDefault();
}
bool isDefaultStart = RegexHelpers.IsDefaultStart(input, options, start);
bool isDefaultCount = RegexHelpers.IsDefaultCount(input, options, count);
if (isDefaultStart)
{
if (isDefaultCount)
{
Assert.Equal(expected, r.Replace(input, replacement));
Assert.Equal(expected, Regex.Replace(input, pattern, replacement, options));
}
Assert.Equal(expected, r.Replace(input, replacement, count));
}
Assert.Equal(expected, r.Replace(input, replacement, count, start));
}
public async Task Count_ReturnsExpectedCount(RegexEngine engine, string pattern, string input, RegexOptions options, int expectedCount)
{
Regex r = await RegexHelpers.GetRegexAsync(engine, pattern, options);
Assert.Equal(expectedCount, r.Count(input));
Assert.Equal(expectedCount, r.Count(input.AsSpan()));
Assert.Equal(r.Count(input), r.Matches(input).Count);
Assert.Equal(r.Count(input.AsSpan()), r.Matches(input).Count);
if (options == RegexOptions.None && engine == RegexEngine.Interpreter)
{
Assert.Equal(expectedCount, Regex.Count(input, pattern));
Assert.Equal(expectedCount, Regex.Count(input.AsSpan(), pattern));
}
switch (engine)
{
case RegexEngine.Interpreter:
case RegexEngine.Compiled:
case RegexEngine.NonBacktracking:
RegexOptions engineOptions = RegexHelpers.OptionsFromEngine(engine);
Assert.Equal(expectedCount, Regex.Count(input, pattern, options | engineOptions));
Assert.Equal(expectedCount, Regex.Count(input.AsSpan(), pattern, options | engineOptions));
Assert.Equal(expectedCount, Regex.Count(input, pattern, options | engineOptions, Regex.InfiniteMatchTimeout));
Assert.Equal(expectedCount, Regex.Count(input.AsSpan(), pattern, options | engineOptions, Regex.InfiniteMatchTimeout));
break;
}
}
public void TurkishI_Is_Differently_LowerUpperCased_In_Turkish_Culture_NonBacktracking()
{
var turkish = new CultureInfo("tr-TR");
string input = "I\u0131\u0130i";
// Use the input as the regex also
// Ignore the Compiled option here because it is a noop in combination with NonBacktracking
Regex cultInvariantRegex = RegexHelpers.CreateRegexInCulture(input, RegexHelpers.RegexOptionNonBacktracking | RegexOptions.IgnoreCase | RegexOptions.CultureInvariant, CultureInfo.InvariantCulture);
Regex turkishRegex = RegexHelpers.CreateRegexInCulture(input, RegexHelpers.RegexOptionNonBacktracking | RegexOptions.IgnoreCase, turkish);
Assert.True(cultInvariantRegex.IsMatch(input));
Assert.True(turkishRegex.IsMatch(input)); // <---------- This result differs from the result in the previous test!!!
// As above and no surprises here
// The regexes recognize different lowercase variants of different versions of i differently
Assert.True(cultInvariantRegex.IsMatch(input.ToLowerInvariant()));
Assert.False(cultInvariantRegex.IsMatch(input.ToLower(turkish)));
Assert.False(turkishRegex.IsMatch(input.ToLowerInvariant()));
Assert.True(turkishRegex.IsMatch(input.ToLower(turkish)));
// The same holds symmetrically for ToUpper
Assert.True(cultInvariantRegex.IsMatch(input.ToUpperInvariant()));
Assert.False(cultInvariantRegex.IsMatch(input.ToUpper(turkish)));
Assert.False(turkishRegex.IsMatch(input.ToUpperInvariant()));
Assert.True(turkishRegex.IsMatch(input.ToUpper(turkish)));
}
public async Task SingleExpected(RegexEngine engine, char c)
{
string s = $@"\u{(int)c:X4}";
var set = new HashSet <char>()
{
c
};
// One
await ValidateSetAsync(engine, $"{s}", RegexOptions.None, set, null);
await ValidateSetAsync(engine, $"[{s}]", RegexOptions.None, set, null);
await ValidateSetAsync(engine, $"[^{s}]", RegexOptions.None, null, set);
if (!RegexHelpers.IsNonBacktracking(engine))
{
// Positive lookahead
await ValidateSetAsync(engine, $"(?={s}){s}", RegexOptions.None, set, null);
await ValidateSetAsync(engine, $"(?=[^{s}])[^{s}]", RegexOptions.None, null, set);
// Negative lookahead
await ValidateSetAsync(engine, $"(?![^{s}]){s}", RegexOptions.None, set, null);
await ValidateSetAsync(engine, $"(?![{s}])[^{s}]", RegexOptions.None, null, set);
}
// Concatenation
await ValidateSetAsync(engine, $"[{s}{s}]", RegexOptions.None, set, null);
await ValidateSetAsync(engine, $"[^{s}{s}{s}]", RegexOptions.None, null, set);
// Alternation
await ValidateSetAsync(engine, $"{s}|{s}", RegexOptions.None, set, null);
await ValidateSetAsync(engine, $"[^{s}]|[^{s}]|[^{s}]", RegexOptions.None, null, set);
await ValidateSetAsync(engine, $"{s}|[^{s}]", RegexOptions.None, null, new HashSet <char>());
}
public async Task Matches(RegexEngine engine, string pattern, string input, RegexOptions options, CaptureData[] expected)
{
Regex regexAdvanced = await RegexHelpers.GetRegexAsync(engine, pattern, options);
VerifyMatches(regexAdvanced.Matches(input), expected);
VerifyMatches(regexAdvanced.Match(input), expected);
}
public void Match_In_Different_Cultures_CriticalCases(string pattern, RegexOptions options, CultureInfo culture, string input, string match_expected)
{
Regex r = RegexHelpers.CreateRegexInCulture(pattern, options, culture);
Match match = r.Match(input);
Assert.Equal(match_expected, match.Value);
}
public void Replace(string pattern, string input, MatchEvaluator evaluator, RegexOptions options, int count, int start, string expected)
{
bool isDefaultStart = RegexHelpers.IsDefaultStart(input, options, start);
bool isDefaultCount = RegexHelpers.IsDefaultCount(input, options, count);
if (options == RegexOptions.None)
{
if (isDefaultStart && isDefaultCount)
{
// Use Replace(string, MatchEvaluator) or Replace(string, string, MatchEvaluator)
Assert.Equal(expected, new Regex(pattern).Replace(input, evaluator));
Assert.Equal(expected, Regex.Replace(input, pattern, evaluator));
}
if (isDefaultStart)
{
// Use Replace(string, MatchEvaluator, string, int)
Assert.Equal(expected, new Regex(pattern).Replace(input, evaluator, count));
}
// Use Replace(string, MatchEvaluator, int, int)
Assert.Equal(expected, new Regex(pattern).Replace(input, evaluator, count, start));
}
if (isDefaultStart && isDefaultCount)
{
// Use Replace(string, MatchEvaluator) or Replace(string, MatchEvaluator, RegexOptions)
Assert.Equal(expected, new Regex(pattern, options).Replace(input, evaluator));
Assert.Equal(expected, Regex.Replace(input, pattern, evaluator, options));
}
if (isDefaultStart)
{
// Use Replace(string, MatchEvaluator, string, int)
Assert.Equal(expected, new Regex(pattern, options).Replace(input, evaluator, count));
}
// Use Replace(string, MatchEvaluator, int, int)
Assert.Equal(expected, new Regex(pattern, options).Replace(input, evaluator, count, start));
}
public void Split(string pattern, string input, RegexOptions options, int count, int start, string[] expected)
{
bool isDefaultStart = RegexHelpers.IsDefaultStart(input, options, start);
bool isDefaultCount = RegexHelpers.IsDefaultStart(input, options, count);
if (options == RegexOptions.None)
{
// Use Split(string), Split(string, string), Split(string, int) or Split(string, int, int)
if (isDefaultStart && isDefaultCount)
{
// Use Split(string) or Split(string, string)
Assert.Equal(expected, new Regex(pattern).Split(input));
Assert.Equal(expected, Regex.Split(input, pattern));
}
if (isDefaultStart)
{
// Use Split(string, int)
Assert.Equal(expected, new Regex(pattern).Split(input, count));
}
// Use Split(string, int, int)
Assert.Equal(expected, new Regex(pattern).Split(input, count, start));
}
if (isDefaultStart && isDefaultCount)
{
// Use Split(string, string, RegexOptions)
Assert.Equal(expected, Regex.Split(input, pattern, options));
}
if (isDefaultStart)
{
// Use Split(string, int)
Assert.Equal(expected, new Regex(pattern, options).Split(input, count));
}
// Use Split(string, int, int, int)
Assert.Equal(expected, new Regex(pattern, options).Split(input, count, start));
}
public async Task Match_In_Different_Cultures_CriticalCases(string pattern, RegexOptions options, RegexEngine engine, CultureInfo culture, string input, string match_expected)
{
Regex r = await RegexHelpers.GetRegexAsync(engine, pattern, options, culture);
Match match = r.Match(input);
Assert.Equal(match_expected, match.Value);
}
public async Task Test(RegexEngine engine, RegexOptions options, string pattern, string input, string captures, string nonBacktrackingCaptures = null)
{
if (input == "NULL")
{
input = "";
}
bool nonBacktracking = engine == RegexEngine.NonBacktracking;
string expected = nonBacktracking && nonBacktrackingCaptures != null ?
nonBacktrackingCaptures : // nonBacktrackingCaptures value overrides the expected result in NonBacktracking mode
captures;
if (expected == "BADBR")
{
await Assert.ThrowsAnyAsync <ArgumentException>(async() => await RegexHelpers.GetRegexAsync(engine, pattern, options));
return;
}
if (nonBacktracking && nonBacktrackingCaptures == "NONBACKTRACKINGINCOMPATIBLE")
{
// In particular: backreferences are not supported in NonBacktracking mode
await Assert.ThrowsAnyAsync <NotSupportedException>(() => RegexHelpers.GetRegexAsync(engine, pattern, options));
return;
}
Regex r = await RegexHelpers.GetRegexAsync(engine, pattern, options);
if (expected == "NOMATCH")
{
Assert.False(r.IsMatch(input));
return;
}
Match match = r.Match(input);
Assert.True(match.Success);
var expectedSet = new HashSet <(int start, int end)>(
expected
.Split(new[] { '(', ')' }, StringSplitOptions.RemoveEmptyEntries)
.Select(s => s.Split(','))
.Select(s => (start: int.Parse(s[0]), end: int.Parse(s[1]))));
var actualSet = new HashSet <(int start, int end)>(
match.Groups
.Cast <Group>()
.Select(g => (start: g.Index, end: g.Index + g.Length)));
// NonBacktracking mode only provides the top-level match.
// The .NET implementation sometimes has extra captures beyond what the original data specifies, so we assert a subset.
if (nonBacktracking ? !actualSet.IsSubsetOf(expectedSet) : !expectedSet.IsSubsetOf(actualSet))
{
throw new Xunit.Sdk.XunitException($"Actual: {string.Join(", ", actualSet)}{Environment.NewLine}Expected: {string.Join(", ", expected)}");
}
}
public async Task TurkishCulture_MatchesWordChar(RegexEngine engine, string input, RegexOptions options, string expectedResult)
{
using (new ThreadCultureChange(new CultureInfo("tr-TR")))
{
Regex regex = await RegexHelpers.GetRegexAsync(engine, @"\w*", options);
Assert.Equal(expectedResult, regex.Match(input).Value);
}
}
public static IEnumerable <object[]> Count_ReturnsExpectedCount_TestData()
{
foreach (RegexEngine engine in RegexHelpers.AvailableEngines)
{
yield return(new object[] { engine, @"", "", 0, RegexOptions.None, 1 });
yield return(new object[] { engine, @"", "a", 0, RegexOptions.None, 2 });
yield return(new object[] { engine, @"", "ab", 0, RegexOptions.None, 3 });
yield return(new object[] { engine, @"", "ab", 1, RegexOptions.None, 2 });
yield return(new object[] { engine, @"\w", "", 0, RegexOptions.None, 0 });
yield return(new object[] { engine, @"\w", "a", 0, RegexOptions.None, 1 });
yield return(new object[] { engine, @"\w", "ab", 0, RegexOptions.None, 2 });
yield return(new object[] { engine, @"\w", "ab", 1, RegexOptions.None, 1 });
yield return(new object[] { engine, @"\w", "ab", 2, RegexOptions.None, 0 });
yield return(new object[] { engine, @"\b\w+\b", "abc def ghi jkl", 0, RegexOptions.None, 4 });
yield return(new object[] { engine, @"\b\w+\b", "abc def ghi jkl", 7, RegexOptions.None, 2 });
yield return(new object[] { engine, @"A", "", 0, RegexOptions.IgnoreCase, 0 });
yield return(new object[] { engine, @"A", "a", 0, RegexOptions.IgnoreCase, 1 });
yield return(new object[] { engine, @"A", "aAaA", 0, RegexOptions.IgnoreCase, 4 });
yield return(new object[] { engine, @".", "\n\n\n", 0, RegexOptions.None, 0 });
yield return(new object[] { engine, @".", "\n\n\n", 0, RegexOptions.Singleline, 3 });
yield return(new object[] { engine, @"[а-я-[аeиоуыэюя]]", "спокойной ночи", 0, RegexOptions.None, 8 });
if (!RegexHelpers.IsNonBacktracking(engine))
{
// Lookbehinds
yield return(new object[] { engine, @"(?<=abc)\w", "abcxabcy", 7, RegexOptions.None, 1 });
// Starting anchors
yield return(new object[] { engine, @"\Gdef", "abcdef", 0, RegexOptions.None, 0 });
yield return(new object[] { engine, @"\Gdef", "abcdef", 3, RegexOptions.None, 1 });
// RightToLeft
yield return(new object[] { engine, @"\b\w+\b", "abc def ghi jkl", 15, RegexOptions.RightToLeft, 4 });
yield return(new object[] { RegexEngine.Interpreter, @"(?<=abc)\w", "abcxabcy", 8, RegexOptions.RightToLeft, 2 });
yield return(new object[] { engine, @"(?<=abc)\w", "abcxabcy", 7, RegexOptions.RightToLeft, 1 });
}
}
}
public async Task CharactersLowercasedOneByOne(RegexEngine engine)
{
using (new ThreadCultureChange("en-US"))
{
Assert.True((await RegexHelpers.GetRegexAsync(engine, "\uD801\uDC00", RegexOptions.IgnoreCase)).IsMatch("\uD801\uDC00"));
Assert.True((await RegexHelpers.GetRegexAsync(engine, "\uD801\uDC00", RegexOptions.IgnoreCase)).IsMatch("abcdefg\uD801\uDC00"));
Assert.True((await RegexHelpers.GetRegexAsync(engine, "\uD801", RegexOptions.IgnoreCase)).IsMatch("\uD801\uDC00"));
Assert.True((await RegexHelpers.GetRegexAsync(engine, "\uDC00", RegexOptions.IgnoreCase)).IsMatch("\uD801\uDC00"));
}
}
public void EnumerateMatches_Count(RegexEngine engine, string pattern, string input, int expectedCount)
{
Regex r = RegexHelpers.GetRegexAsync(engine, pattern).GetAwaiter().GetResult();
int count = 0;
foreach (ValueMatch _ in r.EnumerateMatches(input))
{
count++;
}
Assert.Equal(expectedCount, count);
}
public void TurkishCulture_Handling_Of_IgnoreCase_NonBacktracking()
{
var turkish = new CultureInfo("tr-TR");
string input = "I\u0131\u0130i";
string pattern = "[H-J][\u0131-\u0140][\u0120-\u0130][h-j]";
Regex regex = RegexHelpers.CreateRegexInCulture(pattern, RegexOptions.IgnoreCase | RegexHelpers.RegexOptionNonBacktracking, turkish);
// The pattern must trivially match the input because all of the letters fall in the given intervals
// Ignoring case can only add more letters here -- not REMOVE letters
Assert.True(regex.IsMatch(input));
}
public async Task TurkishCulture_Handling_Of_IgnoreCase(RegexEngine engine)
{
var turkish = new CultureInfo("tr-TR");
string input = "I\u0131\u0130i";
string pattern = "[H-J][\u0131-\u0140][\u0120-\u0130][h-j]";
Regex regex = await RegexHelpers.GetRegexAsync(engine, pattern, RegexOptions.IgnoreCase, turkish);
// The pattern must trivially match the input because all of the letters fall in the given intervals
// Ignoring case can only add more letters here -- not REMOVE letters
Assert.True(regex.IsMatch(input));
}
public async Task CharactersComparedOneByOne_AnchoredPattern(RegexEngine engine, string pattern, string input, string culture, RegexOptions options, bool expected)
{
// Regex compares characters one by one. If that changes, it could impact the behavior of
// a case like this, where these characters are not the same, but the strings compare
// as equal with the invariant culture (and some other cultures as well).
using (new ThreadCultureChange(culture))
{
Regex r = await RegexHelpers.GetRegexAsync(engine, pattern, options);
Assert.Equal(expected, r.IsMatch(input));
}
}
public void EnumerateMatches_ReturnsExpectedCount(RegexEngine engine, string pattern, string input, int startat, RegexOptions options, int expectedCount)
{
Regex r = RegexHelpers.GetRegexAsync(engine, pattern, options).GetAwaiter().GetResult();
int count;
count = 0;
foreach (ValueMatch _ in r.EnumerateMatches(input, startat))
{
count++;
}
Assert.Equal(expectedCount, count);
bool isDefaultStartAt = startat == ((options & RegexOptions.RightToLeft) != 0 ? input.Length : 0);
if (!isDefaultStartAt)
{
return;
}
if (options == RegexOptions.None && engine == RegexEngine.Interpreter)
{
count = 0;
foreach (ValueMatch _ in Regex.EnumerateMatches(input, pattern))
{
count++;
}
Assert.Equal(expectedCount, count);
}
switch (engine)
{
case RegexEngine.Interpreter:
case RegexEngine.Compiled:
case RegexEngine.NonBacktracking:
RegexOptions engineOptions = RegexHelpers.OptionsFromEngine(engine);
count = 0;
foreach (ValueMatch _ in Regex.EnumerateMatches(input, pattern, options | engineOptions))
{
count++;
}
Assert.Equal(expectedCount, count);
count = 0;
foreach (ValueMatch _ in Regex.EnumerateMatches(input, pattern, options | engineOptions, Regex.InfiniteMatchTimeout))
{
count++;
}
Assert.Equal(expectedCount, count);
break;
}
}
public async Task Test(string pattern, string input, string captures)
{
if (input == "NULL")
{
input = "";
}
foreach (RegexEngine engine in RegexHelpers.AvailableEngines)
{
if (captures == "BADBR")
{
await Assert.ThrowsAnyAsync <ArgumentException>(async() => (await RegexHelpers.GetRegexAsync(engine, pattern)).IsMatch(input));
return;
}
Regex r = await RegexHelpers.GetRegexAsync(engine, pattern);
if (captures == "NOMATCH")
{
Assert.False(r.IsMatch(input));
return;
}
Match match = r.Match(input);
Assert.True(match.Success);
var expected = new HashSet <(int start, int end)>(
captures
.Split(new[] { '(', ')' }, StringSplitOptions.RemoveEmptyEntries)
.Select(s => s.Split(','))
.Select(s => (start: int.Parse(s[0]), end: int.Parse(s[1])))
.Distinct()
.OrderBy(c => c.start)
.ThenBy(c => c.end));
var actual = new HashSet <(int start, int end)>(
match.Groups
.Cast <Group>()
.Select(g => (start: g.Index, end: g.Index + g.Length))
.Distinct()
.OrderBy(g => g.start)
.ThenBy(g => g.end));
// The .NET implementation sometimes has extra captures beyond what the original data specifies, so we assert a subset.
if (!expected.IsSubsetOf(actual))
{
throw new Xunit.Sdk.XunitException($"Actual: {string.Join(", ", actual)}{Environment.NewLine}Expected: {string.Join(", ", expected)}");
}
}
}
public static IEnumerable <object[]> TurkishCulture_MatchesWordChar_MemberData()
{
foreach (RegexEngine engine in RegexHelpers.AvailableEngines)
{
yield return(new object[] { engine, "I\u0131\u0130i", RegexOptions.None, "I\u0131\u0130i" });
yield return(new object[] { engine, "I\u0131\u0130i", RegexOptions.IgnoreCase, "I\u0131\u0130i" });
if (!RegexHelpers.IsNonBacktracking(engine))
{
yield return(new object[] { engine, "I\u0131\u0130i", RegexOptions.IgnoreCase | RegexOptions.ECMAScript, "" });
}
}
}
public static IEnumerable <object[]> Replace_MatchEvaluator_TestData()
{
foreach (RegexEngine engine in RegexHelpers.AvailableEngines)
{
yield return(new object[] { engine, "a", "bbbb", new MatchEvaluator(match => "uhoh"), RegexOptions.None, 4, 0, "bbbb" });
yield return(new object[] { engine, "(Big|Small)", "Big mountain", new MatchEvaluator(MatchEvaluator1), RegexOptions.None, 12, 0, "Huge mountain" });
yield return(new object[] { engine, "(Big|Small)", "Small village", new MatchEvaluator(MatchEvaluator1), RegexOptions.None, 13, 0, "Tiny village" });
if ("i".ToUpper() == "I")
{
yield return(new object[] { engine, "(Big|Small)", "bIG horse", new MatchEvaluator(MatchEvaluator1), RegexOptions.IgnoreCase, 9, 0, "Huge horse" });
}
yield return(new object[] { engine, "(Big|Small)", "sMaLl dog", new MatchEvaluator(MatchEvaluator1), RegexOptions.IgnoreCase, 9, 0, "Tiny dog" });
yield return(new object[] { engine, ".+", "XSP_TEST_FAILURE", new MatchEvaluator(MatchEvaluator2), RegexOptions.None, 16, 0, "SUCCESS" });
yield return(new object[] { engine, "[abcabc]", "abcabc", new MatchEvaluator(MatchEvaluator3), RegexOptions.None, 6, 0, "ABCABC" });
yield return(new object[] { engine, "[abcabc]", "abcabc", new MatchEvaluator(MatchEvaluator3), RegexOptions.None, 3, 0, "ABCabc" });
yield return(new object[] { engine, "[abcabc]", "abcabc", new MatchEvaluator(MatchEvaluator3), RegexOptions.None, 3, 2, "abCABc" });
if (!RegexHelpers.IsNonBacktracking(engine))
{
// Regression test:
// Regex treating Devanagari matra characters as matching "\b"
// Unicode characters in the "Mark, NonSpacing" Category, U+0902=Devanagari sign anusvara, U+0947=Devanagri vowel sign E
string boldInput = "\u092f\u0939 \u0915\u0930 \u0935\u0939 \u0915\u0930\u0947\u0902 \u0939\u0948\u0964";
string boldExpected = "\u092f\u0939 <b>\u0915\u0930</b> \u0935\u0939 <b>\u0915\u0930\u0947\u0902</b> \u0939\u0948\u0964";
yield return(new object[] { engine, @"\u0915\u0930.*?\b", boldInput, new MatchEvaluator(MatchEvaluatorBold), RegexOptions.CultureInvariant | RegexOptions.Singleline, boldInput.Length, 0, boldExpected });
// RighToLeft
yield return(new object[] { engine, "a", "bbbb", new MatchEvaluator(match => "uhoh"), RegexOptions.RightToLeft, 4, 3, "bbbb" });
yield return(new object[] { engine, @"foo\s+", "0123456789foo4567890foo ", new MatchEvaluator(MatchEvaluatorBar), RegexOptions.RightToLeft, 32, 32, "0123456789foo4567890bar" });
yield return(new object[] { engine, @"\d", "0123456789foo4567890foo ", new MatchEvaluator(MatchEvaluatorPoundSign), RegexOptions.RightToLeft, 17, 32, "##########foo#######foo " });
yield return(new object[] { engine, @"\d", "0123456789foo4567890foo ", new MatchEvaluator(MatchEvaluatorPoundSign), RegexOptions.RightToLeft, 7, 32, "0123456789foo#######foo " });
yield return(new object[] { engine, @"\d", "0123456789foo4567890foo ", new MatchEvaluator(MatchEvaluatorPoundSign), RegexOptions.RightToLeft, 0, 32, "0123456789foo4567890foo " });
yield return(new object[] { engine, @"\d", "0123456789foo4567890foo ", new MatchEvaluator(MatchEvaluatorPoundSign), RegexOptions.RightToLeft, -1, 32, "##########foo#######foo " });
}
}
}
public void EnumerateMatches(RegexEngine engine, string pattern, string input, RegexOptions options, CaptureData[] expected)
{
Regex regexAdvanced = RegexHelpers.GetRegexAsync(engine, pattern, options).GetAwaiter().GetResult();
int count = 0;
ReadOnlySpan <char> span = input.AsSpan();
foreach (ValueMatch match in regexAdvanced.EnumerateMatches(span))
{
Assert.Equal(expected[count].Index, match.Index);
Assert.Equal(expected[count].Length, match.Length);
Assert.Equal(expected[count].Value, span.Slice(match.Index, match.Length).ToString());
count++;
}
Assert.Equal(expected.Length, count);
}
public void Match(string pattern, string input, RegexOptions options, int beginning, int length, CaptureData[] expected)
{
bool isDefaultStart = RegexHelpers.IsDefaultStart(input, options, beginning);
bool isDefaultCount = RegexHelpers.IsDefaultStart(input, options, length);
if (options == RegexOptions.None)
{
if (isDefaultStart && isDefaultCount)
{
// Use Match(string) or Match(string, string)
VerifyMatch(new Regex(pattern).Match(input), true, expected);
VerifyMatch(Regex.Match(input, pattern), true, expected);
Assert.True(new Regex(pattern).IsMatch(input));
Assert.True(Regex.IsMatch(input, pattern));
}
if (beginning + length == input.Length)
{
// Use Match(string, int)
VerifyMatch(new Regex(pattern).Match(input, beginning), true, expected);
Assert.True(new Regex(pattern).IsMatch(input, beginning));
}
else
{
// Use Match(string, int, int)
VerifyMatch(new Regex(pattern).Match(input, beginning, length), true, expected);
}
}
if (isDefaultStart && isDefaultCount)
{
// Use Match(string) or Match(string, string, RegexOptions)
VerifyMatch(new Regex(pattern, options).Match(input), true, expected);
VerifyMatch(Regex.Match(input, pattern, options), true, expected);
Assert.True(Regex.IsMatch(input, pattern, options));
}
if (beginning + length == input.Length)
{
// Use Match(string, int)
VerifyMatch(new Regex(pattern, options).Match(input, beginning), true, expected);
}
if ((options & RegexOptions.RightToLeft) == 0)
{
// Use Match(string, int, int)
VerifyMatch(new Regex(pattern, options).Match(input, beginning, length), true, expected);
}
}
public async Task UnicodeCategoriesInclusionsExpected(RegexEngine engine, string generalCategory, UnicodeCategory unicodeCategory)
{
Regex r;
char[] allChars = Enumerable.Range(0, char.MaxValue + 1).Select(i => (char)i).ToArray();
int expectedInCategory = allChars.Count(c => char.GetUnicodeCategory(c) == unicodeCategory);
int expectedNotInCategory = allChars.Length - expectedInCategory;
r = await RegexHelpers.GetRegexAsync(engine, @$ "\p{{{generalCategory}}}");
Assert.Equal(expectedInCategory, r.Matches(string.Concat(allChars)).Count);
r = await RegexHelpers.GetRegexAsync(engine, (@$ "\P{{{generalCategory}}}"));
Assert.Equal(expectedNotInCategory, r.Matches(string.Concat(allChars)).Count);
}
private static async Task ValidateSetAsync(string regex, RegexOptions options, HashSet <char> included, HashSet <char> excluded, bool validateEveryChar = false)
{
Assert.True((included != null) ^ (excluded != null));
foreach (RegexEngine engine in RegexHelpers.AvailableEngines)
{
Regex r = await RegexHelpers.GetRegexAsync(engine, regex, options);
if (validateEveryChar)
{
for (int i = 0; i <= char.MaxValue; i++)
{
bool actual = r.IsMatch(((char)i).ToString());
bool expected = included != null?included.Contains((char)i) : !excluded.Contains((char)i);
if (actual != expected)
{
Fail(i);
}
}
}
else if (included != null)
{
foreach (char c in included)
{
if (!r.IsMatch(c.ToString()))
{
Fail(c);
}
}
}
else
{
foreach (char c in excluded)
{
if (r.IsMatch(c.ToString()))
{
Fail(c);
}
}
}
}
void Fail(int c) => throw new XunitException($"Set=\"{regex}\", Options=\"{options}\", {c:X4} => '{(char)c}'");
}
public void EnumerateMatches_CheckIndex(RegexEngine engine)
{
const string Pattern = @"e{2}\w\b";
const string Input = "needing a reed";
Regex r = RegexHelpers.GetRegexAsync(engine, Pattern).GetAwaiter().GetResult();
int count = 0;
string[] expectedMatches = new[] { "eed" };
int[] expectedIndex = new[] { 11 };
ReadOnlySpan <char> span = Input.AsSpan();
foreach (ValueMatch match in r.EnumerateMatches(span))
{
Assert.Equal(expectedMatches[count], span.Slice(match.Index, match.Length).ToString());
Assert.Equal(expectedIndex[count++], match.Index);
}
}
public async Task Replace_MatchEvaluator_Test(RegexEngine engine, string pattern, string input, MatchEvaluator evaluator, RegexOptions options, int count, int start, string expected)
{
bool isDefaultStart = RegexHelpers.IsDefaultStart(input, options, start);
bool isDefaultCount = RegexHelpers.IsDefaultCount(input, options, count);
Regex r = await RegexHelpers.GetRegexAsync(engine, pattern, options);
if (isDefaultStart && isDefaultCount)
{
Assert.Equal(expected, r.Replace(input, evaluator));
}
if (isDefaultStart)
{
Assert.Equal(expected, r.Replace(input, evaluator, count));
}
Assert.Equal(expected, r.Replace(input, evaluator, count, start));
}
public async Task Split(RegexEngine engine, string pattern, string input, RegexOptions options, int count, int start, string[] expected)
{
bool isDefaultStart = RegexHelpers.IsDefaultStart(input, options, start);
bool isDefaultCount = RegexHelpers.IsDefaultCount(input, options, count);
Regex r = await RegexHelpers.GetRegexAsync(engine, pattern, options);
if (isDefaultStart && isDefaultCount)
{
Assert.Equal(expected, r.Split(input));
}
if (isDefaultStart)
{
Assert.Equal(expected, r.Split(input, count));
}
Assert.Equal(expected, r.Split(input, count, start));
}
