This is an automated email from the ASF dual-hosted git repository.
nightowl888 pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/lucenenet.git
The following commit(s) were added to refs/heads/master by this push:
new 86052e5c4 SWEEP: Prefer 'AsSpan' over 'Substring' when span-based
overloads are available. Fixes #675.
86052e5c4 is described below
commit 86052e5c4701b70cb15f9dbbbd48dad8f7f0b512
Author: Shad Storhaug <[email protected]>
AuthorDate: Mon Oct 17 02:01:38 2022 +0700
SWEEP: Prefer 'AsSpan' over 'Substring' when span-based overloads are
available. Fixes #675.
---
Directory.Build.targets | 7 +++++--
.../Analysis/Hunspell/Dictionary.cs | 4 ++++
.../Analysis/Pattern/PatternReplaceCharFilter.cs | 16 ++++++++++++----
.../Language/DoubleMetaphone.cs | 8 ++++++++
.../ByTask/Feeds/EnwikiContentSource.cs | 15 +++++++++++++--
.../ByTask/Tasks/SearchTravRetHighlightTask.cs | 4 ++++
.../Flexible/Standard/Parser/EscapeQuerySyntaxImpl.cs | 11 +++++++++++
.../Flexible/Standard/Parser/StandardSyntaxParser.cs | 12 ++++++++++--
src/Lucene.Net.Replicator/Http/HttpClientBase.cs | 2 +-
.../Suggest/Analyzing/AnalyzingInfixSuggester.cs | 8 ++++++++
src/Lucene.Net/Util/AttributeSource.cs | 13 +++++++++++++
11 files changed, 89 insertions(+), 11 deletions(-)
diff --git a/Directory.Build.targets b/Directory.Build.targets
index 575b43be5..aa039fc02 100644
--- a/Directory.Build.targets
+++ b/Directory.Build.targets
@@ -34,6 +34,7 @@
<DefineConstants>$(DefineConstants);FEATURE_ARGITERATOR</DefineConstants>
<DefineConstants>$(DefineConstants);FEATURE_DICTIONARY_REMOVE_CONTINUEENUMERATION</DefineConstants>
+
<DefineConstants>$(DefineConstants);FEATURE_STRING_CONCAT_READONLYSPAN</DefineConstants>
</PropertyGroup>
@@ -53,8 +54,10 @@
<DefineConstants>$(DefineConstants);FEATURE_CONDITIONALWEAKTABLE_ENUMERATOR</DefineConstants>
<DefineConstants>$(DefineConstants);FEATURE_CONDITIONALWEAKTABLE_ADDORUPDATE</DefineConstants>
- <DefineConstants>$(DefineConstants);FEATURE_SPAN</DefineConstants>
-
+
<DefineConstants>$(DefineConstants);FEATURE_NUMBER_PARSE_READONLYSPAN</DefineConstants>
+
<DefineConstants>$(DefineConstants);FEATURE_STREAM_READ_SPAN</DefineConstants>
+
<DefineConstants>$(DefineConstants);FEATURE_STRINGBUILDER_APPEND_READONLYSPAN</DefineConstants>
+
</PropertyGroup>
<!-- Features in .NET Standard 2.x, .NET Core 2.x, .NET Core 3.x, .NET 5.x,
and .NET 6.x -->
diff --git a/src/Lucene.Net.Analysis.Common/Analysis/Hunspell/Dictionary.cs
b/src/Lucene.Net.Analysis.Common/Analysis/Hunspell/Dictionary.cs
index 57027293c..5008384c0 100644
--- a/src/Lucene.Net.Analysis.Common/Analysis/Hunspell/Dictionary.cs
+++ b/src/Lucene.Net.Analysis.Common/Analysis/Hunspell/Dictionary.cs
@@ -967,7 +967,11 @@ namespace Lucene.Net.Analysis.Hunspell
sb.Length = 0;
sb.Append(cleansed);
}
+#if FEATURE_STRINGBUILDER_APPEND_READONLYSPAN
+ sb.Append(line.AsSpan(flagSep));
+#else
sb.Append(line.Substring(flagSep));
+#endif
writer.Write(sb.ToString().GetBytes(Encoding.UTF8));
}
}
diff --git
a/src/Lucene.Net.Analysis.Common/Analysis/Pattern/PatternReplaceCharFilter.cs
b/src/Lucene.Net.Analysis.Common/Analysis/Pattern/PatternReplaceCharFilter.cs
index 75c6c3b8d..d42d9f3ea 100644
---
a/src/Lucene.Net.Analysis.Common/Analysis/Pattern/PatternReplaceCharFilter.cs
+++
b/src/Lucene.Net.Analysis.Common/Analysis/Pattern/PatternReplaceCharFilter.cs
@@ -87,7 +87,7 @@ namespace Lucene.Net.Analysis.Pattern
{
buffered.Append(temp, 0, cnt);
}
- transformedInput = new
StringReader(ProcessPattern(buffered).ToString());
+ transformedInput = new StringReader(ProcessPattern(buffered)); //
LUCENENET specific: ProcessPattern already returns string, no need to call
ToString() on it
}
public override int Read()
@@ -110,7 +110,7 @@ namespace Lucene.Net.Analysis.Pattern
/// </summary>
private string ProcessPattern(StringBuilder input)
{
- // LUCENENET TODO: Replacing characters in a StringBuilder is not
natively
+ // LUCENENET TODO: Replacing characters in a StringBuilder via
regex is not natively
// supported in .NET, so this is the approach we are left with.
// At some point it might make sense to try to port the
// MONO implementation over that DOES support Regex on a
StringBuilder.
@@ -131,7 +131,11 @@ namespace Lucene.Net.Analysis.Pattern
int lengthBeforeReplacement = cumulativeOutput.Length +
skippedSize;
// Add the part that didn't match the regex
+#if FEATURE_STRINGBUILDER_APPEND_READONLYSPAN
+ cumulativeOutput.Append(inputStr.AsSpan(lastMatchEnd,
m.Index - lastMatchEnd));
+#else
cumulativeOutput.Append(inputStr.Substring(lastMatchEnd,
m.Index - lastMatchEnd));
+#endif
int groupSize = m.Length;
lastMatchEnd = m.Index + m.Length;
@@ -171,12 +175,16 @@ namespace Lucene.Net.Analysis.Pattern
} while ((m = m.NextMatch()).Success);
// Append the remaining output, no further changes to indices.
- cumulativeOutput.Append(input.ToString(lastMatchEnd,
input.Length - lastMatchEnd));
+#if FEATURE_STRINGBUILDER_APPEND_READONLYSPAN
+ cumulativeOutput.Append(inputStr.AsSpan(lastMatchEnd,
input.Length - lastMatchEnd));
+#else
+ cumulativeOutput.Append(inputStr.Substring(lastMatchEnd,
input.Length - lastMatchEnd));
+#endif
return cumulativeOutput.ToString();
}
// No match - just return the original string.
- return input.ToString();
+ return inputStr; // LUCENENET: Since we have already dumped the
string from input, just return it directly.
}
}
}
\ No newline at end of file
diff --git a/src/Lucene.Net.Analysis.Phonetic/Language/DoubleMetaphone.cs
b/src/Lucene.Net.Analysis.Phonetic/Language/DoubleMetaphone.cs
index 0c2aba590..192ae286e 100644
--- a/src/Lucene.Net.Analysis.Phonetic/Language/DoubleMetaphone.cs
+++ b/src/Lucene.Net.Analysis.Phonetic/Language/DoubleMetaphone.cs
@@ -1240,7 +1240,11 @@ namespace Lucene.Net.Analysis.Phonetic.Language
}
else
{
+#if FEATURE_STRINGBUILDER_APPEND_READONLYSPAN
+ this.primary.Append(value.AsSpan(0, addChars - 0));
+#else
this.primary.Append(value.Substring(0, addChars - 0));
+#endif
}
}
@@ -1253,7 +1257,11 @@ namespace Lucene.Net.Analysis.Phonetic.Language
}
else
{
+#if FEATURE_STRINGBUILDER_APPEND_READONLYSPAN
+ this.alternate.Append(value.AsSpan(0, addChars - 0));
+#else
this.alternate.Append(value.Substring(0, addChars - 0));
+#endif
}
}
diff --git a/src/Lucene.Net.Benchmark/ByTask/Feeds/EnwikiContentSource.cs
b/src/Lucene.Net.Benchmark/ByTask/Feeds/EnwikiContentSource.cs
index 15c5b11f3..8d81b7c07 100644
--- a/src/Lucene.Net.Benchmark/ByTask/Feeds/EnwikiContentSource.cs
+++ b/src/Lucene.Net.Benchmark/ByTask/Feeds/EnwikiContentSource.cs
@@ -116,16 +116,27 @@ namespace Lucene.Net.Benchmarks.ByTask.Feeds
internal string Time(string original)
{
- StringBuilder buffer = new StringBuilder();
+ StringBuilder buffer = new StringBuilder(24); // LUCENENET: We
always have 24 chars, so allocate it up front
+#if FEATURE_STRINGBUILDER_APPEND_READONLYSPAN
+ buffer.Append(original.AsSpan(8, 10 - 8));
+ buffer.Append('-');
+ buffer.Append(months[int.Parse(original.AsSpan(5, 7 - 5),
NumberStyles.Integer, CultureInfo.InvariantCulture) - 1]);
+ buffer.Append('-');
+ buffer.Append(original.AsSpan(0, 4 - 0));
+ buffer.Append(' ');
+ buffer.Append(original.AsSpan(11, 19 - 11));
+ buffer.Append(".000");
+#else
buffer.Append(original.Substring(8, 10 - 8));
buffer.Append('-');
- buffer.Append(months[Convert.ToInt32(original.Substring(5, 7 -
5), CultureInfo.InvariantCulture) - 1]);
+ buffer.Append(months[int.Parse(original.Substring(5, 7 - 5),
NumberStyles.Integer, CultureInfo.InvariantCulture) - 1]);
buffer.Append('-');
buffer.Append(original.Substring(0, 4 - 0));
buffer.Append(' ');
buffer.Append(original.Substring(11, 19 - 11));
buffer.Append(".000");
+#endif
return buffer.ToString();
}
diff --git
a/src/Lucene.Net.Benchmark/ByTask/Tasks/SearchTravRetHighlightTask.cs
b/src/Lucene.Net.Benchmark/ByTask/Tasks/SearchTravRetHighlightTask.cs
index 1f54d9bcd..b207b930c 100644
--- a/src/Lucene.Net.Benchmark/ByTask/Tasks/SearchTravRetHighlightTask.cs
+++ b/src/Lucene.Net.Benchmark/ByTask/Tasks/SearchTravRetHighlightTask.cs
@@ -157,7 +157,11 @@ namespace Lucene.Net.Benchmarks.ByTask.Tasks
else if (splits[i].StartsWith("mergeContiguous[",
StringComparison.Ordinal) == true)
{
int len = "mergeContiguous[".Length;
+#if FEATURE_NUMBER_PARSE_READONLYSPAN
+ m_mergeContiguous = bool.Parse(splits[i].AsSpan(len,
(splits[i].Length - 1) - len));
+#else
m_mergeContiguous = bool.Parse(splits[i].Substring(len,
(splits[i].Length - 1) - len));
+#endif
}
else if (splits[i].StartsWith("fields[",
StringComparison.Ordinal) == true)
{
diff --git
a/src/Lucene.Net.QueryParser/Flexible/Standard/Parser/EscapeQuerySyntaxImpl.cs
b/src/Lucene.Net.QueryParser/Flexible/Standard/Parser/EscapeQuerySyntaxImpl.cs
index 595f50876..182a82caa 100644
---
a/src/Lucene.Net.QueryParser/Flexible/Standard/Parser/EscapeQuerySyntaxImpl.cs
+++
b/src/Lucene.Net.QueryParser/Flexible/Standard/Parser/EscapeQuerySyntaxImpl.cs
@@ -170,10 +170,17 @@ namespace Lucene.Net.QueryParsers.Flexible.Standard.Parser
}
if (found)
{
+#if FEATURE_STRINGBUILDER_APPEND_READONLYSPAN
+ result.Append(@string.ToString().AsSpan(copyStart,
firstIndex - copyStart));
+ result.Append(escapeChar);
+ result.Append(@string.ToString().AsSpan(firstIndex,
+ (firstIndex + sequence1Length) - firstIndex));
+#else
result.Append(@string.ToString().Substring(copyStart,
firstIndex - copyStart));
result.Append(escapeChar);
result.Append(@string.ToString().Substring(firstIndex,
(firstIndex + sequence1Length) - firstIndex));
+#endif
copyStart = start = firstIndex + sequence1Length;
}
else
@@ -184,7 +191,11 @@ namespace Lucene.Net.QueryParsers.Flexible.Standard.Parser
if (result.Length == 0 && copyStart == 0)
return @string;
+#if FEATURE_STRINGBUILDER_APPEND_READONLYSPAN
+ result.Append(@string.ToString().AsSpan(copyStart));
+#else
result.Append(@string.ToString().Substring(copyStart));
+#endif
return result.ToString().AsCharSequence();
}
diff --git
a/src/Lucene.Net.QueryParser/Flexible/Standard/Parser/StandardSyntaxParser.cs
b/src/Lucene.Net.QueryParser/Flexible/Standard/Parser/StandardSyntaxParser.cs
index 7bb9afd6d..4421b52e3 100644
---
a/src/Lucene.Net.QueryParser/Flexible/Standard/Parser/StandardSyntaxParser.cs
+++
b/src/Lucene.Net.QueryParser/Flexible/Standard/Parser/StandardSyntaxParser.cs
@@ -633,7 +633,11 @@ namespace Lucene.Net.QueryParsers.Flexible.Standard.Parser
if (fuzzy)
{
// LUCENENET specific: parse without throwing
exceptions
- float fms =
float.TryParse(fuzzySlop.Image.Substring(1), NumberStyles.Float,
CultureInfo.InvariantCulture, out float temp) ? temp : defaultMinSimilarity;
+#if FEATURE_NUMBER_PARSE_READONLYSPAN
+ float fms = float.TryParse(fuzzySlop.Image.AsSpan(1),
NumberStyles.Float, CultureInfo.InvariantCulture, out float temp) ? temp :
defaultMinSimilarity;
+#else
+ float fms =
float.TryParse(fuzzySlop.Image.Substring(1), NumberStyles.Float,
CultureInfo.InvariantCulture, out float temp) ? temp: defaultMinSimilarity;
+#endif
if (fms < 0.0f)
{
{ if (true) throw new
ParseException(QueryParserMessages.INVALID_SYNTAX_FUZZY_LIMITS); }
@@ -766,8 +770,12 @@ namespace Lucene.Net.QueryParsers.Flexible.Standard.Parser
if (fuzzySlop != null)
{
- // LUCENENET: don't let parsing throw exceptions
+ // LUCENENET specific: parse without throwing
exceptions
+#if FEATURE_NUMBER_PARSE_READONLYSPAN
+ if (float.TryParse(fuzzySlop.Image.AsSpan(1),
NumberStyles.Float, CultureInfo.InvariantCulture, out float temp))
+#else
if (float.TryParse(fuzzySlop.Image.Substring(1),
NumberStyles.Float, CultureInfo.InvariantCulture, out float temp))
+#endif
{
try
{
diff --git a/src/Lucene.Net.Replicator/Http/HttpClientBase.cs
b/src/Lucene.Net.Replicator/Http/HttpClientBase.cs
index 880310841..2ea4aa574 100644
--- a/src/Lucene.Net.Replicator/Http/HttpClientBase.cs
+++ b/src/Lucene.Net.Replicator/Http/HttpClientBase.cs
@@ -440,7 +440,7 @@ namespace Lucene.Net.Replicator.Http
return res;
}
-#if FEATURE_SPAN
+#if FEATURE_STREAM_READ_SPAN
public override int Read(Span<byte> buffer)
{
int res = input.Read(buffer);
diff --git
a/src/Lucene.Net.Suggest/Suggest/Analyzing/AnalyzingInfixSuggester.cs
b/src/Lucene.Net.Suggest/Suggest/Analyzing/AnalyzingInfixSuggester.cs
index 789a84254..8c1c86c5e 100644
--- a/src/Lucene.Net.Suggest/Suggest/Analyzing/AnalyzingInfixSuggester.cs
+++ b/src/Lucene.Net.Suggest/Suggest/Analyzing/AnalyzingInfixSuggester.cs
@@ -783,11 +783,19 @@ namespace Lucene.Net.Search.Suggest.Analyzing
// TODO: apps can try to invert their analysis logic
// here, e.g. downcase the two before checking prefix:
sb.Append("<b>");
+#if FEATURE_STRINGBUILDER_APPEND_READONLYSPAN
+ sb.Append(surface.AsSpan(0, prefixToken.Length - 0));
+#else
sb.Append(surface.Substring(0, prefixToken.Length - 0));
+#endif
sb.Append("</b>");
if (prefixToken.Length < surface.Length)
{
+#if FEATURE_STRINGBUILDER_APPEND_READONLYSPAN
+ sb.Append(surface.AsSpan(prefixToken.Length));
+#else
sb.Append(surface.Substring(prefixToken.Length));
+#endif
}
}
diff --git a/src/Lucene.Net/Util/AttributeSource.cs
b/src/Lucene.Net/Util/AttributeSource.cs
index 2e4f5c0b4..fd5300ece 100644
--- a/src/Lucene.Net/Util/AttributeSource.cs
+++ b/src/Lucene.Net/Util/AttributeSource.cs
@@ -152,16 +152,29 @@ namespace Lucene.Net.Util
int lastPlus =
attributeInterfaceType.FullName.LastIndexOf('+');
if (lastPlus == -1)
{
+#if FEATURE_STRING_CONCAT_READONLYSPAN
+ return string.Concat(
+ attributeInterfaceType.Namespace,
+ ".",
+ attributeInterfaceType.Name.AsSpan(1));
+#else
return string.Concat(
attributeInterfaceType.Namespace,
".",
attributeInterfaceType.Name.Substring(1));
+#endif
}
else
{
+#if FEATURE_STRING_CONCAT_READONLYSPAN
+ return string.Concat(
+ attributeInterfaceType.FullName.AsSpan(0, lastPlus
+ 1),
+ attributeInterfaceType.Name.AsSpan(1));
+#else
return string.Concat(
attributeInterfaceType.FullName.Substring(0,
lastPlus + 1),
attributeInterfaceType.Name.Substring(1));
+#endif
}
}
}
