private static UnicodeTrie GenerateBreakTypeTrie()
{
var graphemeBreakClassValues = UnicodeEnumsGenerator.GetPropertyValueAliases("# Grapheme_Cluster_Break (GCB)");
var graphemeBreakClassMapping = graphemeBreakClassValues.Select(x => x.name).ToList();
var trieBuilder = new UnicodeTrieBuilder();
var graphemeBreakData = ReadBreakData(
"https://www.unicode.org/Public/UCD/latest/ucd/auxiliary/GraphemeBreakProperty.txt");
foreach (var(start, end, graphemeBreakType) in graphemeBreakData)
{
if (!graphemeBreakClassMapping.Contains(graphemeBreakType))
{
continue;
}
if (start == end)
{
trieBuilder.Set(start, (uint)graphemeBreakClassMapping.IndexOf(graphemeBreakType));
}
else
{
trieBuilder.SetRange(start, end, (uint)graphemeBreakClassMapping.IndexOf(graphemeBreakType));
}
}
var emojiBreakData = ReadBreakData("https://unicode.org/Public/emoji/12.0/emoji-data.txt");
foreach (var(start, end, graphemeBreakType) in emojiBreakData)
{
if (!graphemeBreakClassMapping.Contains(graphemeBreakType))
{
continue;
}
if (start == end)
{
trieBuilder.Set(start, (uint)graphemeBreakClassMapping.IndexOf(graphemeBreakType));
}
else
{
trieBuilder.SetRange(start, end, (uint)graphemeBreakClassMapping.IndexOf(graphemeBreakType));
}
}
return(trieBuilder.Freeze());
}
public void SetRange()
{
var trie = new UnicodeTrieBuilder(10, 666);
trie.SetRange(13, 6666, 7788, false);
trie.SetRange(6000, 7000, 9900, true);
Assert.Equal(10u, trie.Get(12));
Assert.Equal(7788u, trie.Get(13));
Assert.Equal(7788u, trie.Get(5999));
Assert.Equal(9900u, trie.Get(6000));
Assert.Equal(9900u, trie.Get(7000));
Assert.Equal(10u, trie.Get(7001));
Assert.Equal(666u, trie.Get(0x110000));
}
/// <summary>
/// Generates the UnicodeTrie for the LineBreak code point ranges.
/// </summary>
public static void GenerateLineBreakTrie()
{
var regex = new Regex(@"^([0-9A-F]+)(?:\.\.([0-9A-F]+))?\s*;\s*(.*?)\s*#");
var builder = new UnicodeTrieBuilder((uint)LineBreakClass.XX);
using (StreamReader sr = GetStreamReader("LineBreak.txt"))
{
string line;
while ((line = sr.ReadLine()) != null)
{
Match match = regex.Match(line);
if (match.Success)
{
var start = match.Groups[1].Value;
var end = match.Groups[2].Value;
var point = match.Groups[3].Value;
if (end?.Length == 0)
{
end = start;
}
builder.SetRange(int.Parse(start, NumberStyles.HexNumber), int.Parse(end, NumberStyles.HexNumber), (uint)Enum.Parse <LineBreakClass>(point), true);
}
}
}
UnicodeTrie trie = builder.Freeze();
using FileStream stream = GetStreamWriter("LineBreak.trie");
trie.Save(stream);
}
public void RunRangeChecks(TestRange[] testRanges, CheckValue[] checkValues)
{
uint initialValue = testRanges[0].value;
uint errorValue = 0x0bad;
var builder = new UnicodeTrieBuilder(initialValue, errorValue);
for (int i = 1; i < testRanges.Length; i++)
{
var r = testRanges[i];
builder.SetRange(r.start, r.end - 1, r.value, r.overwrite);
}
var frozen = builder.Freeze();
int cp = 0;
for (int i = 0; i < checkValues.Length; i++)
{
var v = checkValues[i];
for (; cp < v.codePoint; cp++)
{
Assert.Equal(v.value, builder.Get(cp));
Assert.Equal(v.value, frozen.Get(cp));
}
}
}
private static UnicodeTrie GenerateBreakTypeTrie()
{
var trieBuilder = new UnicodeTrieBuilder();
var graphemeBreakData = ReadBreakData(Path.Combine(UnicodeDataGenerator.Ucd, "auxiliary/GraphemeBreakProperty.txt"));
var emojiBreakData = ReadBreakData(Path.Combine(UnicodeDataGenerator.Ucd, "emoji/emoji-data.txt"));
foreach (var breakData in new [] { graphemeBreakData, emojiBreakData })
{
foreach (var(start, end, graphemeBreakType) in breakData)
{
if (!Enum.TryParse <GraphemeBreakClass>(graphemeBreakType, out var value))
{
continue;
}
if (start == end)
{
trieBuilder.Set(start, (uint)value);
}
else
{
trieBuilder.SetRange(start, end, (uint)value);
}
}
}
return(trieBuilder.Freeze());
}
public void RunRangeChecks(TestRange[] testRanges, CheckValue[] checkValues)
{
uint initialValue = testRanges[0].Value;
const uint errorValue = 0x0bad;
var builder = new UnicodeTrieBuilder(initialValue, errorValue);
for (int i = 1; i < testRanges.Length; i++)
{
TestRange r = testRanges[i];
builder.SetRange(r.Start, r.End - 1, r.Value, r.Overwrite);
}
UnicodeTrie frozen = builder.Freeze();
int cp = 0;
for (int i = 0; i < checkValues.Length; i++)
{
CheckValue v = checkValues[i];
while (cp < v.CodePoint)
{
Assert.Equal(v.Value, builder.Get(cp));
Assert.Equal(v.Value, frozen.Get(cp));
cp++;
}
}
}
public void SetRangeCompacted()
{
var builder = new UnicodeTrieBuilder(10, 666);
builder.SetRange(13, 6666, 7788, false);
builder.SetRange(6000, 7000, 9900, true);
UnicodeTrie trie = builder.Freeze();
Assert.Equal(10u, trie.Get(12));
Assert.Equal(7788u, trie.Get(13));
Assert.Equal(7788u, trie.Get(5999));
Assert.Equal(9900u, trie.Get(6000));
Assert.Equal(9900u, trie.Get(7000));
Assert.Equal(10u, trie.Get(7001));
Assert.Equal(666u, trie.Get(0x110000));
}
public void SetRangeSerialized()
{
var builder = new UnicodeTrieBuilder(10, 666);
builder.SetRange(13, 6666, 7788, false);
builder.SetRange(6000, 7000, 9900, true);
var data = builder.ToBuffer();
var trie = new UnicodeTrie(data);
Assert.Equal(10u, trie.Get(12));
Assert.Equal(7788u, trie.Get(13));
Assert.Equal(7788u, trie.Get(5999));
Assert.Equal(9900u, trie.Get(6000));
Assert.Equal(9900u, trie.Get(7000));
Assert.Equal(10u, trie.Get(7001));
Assert.Equal(666u, trie.Get(0x110000));
}
public void SetRangeSerialized()
{
var builder = new UnicodeTrieBuilder(10, 666);
builder.SetRange(13, 6666, 7788, false);
builder.SetRange(6000, 7000, 9900, true);
using var ms = new MemoryStream();
builder.Freeze().Save(ms);
ms.Position = 0;
var trie = new UnicodeTrie(ms);
Assert.Equal(10u, trie.Get(12));
Assert.Equal(7788u, trie.Get(13));
Assert.Equal(7788u, trie.Get(5999));
Assert.Equal(9900u, trie.Get(6000));
Assert.Equal(9900u, trie.Get(7000));
Assert.Equal(10u, trie.Get(7001));
Assert.Equal(666u, trie.Get(0x110000));
}