public static EasilyConfusedKana FromFile(string path)
{
return(new EasilyConfusedKana(
File.ReadLines(path, Encoding.UTF8)
.Where(line => !line.StartsWith("#"))
.Select(line => line.AsCodePoints().Select(cp => CodePoint.FromInt(cp)))));
}
public IEnumerable <IGrouping <string, CodePoint> > FindRelated(CodePoint codePoint)
{
return(new IGrouping <string, CodePoint>[]
{
new CategoryGrouping <CodePoint>("Similarly looking", FindSimilar(codePoint))
});
}
public RadicalSearcherResult(int start, int length, string text, CodePoint radical)
{
Start = start;
Length = length;
Text = text ?? throw new ArgumentNullException(nameof(text));
Radical = radical ?? throw new ArgumentNullException(nameof(radical));
}
public IEnumerable <Radical> AllRadicals()
{
return(radkfile.Radicals
.Select(rad => (codePoint: CodePoint.FromInt(rad.CodePoint), strokeCount: rad.StrokeCount))
.Select(p => new Radical(p.codePoint, p.strokeCount))
.OrderBy(r => r.StrokeCount));
}
public IEnumerable <IGrouping <string, CodePoint> > FindRelated(CodePoint codePoint)
{
similar.TryGetValue(codePoint.ToString(), out var resultList);
IEnumerable <string> result = resultList ?? Enumerable.Empty <string>();
return(EnumerableExt.OfSingle(new CategoryGrouping <CodePoint>("Similar Kanji",
result.Select(r => CodePoint.FromString(r)))));
}
public IEnumerable <CodePoint> LookupByRadicals(IEnumerable <CodePoint> radicals)
{
return(radkfile.LookupMatching(radicals.Select(r => r.ToString()))
.OrderBy(r => kanjidict.Lookup(r)
.Map(e => e.StrokeCount)
.ValueOr(int.MaxValue))
.Select(cp => CodePoint.FromString(cp)));
}
public IEnumerable <CodePoint> LookupRelatedCharacters(CodePoint point)
{
return(EnumerableExt.IntersperseSequencesWith(new[]
{
kanaProperties.FindSimilar(point),
confused.FindSimilar(point)
},
null as CodePoint));
}
public WordInfo(string word, PartOfSpeech partOfSpeech = PartOfSpeech.Unknown, string notInflected = null, bool?isIndependent = null, Option <EdictType> type = default(Option <EdictType>), IEnumerable <PartOfSpeechInfo> posInfo = null)
{
RawWord = word;
CodePoints = new List <CodePoint>(
word.AsCodePoints().Select(cp => CodePoint.FromInt(cp)));
EstimatedPartOfSpeech = partOfSpeech;
NotInflected = notInflected;
Independent = isIndependent;
Type = type;
this.PartOfSpeechInfo = posInfo?.ToList() ?? Enumerable.Empty <PartOfSpeechInfo>();
}
public Option <IEnumerable <CodePoint> > LookupRadicalsByKanji(Kanji kanji)
{
if (remapper == null)
{
return(kradfile.LookupRadicals(kanji.ToString())
.Map(radicals => radicals.Select(cp => CodePoint.FromString(cp))));
}
return(remapper
.LookupRadicals(kanji.ToString())
.Map(radicals => radicals.Select(cp => CodePoint.FromString(cp))));
}
public IEnumerable <CodePoint> FindSimilar(CodePoint point)
{
similarityGroups.TryGetValue(point, out var listOfSimilar);
var similar = listOfSimilar ?? Enumerable.Empty <CodePoint>();
return(similar
.Except(Enumerable.Repeat(point, 1))
.OrderBy(other =>
{
return Math.Abs(position[point.GetType()] - position[other.GetType()]);
}));
}
public Result SelectRadical(IEnumerable<CodePoint> radicals, int sortingCriteriaIndex)
{
var result = new List<CodePoint>();
var possibleRadicals = new KeyValuePair<CodePoint, bool>[radicalCount];
var key = new Vector<ulong>[elementSize].AsSpan();
var vec = AsScalarSpan(key);
foreach (var radical in radicals)
{
var radicalIndex = radicalToIndex[radical.Utf32];
vec[radicalIndex / ulongBitCount] |= (ulong)(1UL << radicalIndex);
}
var s = sortingCriteriaIndex;
var radk = radkinfo[s];
var target = new Vector<ulong>[radk.Length];
for (int i = 0; i < kanjiCount; ++i)
{
for (int j = 0; j < elementSize; ++j)
{
target[i * elementSize + j] = radk[i * elementSize + j] & key[j];
}
}
var possible = new Vector<ulong>[elementSize].AsSpan();
for (int i = 0; i < kanjiCount; ++i)
{
bool isPresent = true;
for (int j = 0; j < elementSize; ++j)
{
if (target[i * elementSize + j] != key[j])
isPresent = false;
}
if (isPresent)
{
result.Add(CodePoint.FromInt(indexToKanji[s][i]));
for (int j = 0; j < elementSize; ++j)
{
possible[j] |= radk[i * elementSize + j];
}
}
}
var possibleUlong = AsScalarSpan(possible);
for (int radicalIndex = 0; radicalIndex < radicalCount; ++radicalIndex)
{
var mask = (ulong)(1UL << radicalIndex);
bool isPresent = (possibleUlong[radicalIndex / ulongBitCount] & mask) == mask;
possibleRadicals[radicalIndex] = new KeyValuePair<CodePoint, bool>(CodePoint.FromInt(indexToRadical[radicalIndex]), isPresent);
}
return new Result(result, possibleRadicals);
}
public WordInfo(
string word,
PartOfSpeech partOfSpeech = PartOfSpeech.Unknown,
string dictionaryForm = null,
Option <EdictPartOfSpeech> type = default,
string reading = null,
string dictionaryFormReading = null)
{
RawWord = word;
CodePoints = new List <CodePoint>(
word.AsCodePoints().Select(cp => CodePoint.FromInt(cp)));
EstimatedPartOfSpeech = partOfSpeech;
DictionaryForm = dictionaryForm;
Type = type;
Reading = reading;
DictionaryFormReading = dictionaryFormReading;
}
public IEnumerable <CodePoint> LookupKanjiByRadicals(IEnumerable <CodePoint> radicals, IKanjiOrdering ordering)
{
if (remapper == null)
{
return(radkfile
.LookupMatching(radicals.Select(s => s.ToString()))
.Select(r => CodePoint.FromString(r))
.OrderBy(x => x, ordering)
.ToList());
}
return(remapper
.LookupKanji(radicals.Select(s => s.ToString()))
.Select(r => CodePoint.FromString(r))
.OrderBy(x => x, ordering)
.ToList());
}
public IEnumerable <CodePoint> FindSimilar(CodePoint point)
{
var oppositeSizedCp = OppositeSizedVersionOf(point.Utf32);
var oppositeSized = (oppositeSizedCp != null
? Enumerable.Repeat(char.ConvertFromUtf32(oppositeSizedCp.Value), 1)
: Enumerable.Empty <string>())
.Select(s => CodePoint.FromString(s));
mapping.TryGetValue(point.ToString(), out var restStr);
restStr = restStr ?? new List <string>();
// TOFIX: support combo kana
var rest = restStr
.Where(s => s.Length == 1)
.Select(s => CodePoint.FromString(s));
return(oppositeSized.Concat(rest));
}
public IEnumerable <IGrouping <string, CodePoint> > FindRelated(CodePoint codePoint)
{
var result = new List <IGrouping <string, CodePoint> >();
if (hiraganaKatakanaMap.TryGetValue(codePoint.Utf32, out var katakana))
{
result.Add(new CategoryGrouping <CodePoint>("Katakana", new[] { CodePoint.FromInt(katakana), }));
}
if (hiraganaKatakanaMap.TryGetKey(codePoint.Utf32, out var hiragana))
{
result.Add(new CategoryGrouping <CodePoint>("Hiragana", new[] { CodePoint.FromInt(hiragana), }));
}
if (smallLargeMap.TryGetValue(codePoint.Utf32, out var large))
{
result.Add(new CategoryGrouping <CodePoint>("Large", new [] { CodePoint.FromInt(large), }));
}
if (smallLargeMap.TryGetKey(codePoint.Utf32, out var small))
{
result.Add(new CategoryGrouping <CodePoint>("Small", new[] { CodePoint.FromInt(small), }));
}
if (regularDakutenMap.TryGetValue(codePoint.Utf32, out var dakuten) ||
(regularHandakutenMap.TryGetKey(codePoint.Utf32, out var r1) &&
regularDakutenMap.TryGetValue(r1, out dakuten)))
{
result.Add(new CategoryGrouping <CodePoint>("Dakuten", new [] { CodePoint.FromInt(dakuten), }));
}
if (regularHandakutenMap.TryGetValue(codePoint.Utf32, out var handakuten) ||
(regularDakutenMap.TryGetKey(codePoint.Utf32, out var r2) &&
regularHandakutenMap.TryGetValue(r2, out handakuten)))
{
result.Add(new CategoryGrouping <CodePoint>("Handakuten", new[] { CodePoint.FromInt(handakuten), }));
}
if (regularDakutenMap.TryGetKey(codePoint.Utf32, out var regular) ||
regularHandakutenMap.TryGetKey(codePoint.Utf32, out regular))
{
result.Add(new CategoryGrouping <CodePoint>("Regular", new[] { CodePoint.FromInt(regular), }));
}
return(result);
}
public Radical(CodePoint cp, int strokeCount)
{
CodePoint = cp;
StrokeCount = strokeCount;
}
public static Func <string, bool> CreateMatcher(
this ILanguageService lang,
IReadOnlyDictionary <CodePoint, KanjiPlaceholder> haystack,
string template)
{
var privateUseAreaMatch = new Regex(@"\p{Co}");
var regex = new Regex("^" + privateUseAreaMatch.Replace(Regex.Escape(template).Replace(@"/\\", "."), ".") + "$");
return(word => regex.IsMatch(word) && KanjiPlaceholdersMatch(template, word));
bool KanjiPlaceholdersMatch(string t, string c)
{
t = t.Replace(@"/\\", ".");
foreach (var(templateChar, concreteChar) in t.AsCodePoints().Zip(c.AsCodePoints(),
(l, r) => (CodePoint.FromInt(l), CodePoint.FromInt(r))))
{
if (!haystack.TryGetValue(templateChar, out var placeholder))
{
continue;
}
var concreteRadicals = new HashSet <CodePoint>(concreteChar is Kanji k
? lang.LookupRadicals(k).ValueOr(Enumerable.Empty <CodePoint>())
: Enumerable.Empty <CodePoint>());
if (placeholder.Radicals.All(templateRadical =>
concreteRadicals.Contains(templateRadical.CodePoint)))
{
continue;
}
else
{
return(false);
}
}
return(true);
}
}
public int Compare(CodePoint x, CodePoint y)
{
return(comparer.Compare(x, y));
}
public KanjiRadicalLookup(IEnumerable<Radkfile.Entry> entries, KanjiDict kanjiDict)
{
SortingCriteria = new ReadOnlyListWithSelector<IKanjiOrdering>(new IKanjiOrdering[]
{
KanjiOrdering.Create("Sort by stroke count", kanjiDict, x => x.StrokeCount),
KanjiOrdering.Create("Sort by frequency", kanjiDict, x => x.FrequencyRating)
});
SortingCriteria.SelectedIndex = 0;
var entryList = entries.ToList();
radicalCount = entryList.Count;
elementSize = DivideRoundUp(radicalCount, vectorBitCount);
elementSize = elementSize == 0 ? 1 : elementSize;
var kradMapping = entryList
.ToDictionary(entry => entry.Radical.CodePoint, entry => entry.KanjiCodePoints.AsEnumerable())
.InvertMappingToSequence();
var kanjiCodePoints = entryList
.SelectMany(entry => entry.KanjiCodePoints)
.Distinct()
.ToArray();
kanjiCount = kanjiCodePoints.Length;
indexToKanji = SortingCriteria
.Select(sortingCriterion => kanjiCodePoints
.OrderBy(x => x, Comparer<int>.Create((l, r) => sortingCriterion.Compare(
CodePoint.FromInt(l),
CodePoint.FromInt(r))))
.ToArray())
.ToArray();
indexToRadical = entryList
.Select(entry => entry.Radical.CodePoint)
.ToArray();
radicalToIndex = indexToRadical
.Indexed()
.ToDictionary(p => p.element, p => p.index);
var kanjiToIndex = indexToKanji
.Select(a => a
.Indexed()
.ToDictionary(p => p.element, p => p.index))
.ToArray();
radkinfo = Enumerable.Range(0, SortingCriteria.Count)
.Select(CreateRadkInfo)
.ToArray();
Vector<ulong>[] CreateRadkInfo(int x)
{
var r = new Vector<ulong>[kanjiCount * elementSize];
foreach (var kanji in kanjiCodePoints)
{
var v = new Vector<ulong>[elementSize];
var kanjiIndex = kanjiToIndex[x][kanji];
var vec = AsScalarSpan(v);
foreach (var radical in kradMapping[kanji])
{
var radicalIndex = radicalToIndex[radical];
vec[radicalIndex / ulongBitCount] |= (ulong)(1UL << radicalIndex);
}
for (int i = 0; i < elementSize; ++i)
{
r[kanjiIndex * elementSize + i] = v[i];
}
}
return r;
}
}
public CodePoint LookupCharacter(string character, int position = 0)
{
return(CodePoint.FromString(character, position));
}
public IEnumerable <IGrouping <string, CodePoint> > FindRelated(CodePoint codePoint)
{
return(relatedProviders.SelectMany(r => r.FindRelated(codePoint)));
}
public CodePoint LookupCharacter(int codePoint)
{
return(CodePoint.FromInt(codePoint));
}
public IEnumerable <CodePoint> FindSimilar(CodePoint point)
{
return(Enumerable.Empty <CodePoint>());
}