User Comment Type - IMEUCT

typedef enum

} IMEUCT;

Notes on User Comment

IMEUCT specifies the type of user comment. IFED_UCT_STRING_SJIS is a string in S-Jis format and IFED_UCT_STRING_UNICODE is a string in Unicode format. If a string is null-terminated, the size of data must include the last null char. IFED_UCT_USER_DEFINED means a client can register a word with user comment of this type and use it for whatever reasons the client may have. The IME kernel will not use a word of this type during morphological analysis. When IMEWRD structure is returned via GetNextWord, the pvComment field will be set to NULL and cbComment will be zero if there are no user comment for that word. Otherwise, the IMEUCT field specifies the type of user comment. The pvComment is allocated by IFEDictionary object. The size of buffer is the actual data size (cbComment) plus two, and the last two bytes are set to zero. This allows non-zero-terminated strings to be made zero-terminated without copying string to another location with a bigger buffer. The client must free the pvComment buffer.

Dictionary Header - IMESHF

typedef struct {
    WORD cbShf; //size of SHF
    WORD verDic; //user specified dictionary version
    CHAR szTitle[48]; //title of dictionary
    CHAR szDescription[256]; //description of dictionary
    CHAR szCopyright[128]; //copyright info
} IMESHF;

Creating IFEDictionary Instance

This creates an IFEDictionary object, and returns a pointer to the object. This is one way of obtaining an interface pointer to IFEDictionary. Another way of obtaining this pointer is via IFECommon interface, using the QueryInterface method with IID_IFEDictionary interface id defined in msime.h.

This is a DLL function.

HRESULT CreateIFEDictionaryInstance

Parameters:

void **ppvObj

(OUT) A pointer to the created object is placed

Return Values:

HRESULT

S_OK E_OUTOFMEMORY E_FAIL

Operating Dictionary file object

HRESULT IFEDictionary::Open

This opens an existing dictionary file and associates it with this IFEDictionary object.

Parameters:

CHAR * pchFileName	(IN) Points to a file name (NULL terminated) string to be opened. If pchFileName is NULL or an empty string, user dictionary opened by the IME kernel will be used. If pchFileName is an empty string, name of user dictionary will be copied into pchFileName. In this case, the size of pchFileName must be MAX_PATH.
IMESHF * pSHF	(OUT) header of the opened file is stored at pSHF, when the opened file is a NANAMI binary dictionary. Can be NULL.

Return Values:

HRESULT

S_OK JDIC_S_EMPTY_DICTIONARY IFED_E_NOT_FOUND IFED_E_INVALID_FORMAT IFED_E_OPEN_FAILED E_FAIL

To implement a multiple dictionary facility, multiple open/release procedure must be carried out.

HRESULT IFEDictionary::Close

This closes the file associated to this object.

Parameters:    None

Return Values

HRESULT

S_OK E_FAIL

HRESULT IFEDictionary::GetHeader

This method obtains the header of a dictionary without opening the dictionary.

Parameters:

CHAR * pchFileName	(IN) Points to a file name (NULL terminated) string.
IMESHF * pSHF	(OUT) header of the dictionary is stored at pSHF, when the opened file is a NANAMI binary dictionary. Can be NULL.
IMEFMT *pjfmt	IFED_UNKNOWN, IFED_ACTIVE_DICT, IFED_MSIME_BIN_SYSTEM, IFED_MSIME_BIN_USER, IFED_MSIME_SYSTEM_CE, IFED_MSIME2_BIN_SYSTEM, IFED_MSIME2_BIN_USER, IFED_MSIME95_BIN_SYSTEM, IFED_MSIME95_BIN_USER, IFED_MSIME97_BIN_SYSTEM, IFED_MSIME97_BIN_USER,
IMEDICAT *pdicat	(OUT) dictionary type is returned. dicatNone  undefined dicatGeneral  general dictionary dicatNamePlace  name/place dictionary dicatSpeech  speech dictionary dicatReverse  reverse dictionary dicatEnglish  English dictionary dicatALL  or of above flags

Return Values:

HRESULT

S_OK IFED_E_INVALID_FORMAT E_FAIL

HRESULT IFEDictionary::SetHeader

This sets/modifies the dictionary header of "this" IFEDictionary object. This set/modifies a header of dictionary object.

Parameters:

IMESHF * pSHF

(IN) Points to a dictionary header structure.

Return Values:

HRESULT

S_OK E_FAILED

HRESULT IFEDictionary::Create

This creates a User dictionary and associates it with "this" IFEDictionary object.

Parameters:

CHAR * pchFileName	(IN) Points to a file path (Null terminated) string to be created.
IMESHF * pSHF	(IN) Points to a dictionary header structure.

Return Values:

HRESULT

S_OK IFED_S_EMPTY_DICTIONARY E_OUTOFMEMORY E_FAIL

ULONG IFEDictionary::Release

This frees the IFEDictionary object.

Getting and Registering Words

HRESULT IFEDictionary::GetPosTable

This method obtains the public POS (part of speech) table. Since this table is language dependent and IME dependent, clients must include an appropriate header for a particular language. For Japanese, the header is named " msime.h" and includes the following data structure.

POSTBL

//POS table data structure

typedef struct _POSTBL {
    WORD nPos; //pos number
    BYTE *szName; //name of pos
} POSTBL;

Parameters:

POSTBL **prgPosTbl	(OUT) pointer to the array of POSTBL.
int *pcPosTbl	(OUT) pointer to the count of part of speech in POSTBL; can be NULL.

Return Values:

HRESULT

S_OK E_FAIL

HRESULT IFEDictionary::GetWords

This method gets word entries from a dictionary. The selection can be done by some combinations of reading with/without a wildcard at tail, display with/without a wildcard at its tail, and POS. In addition, retrievals by a reading can be done by specifying a range in the Hiragana go-jyuu-on ordering.

Parameters:

WCHAR * pwchFirst,	(IN) Points to a reading string against which IFEDictionary entries are matched. The value must be one of the followings: NULL (meaning "low-value") Hiragana string (full reading to be retrieved) Hiragana string ended by "*" (specifying only leading characters of reading) This can be a starting reading string when a range of words is to be retrieved. In this case, a wildcard must not be used.
WCHAR * pwchLast,	(IN) Points to a reading string. This is used to ending reading string. This must contain the same value of pwchReading when a retrieval is done by a single value, i.e., by not a range value. The value must be one of the followings: NULL (meaning "high-value") Hiragana string (full reading to be retrieved) Hiragana string ended by "*" (specifying only leading characters of reading)
WCHAR * pwchDisplay,	(IN) Points to a display string against which IFEDictionary entries are matched. The value must be one of the followings: NULL (meaning "*") Any Japanese string Japanese string ended by "*"
ULONG ulPos,	The filters of MS-IME public POS. IFED_POS_NONE IFED_POS_NOUN IFED_POS_VERB IFED_POS_ADJECTIVE IFED_POS_ADJECTIVE_VERB IFED_POS_ADVERB IFED_POS_ADNOUN IFED_POS_CONJUNCTION IFED_POS_INTERJECTION IFED_POS_INDEPENDENT IFED_POS_INFLECTIONALSUFFIX IFED_POS_PREFIX IFED_POS_SUFFIX IFED_POS_AFFIX IFED_POS_TANKANJI IFED_POS_IDIOMS IFED_POS_SYMBOLS IFED_POS_PARTICLE IFED_POS_AUXILIARY_VERB IFED_POS_SUB_VERB IFED_POS_DEPENDENT IFED_POS_ALL
ULONG ulSelect	(IN) Bit mask which specifies the query output of a word. Must be logical OR of the followings: IFED_SELECT_NONE IFED_SELECT_READING IFED_SELECT_DISPLAY IFED_SELECT_POS IFED_SELECT_ALL Other bits are reserved for future extensions (translation, definition, etc.)
ULONG ulWordSrc	(IN) Bit flag to specify word source. Must be logical OR of the followings when the IFEDictionary is a user dictionary. IFED_REG_USER IFED_REG_AUTO IFED_REG_ALL Other bits are reserved for future extensions. This parameter must be NULL when the IFEDictionary is a system dictionary object.
UCHAR * pchBuffer	(IN/OUT) The caller provides the buffer and this function fills up the data.
ULONG cbBuffer	(IN) size of pchBuffer
ULONG * pcWrd	(OUT) Number of IMEWRD structure returned in buffer. If more entries is found than the pchBuffer can store, an IFED_S_MORE_ENTRIES will be returned.

Return Values:

HRESULT

S_OK IFED_S_MORE_ENTRIES - Client must call NextWords to get further IMEWRD structures. IFED_E_NO_ENTRY E_OUTOFMEMORY E_FAIL

HRESULT IFEDictionary::NextWords

The following must be used after GetWords to get further words.

Parameters:

UCHAR * pchBuffer	(IN/OUT) The caller provides the buffer and this function fills up the data.
ULONG cbBuffer	(IN) size of pchBuffer
ULONG * pcWrd	(OUT) Number of IMEWRD structure returned in buffer. If more entries is found than the pchBuffer can store, an IFED_S_MORE_ENTRIES will be returned.

Return Values:

HRESULT

S_OK IFED_S_MORE_ENTRIES - Client must call NextWords to get further IMEWRD structures. E_FAIL

HRESULT IFEDictionary::ExistWord

This determines if the specified word exists in the IFEDictionary.

Parameters:

IMEWRD *pwrd

(IN) word being checked

Return Values:

HRESULT

S_OK - word exists S_FALSE - word does not exist E_FAIL - error condition

HRESULT IFEDictionary::RegisterWord

This registers a word in the IFEDictionary.

Parameters:

IMEREG reg	(IN) Register type JDIC_REG_HEAD - register the word at head JDIC_REG_TAIL - register the word at tail IFED_REG_DEL - delete the word.
IMEWRD * pWrd	(IN) Points to a IMEWRD structure.

Return Values

HRESULT

S_OK - success IFED_E_NOT_USER_DIC - this object is not a user dictionary IFED_S_WORD_EXISTS - word is already registered IFED_E_USER_COMMENT - failed to insert user comment S_FALSE - failed to register/delete a word E_FAIL - unexpected error