//エントリポイント\r
int i, i_max;\r
CHNLIB_String *input, *temp;\r
- CHNLIB_UIPArray *separated;\r
+ CHNLIB_UIPArray *separated, *sorted;\r
int passthink;\r
+ FILE *readfp;\r
\r
CHNLIB_Environment_SetCurrentWorkingDirectory(argv[0]);\r
\r
AI_System_InitializeSystemWorkingSet();\r
\r
AI_System_LoadMemory(AI_CONFIG_FILE_NAME);\r
+ \r
+ readfp = NULL;\r
\r
for(;;){\r
passthink = False;\r
- input = CHNLIB_ReadLine(stdin);\r
+ if(readfp == NULL){\r
+ input = CHNLIB_ReadLine(stdin);\r
+ } else{\r
+ input = CHNLIB_ReadLine(readfp);\r
+ if(input == NULL){\r
+ readfp = NULL;\r
+ } else{\r
+ puts(CHNLIB_String_GetReferencePointerOfCString(input));\r
+ }\r
+ }\r
\r
if(CHNLIB_String_CompareStringWithCString(input, "::")){\r
//システムコマンド解釈\r
} else if(CHNLIB_String_CompareStringWithCString(temp, "exit")){\r
break;\r
} else if(CHNLIB_String_CompareStringWithCString(temp, "wordlist")){\r
- i_max = CHNLIB_UIPArray_GetNumberOfDatas(WorkingSet.RootWordList);\r
+ sorted = CHNLIB_UIPArray_SortInDescendingOrderByData32(WorkingSet.RootWordList);\r
+ i_max = CHNLIB_UIPArray_GetNumberOfDatas(sorted);\r
for(i = 0; i < i_max; i++){\r
- printf("word%3d:%s\n", i, CHNLIB_String_GetReferencePointerOfCString(CHNLIB_UIPArray_GetPointerByIndex(WorkingSet.RootWordList, i)));\r
+ printf("word%3d:%3d:%s\n", i, CHNLIB_UIPArray_GetData32ByIndex(sorted, i), CHNLIB_String_GetReferencePointerOfCString(CHNLIB_UIPArray_GetPointerByIndex(sorted, i)));\r
+ }\r
+ CHNLIB_UIPArray_FreeOnlyArray(&sorted);\r
+ } else if(CHNLIB_String_CompareStringWithCString(temp, "readfile")){\r
+ //::readfile:filename\r
+ readfp = fopen(CHNLIB_String_GetReferencePointerOfCString(CHNLIB_UIPArray_GetPointerByIndex(separated, 4)), "rb");\r
+ if(readfp == NULL){\r
+ puts("File open error.");\r
}\r
} else{\r
passthink = False;\r
}\r
}\r
\r
+ //WorkingSet.SystemWordList0を使っているので、解放しないように注意\r
CHNLIB_UIPArray_FreeSelectedAll(&separated);\r
}\r
\r
if(!passthink){\r
//AIへの入力\r
- separated = CHNLIB_UIPArray_Initialize();\r
- separated = AI_Think_SlideLookUpWordByHistory(input);\r
- i_max = CHNLIB_UIPArray_GetNumberOfDatas(separated);\r
- printf("Index(Decimal),CountOfContain(Decimal), String\n");\r
- for(i = 0; i < i_max; i++){\r
- printf("%d,%d,%s\n", i, CHNLIB_UIPArray_GetData32ByIndex(separated, i), CHNLIB_String_GetReferencePointerOfCString(CHNLIB_UIPArray_GetPointerByIndex(separated, i)));\r
- }\r
- CHNLIB_UIPArray_FreeAll(&separated);\r
+ //::readfile:AITestData_ja.txt\r
+ AI_Think_LearnWordFromInputString(input);\r
+ \r
CHNLIB_UIPArray_AppendLast(&WorkingSet.InputHistory, CHNLIB_UIPArray_GetNumberOfDatas(WorkingSet.InputHistory), input);\r
}\r
}\r
return 0;\r
}\r
\r
+void AI_Think_LearnWordFromInputString(CHNLIB_String *input)\r
+{\r
+ //入力文字列から単語を抽出して記憶する。\r
+ CHNLIB_UIPArray *candidateWordList;\r
+ int i, i_max;\r
+ int tagIndex, maxExistingWordIndex;\r
+ \r
+ candidateWordList = CHNLIB_UIPArray_Initialize();\r
+ candidateWordList = AI_Think_SlideLookUpWordByHistory(input);\r
+ i_max = CHNLIB_UIPArray_GetNumberOfDatas(candidateWordList);\r
+ maxExistingWordIndex = CHNLIB_UIPArray_GetNumberOfDatas(WorkingSet.RootWordList) - 1;\r
+ \r
+ printf("Index(Decimal),CountOfContain(Decimal), String\n");\r
+ for(i = 0; i < i_max; i++){\r
+ printf("%d,%d,%s\n", i, CHNLIB_UIPArray_GetData32ByIndex(candidateWordList, i), CHNLIB_String_GetReferencePointerOfCString(CHNLIB_UIPArray_GetPointerByIndex(candidateWordList, i)));\r
+ \r
+ //単語リストに登録。\r
+ tagIndex = AI_Memory_AddRootWordData(CHNLIB_UIPArray_GetPointerByIndex(candidateWordList, i));\r
+ \r
+ CHNLIB_UIPArray_SetData32ByIndex(WorkingSet.RootWordList, tagIndex, CHNLIB_UIPArray_GetData32ByIndex(WorkingSet.RootWordList, tagIndex) + CHNLIB_UIPArray_GetData32ByIndex(candidateWordList, i));\r
+ \r
+ if(tagIndex <= maxExistingWordIndex){\r
+ //既存のタグなので、解放されるようにdata32をFalseにしておく。\r
+ CHNLIB_UIPArray_SetData32ByIndex(candidateWordList, i, False);\r
+ }\r
+ //新規追加されたタグについては0以上の値つまりTrueが設定されているはずなので、解放されない。\r
+ }\r
+ //data32 == Falseのタグ、つまり既存のタグと同じだったもののみ解放する。\r
+ CHNLIB_UIPArray_FreeSelectedAll(&candidateWordList);\r
+\r
+ return;\r
+}\r
+\r
CHNLIB_UIPArray *AI_Think_SlideLookUpWordByHistory(CHNLIB_String *input)\r
{\r
//[UTF-8]\r
CHNLIB_UIPArray_SetData32ByIndex(candidatewordlist, k, j);\r
}\r
\r
+ //重複抽出フィルタリング\r
+ AI_Think_CandidateWordList_Filter01(&candidatewordlist, 1);\r
+ AI_Think_CandidateWordList_Filter00(&candidatewordlist);\r
+ \r
return candidatewordlist;\r
}\r
\r
-int AI_Think_CandidateWordList_Filter00(CHNLIB_UIPArray *candidatewordlist)\r
+int AI_Think_CandidateWordList_Filter00(CHNLIB_UIPArray **candidatewordlist)\r
{\r
//[Not implemented]\r
//[UTF-8]\r
//AI_Think_SlideLookUpWordByHistoryで返されたArray内の候補単語に対して、フィルターをかける。\r
+ //00:長い単語に含まれており、かつ出現頻度が長い単語と等しい単語を削除\r
int i, i_max;\r
CHNLIB_String *nowstr, *basestr;\r
- int nowstrlen, basestrlen;\r
+ int basestrCoC; //CountOfContainInHistoryStrings\r
\r
- i_max = CHNLIB_UIPArray_GetNumberOfDatas(candidatewordlist);\r
- basestr = CHNLIB_UIPArray_GetPointerByIndex(WorkingSet.InputHistory, 0);\r
- //basestrlen =\r
+ i_max = CHNLIB_UIPArray_GetNumberOfDatas(*candidatewordlist);\r
+ basestr = CHNLIB_UIPArray_GetPointerByIndex(*candidatewordlist, 0);\r
+ basestrCoC = CHNLIB_UIPArray_GetData32ByIndex(*candidatewordlist, 0);\r
for(i = 1; i < i_max; i++){\r
- nowstr = CHNLIB_UIPArray_GetPointerByIndex(WorkingSet.InputHistory, i);\r
+ nowstr = CHNLIB_UIPArray_GetPointerByIndex(*candidatewordlist, i);\r
+ if(CHNLIB_UTF8_GetCountOfContain(CHNLIB_String_GetReferencePointerOfCString(basestr), CHNLIB_String_GetReferencePointerOfCString(nowstr)) > 0){\r
+ //nowstrはbasestrに含まれている\r
+ if(basestrCoC == CHNLIB_UIPArray_GetData32ByIndex(*candidatewordlist, i)){\r
+ //かつ出現頻度が等しいので不要な単語\r
+ //出現頻度を0にして、後で削除する。\r
+ CHNLIB_UIPArray_SetData32ByIndex(*candidatewordlist, i, 0);\r
+ }\r
+ }\r
\r
+ if(CHNLIB_UIPArray_GetData32ByIndex(*candidatewordlist, i) != 0){\r
+ //単語は削除されなかった、つまり異なる単語なので、basestrを更新\r
+ basestr = nowstr;\r
+ basestrCoC = CHNLIB_UIPArray_GetData32ByIndex(*candidatewordlist, i);\r
+ }\r
}\r
+ \r
+ for(i = 1; i < i_max; i++){\r
+ if(CHNLIB_UIPArray_GetData32ByIndex(*candidatewordlist, i) == 0){\r
+ //削除対象\r
+ CHNLIB_UIPArray_RemoveByIndex(candidatewordlist, i);\r
+ i--;\r
+ i_max--;\r
+ }\r
+ }\r
+ \r
return 0;\r
}\r
\r
-void AI_Memory_AddRootWordData(CHNLIB_String *tag)\r
+int AI_Think_CandidateWordList_Filter01(CHNLIB_UIPArray **candidatewordlist, int length)\r
{\r
- //WorkingSet.RootWordListに文字列を追加する。\r
- //重複がある場合は警告を出し、追加しない。\r
+ //[Not implemented]\r
+ //[UTF-8]\r
+ //AI_Think_SlideLookUpWordByHistoryで返されたArray内の候補単語に対して、フィルターをかける。\r
+ //01:length以下の文字数の単語を削除\r
+ \r
int i, i_max;\r
\r
- i_max = CHNLIB_UIPArray_GetNumberOfDatas(WorkingSet.RootWordList);\r
+ i_max = CHNLIB_UIPArray_GetNumberOfDatas(*candidatewordlist);\r
for(i = 0; i < i_max; i++){\r
- if(CHNLIB_String_CompareString_Strict(CHNLIB_UIPArray_GetPointerByIndex(WorkingSet.RootWordList, i), tag)){\r
- CHNLIB_ReportError("Word[%s] has already existed.", CHNLIB_DEBUG_ARGUMENTS, CHNLIB_String_GetReferencePointerOfCString(tag));\r
- return;\r
+ if(CHNLIB_UTF8_GetStringLengthByCharacter(CHNLIB_String_GetReferencePointerOfCString(CHNLIB_UIPArray_GetPointerByIndex(*candidatewordlist, i))) <= length){\r
+ CHNLIB_UIPArray_RemoveByIndex(candidatewordlist, i);\r
+ i--;\r
+ i_max--;\r
}\r
}\r
\r
- CHNLIB_UIPArray_AppendLast(&WorkingSet.RootWordList, i_max, tag);\r
- \r
- CHNLIB_UIPArray_AppendLast_ProtectFromDuplication(&WorkingSet.RootWordList, CHNLIB_UIPArray_GetNumberOfDatas(WorkingSet.RootWordList), tag, &AI_Memory_AddRootWordData_IsDuplicated);\r
+ return 0;\r
+}\r
+\r
+int AI_Memory_AddRootWordData(CHNLIB_String *tag)\r
+{\r
+ //WorkingSet.RootWordListに文字列を追加する。\r
+ //重複がある場合は警告を出し、追加しない。\r
+ //戻り値は追加されたまたは重複したタグのIndex.\r
\r
- return;\r
+ return CHNLIB_UIPArray_AppendLast_ProtectFromDuplication(&WorkingSet.RootWordList, 0, tag, &AI_Memory_AddRootWordData_IsDuplicated);\r
}\r
\r
int AI_Memory_AddRootWordData_IsDuplicated(const void *listtag, const void *newtag)\r
OSDN Git Service
