public static async void test()
{
var ctx = new FulltextContext();
//ctx.recreate();
ctx.Database.ExecuteSqlCommand("delete Dicts");
Dict dict;
ctx.Dicts.Add(dict = new Dict {
Name = "Dict 1"
});
for (var idx = 0; idx < 100; idx++)
{
var engPhrase = await Insert(ctx, "Now are you?", null, dict, new PhraseSide { src = Langs.en_gb, dest = Langs.en_gb });
var phrase = await Insert(ctx, "Ahoj, jak se máš?", null, dict, new PhraseSide { src = Langs.en_gb, dest = Langs.cs_cz }, engPhrase.Id);
ctx.SaveChanges();
var search = await SearchPhrase(new PhraseSide { src = Langs.en_gb, dest = Langs.cs_cz }, "měj", true);
search = await SearchPhrase(new PhraseSide { src = Langs.en_gb, dest = Langs.cs_cz }, "měj", false);
await Insert(ctx, "Ahoj, jak se máš? Asi dobře Kadle.", phrase.Id, null, null);
await Insert(ctx, "Asi dobře, Karle.", phrase.Id, null, null);
await Insert(ctx, null, engPhrase.Id, null, null);
ctx.SaveChanges();
//await Insert(null, phrase.Id, null);
}
return;
//Parallel.ForEach(Enumerable.Range(0, 1000), async idx => {
// var res = await RunSpellCheck.Check(Langs.de_de, new string[] {
// "einem", "Pferd", "die", "Sporen", "geben"
// //"einem", "pferd", "die", "sporen", "geben"
// //"Einem Pferd die Sporen geben."
// //"Klassenbuch"
// });
// res = null;
//});
//Parallel.ForEach(Enumerable.Range(0, 1000), async idx => {
// var s = "yyy"; for (var i = 0; i < idx; i++) s += "x";
// var res = await RunSpellCheckWords.Check(Langs.cs_cz, new string[] {
// //"einem", "Pferd", "die", "Sporen", "geben", s
// s, "ahoj", "jak", "se", "máš"
// //"einem", "pferd", "die", "sporen", "geben"
// //"Einem Pferd die Sporen geben."
// //"Klassenbuch"
// });
// if (idx >= 999)
// res = null;
//});
return;
//public static async LangsLib.PhraseWord[] spellCheckedWordBreak(Langs lang, string text) {
//for (var idx = 0; idx < 1000; idx++) {
// var res = await SpellChecker.SpellLang.Check(Langs.de_de, new string[] { });
//}
//var stemmer = new StemmerBreaker.Runner(lang);
//var words = stemmer.wordBreak(text);
//var errors = await RunSpellCheck.Check(lang, words.Select(idx => text.Substring(idx.srcPos, idx.srcLen))) as SpellLangResult;
//if (errors != null) { };
//return new LangsLib.PhraseWord[0];
Langs[] langs = new Langs[] { Langs.cs_cz, Langs.de_de, Langs.ru_ru, Langs.pt_pt, Langs.sk_sk, Langs.fr_fr, Langs.it_it, Langs.es_es };
//for (var idx = 0; idx < 1000; idx++) {
// var s = ""; for (var i = 0; i < idx; i++) s += "x";
// var res = await SpellChecker.SpellLang.Check(langs[idx % 8], new string[] {
// //s + " ahoj" + " jak" + " se" + " máš"
// s, "ahoj", "jak", "se", "máš"
// });
//}
//Parallel.ForEach(Enumerable.Range(0, 1000), async idx => {
// var s = ""; for (var i = 0; i < idx; i++) s += "x";
// var res = await SpellChecker.RunSpellCheck.Check(langs[idx % 8], new string[] {
// //s + " ahoj" + " jak" + " se" + " máš"
// s, "ahoj", "jak", "se", "máš"
// });
//});
//return null;
//SpellChecker.RunSpellCheck.Check(lang, new string[] { null });
//spellRes = SpellChecker.SpellLang.Check(lang, text);
//using (var rn = new StemmerBreaker.Runner(lang))
// rn.wordBreak(text);
//spellRes = SpellChecker.SpellLang.Check(lang, text);
//using (var rn = new StemmerBreaker.Runner(lang))
// rn.wordBreak(text);
//spellRes = SpellChecker.SpellLang.Check(lang, text);
//IEnumerable<StemmerBreaker.Put> stBrRes = stBr.wordBreak(text);
//if (spellRes != null) stBrRes = stBrRes.Where(wb => spellRes.All(br => br.pos!= wb.srcPos));
//var words = stBrRes.Select(wb => text.Substring(wb.srcPos, wb.srcLen)).ToArray();
//return null;
//public static string[] toStrings(string text, List<Put> idxs)
//{
// return idxs.Select(idx => text.Substring(idx.srcPos, idx.srcLen)).ToArray();
//}
}
public RunInsertPhrase(FulltextContext ctx, string newWords, int?phraseId, Dict dict, PhraseSide?phraseSide, int?srcSideId)
{
this.phraseId = phraseId; this.phraseSide = phraseSide; this.newWords = newWords; this.srcSideId = srcSideId; this.ctx = ctx; this.dict = dict;
}
public static Task <Phrase> Insert(FulltextContext ctx, string newWords, int?phraseId, Dict dict, PhraseSide?phraseSide, int?srcSideId = null)
{
return(Lib.Run(new RunInsertPhrase(ctx, newWords, phraseId, dict, phraseSide, srcSideId)));
}
public static Phrase STAInsert(FulltextContext ctx, string newPhraseText /*NullOrEmpty => delete, else update or insert*/, int?phraseId /*==null => insert else update or delete*/, Dict dict /*my dict for insert*/, PhraseSide?phraseSide /*for Insert: dict and its side, e.g. czech part of English-Czech dict*/, int?srcSideId /*for inserting Destination side*/)
{
if (string.IsNullOrEmpty(newPhraseText)) //DELETE
{
if (phraseId == null || phraseSide != null || dict != null)
{
throw new Exception("phraseId == null || phraseSide!=null || dict!=null");
}
var delPh = ctx.Phrases.Include(p => p.Dests).First(p => p.Id == phraseId);
ctx.Phrases.RemoveRange(delPh.Dests);
ctx.Phrases.Remove(delPh);
return(null);
}
Phrase dbPhrase; PhraseWords oldPhrase = null; PhraseSide ps;
if (phraseId != null) //UPDATE
{
if (srcSideId != null || dict != null)
{
throw new Exception("srcSideId != null || dict!=null");
}
dbPhrase = ctx.Phrases.Include(p => p.Words).First(p => p.Id == phraseId);
oldPhrase = new PhraseWords(dbPhrase.Text, TPosLen.fromBytes(dbPhrase.TextIdxs));
ps = new PhraseSide {
src = (Langs)dbPhrase.Dict.SrcLang, dest = (Langs)dbPhrase.DestLang
};
}
else //INSERT
{
if (phraseSide == null || dict == null)
{
throw new Exception("phraseSide == null || dict == null");
}
ps = (PhraseSide)phraseSide;
if (ps.src != ps.dest && srcSideId == null)
{
throw new Exception("ps.src!=ps.dest && srcSideId == null");
}
ctx.Phrases.Add(dbPhrase = new Phrase {
SrcLang = (byte)ps.src, DestLang = (byte)ps.dest, SrcRef = srcSideId, Dict = dict
});
}
var lang = ps.langOfText(); var newPhrase = new PhraseWords(newPhraseText);
Action <SelectedWord[]> spellCheckAndDBInsert = wordIdxs => {
STASpellCheck(lang, wordIdxs, newPhrase); //low level spell check
for (var i = 0; i < wordIdxs.Length; i++)
{
if (newPhrase.Idxs[wordIdxs[i].idx].Len > 0) //new correct words to fulltext DB
{
ctx.PhraseWords.Add(new PhraseWord()
{
SrcLang = (byte)ps.src, DestLang = (byte)ps.dest, Word = wordIdxs[i].ftxWord, Phrase = dbPhrase
});
}
}
};
//Word breaking
STAWordBreak(lang, newPhrase);
var newWordIdx = getCorrectPhraseWords(newPhrase);
if (oldPhrase == null) //INSERT
{
spellCheckAndDBInsert(newWordIdx); //Spell check and add to DB
}
else //UPDATE
//Delete olds from DB
{
var oldWordIdx = getCorrectPhraseWords(oldPhrase);
var dbOldWords = dbPhrase.Words;
foreach (var w in oldWordIdx.Except(newWordIdx))
{
ctx.PhraseWords.Remove(dbOldWords.First(db => db.Word == w.ftxWord));
}
//Spell check and add to DB
var newWithoutOldWordIdx = newWordIdx.Except(oldWordIdx).ToArray();
spellCheckAndDBInsert(newWithoutOldWordIdx);
}
dbPhrase.Text = newPhrase.Text;
dbPhrase.TextIdxs = TPosLen.toBytes(newPhrase.Idxs);
dbPhrase.Base = newPhrase.Idxs.Select(idx => newPhrase.Text.Substring(idx.Pos, Math.Abs(idx.Len)).ToLower()).DefaultIfEmpty().Aggregate((r, i) => r + "|" + i);
if (string.IsNullOrEmpty(dbPhrase.Base))
{
dbPhrase.Base = ""; //error
}
if (dbPhrase.Base.Length > Phrase.maxPhraseBaseLen)
{
dbPhrase.Base = dbPhrase.Base.Substring(0, Phrase.maxPhraseBaseLen);
}
return(dbPhrase);
}
