Destructor.

Activates language in requested input mode. After calling this method language is ready to be used with all input modes it supports.

Adds new auto substitution entry to database. If entry for given shorcut already exists, then the old entry will be automatically deleted.

DEPRECATED (will leave if called).

Loads and constructs new core object. Core object is added to the list of available cores and is ready to be used after that.

Some prediction engines support next word prediction feature but require database to be pre-filled with suitable phrase data. This method adds one new phrase or sentence to phrase dictionary. Active core object needs to support phrase based next word prediction, otherwise an error code is returned.

Adds new entry (in most cases a word) to default user dictionary of currently active core object.

Adds entry to specific user dictionary.

This method handles key press and forwards it to correct core object depending on active language and input mode. The current word buffer is updated accordingly. If input sequence buffer has reached its maximum length then nothing will be done.

Attach default user dictionary.

Attach user dictionary in given file to core for requested input mode.

Returns value indicating currently available character conversions.

Front end processeor uses this method for indicating PtiEngine that all key press related timers should be canceled.

Returns pointer to current Chinese candidate page.

Returns active text case.

Returns text case buffer. Case buffer contains shift state for each key press used for producing active word, ie. it remembers case values for each AppendKeyPress call. Case buffer is cleared when active word is commited or cleared. Return value is TPtrC8 containing string of EPtiTextCase values.

Converts given string of characters to one coding system to another. See definition of TPtiCharConversion for list of supported conversion types. It is possible that only a small subset of supported conversion types is actually available at run time (that depends on available core objects). AvailableCharConversions() method can be used for querying available conversion types before using this method.

Returns key for given character. Returned key depends on current language and input mode.

Clears current word buffer. Calling this method means that current word was reject and will not be part of edited text. Either this method or CommitCurrentWord() must be called before starting a new word.

Closes active language. After calling this method there won't be active language and most PtiEngine API methods will return error until ActivateLanguageL is called again. Core objects for active language are asked to release related resources.

Commits current word. Commiting means that core object is isntructed to end inline editing operation and accepted active word as part of edited text. Core object may then update frequency information, add unrecognized word to user dictioary or perform any other operation related to commiting a word. Word buffer is cleared. Either this method or ClearCurrentWord() must be called before starting a new word.

Returns pointer to composition data interface (used with Japanese input).

Returns core info structure for given input mode.

Creates empty user default dictionary file for given core object and initializes it to PtiEngine user dictionary format. If file already exists, then this method does nothing. Normally this method is only used by core objects.

Creates new user dictionary file, inserts given list of words into it and attaches it to active core object for requested input mode. Active language must support requested input mode.

Returns current input sequence (a list of key presses). Bytes in returned descriptor are TPtiKey enum values.

Returns pointer to currently active language.

Returns pointer to current word buffer.

Returns default user dictionary for given input mode.

Remove auto substitution entry.

Deletes last character in current word buffer and updates candidate list accordingly.

Detaches currently attached user dictionary.

Detaches currently attached user dictionary.

Enables or disables tone marks.

This method is same as GetNumericModeKeysForQwertyL, expect that instead of returning strict list of key bindings used in phone number editor, it returns list of all possible characters used in any of the "number only" editor variations. Returned list includes key bindings for characters: "*+pw#1234567890;.,-E?/" (Not necessarily in this order). See also GetNumericModeKeysForQwertyL.

This version first tries to return mappings according to currently active physical keyboard. It current keyboard is not qwerty based, it searches data for the first qwerty based keyboard type it can find. That is done in same order as keyboard types are defined in PtiDefs.h. There is also another version this method, which gets keyboard type as a parameter.

Same as previous version of ExtendedNumericModeKeysForQwertyL (see description above), but keyboard type is given as a parameter and mappings are return only for given keyboard type.

Returns pointer to first word in candidate list. If there isn't any candidate words the returned pointer will point to empty descriptor.

Return auto substitution entry for given index.

Returns list of available input languages.

Returns list of available input languages.

Returns list of candidate words for current input sequence. If word completion feature is on, then words accepted to result list may contain more letters than the number of key presses in current input sequence.

Fills list with phrase candidates for currently selected phonetic spelling.

Get hwr available languages list

Get first hwr implementation support the specified language

Get hwr implementation by give implementation uid

Returns pointer to requested language.

Returns localized language name for given language. This method is quite inefficient (always reads name table from resource file), when possible use Getlanguage()->LocalizedName() instead. This method can be used also when requested language is not among available input languages.

Returns mode name index table for given Chinese variant.

Returns next word candidate list. This method requires that current core object supports next word prediction feature and it is turned on.

Qwerty input mode has different keymapping layout for each language. Therefore the characters for numeric input mode may be mapped to different keys depending on language. There are several situations where client application needs to know which key and case combination produce numeric characters for given language. This convinience method can be used for extracting that information easily (it is also possible to achieve same result directly via CPtiCoreLanguage object). Result array will be left empty if requested language is not available or it doesn't support qwerty input mode. Returned list includes key bindings for characters: "0123456789pw+#*" (Not necessarily in this order). See also ExtendedNumericModeKeysForQwertyL.

This version first tries to return mappings according to currently active physical keyboard. It current keyboard is not qwerty based, it searches data for the first qwerty based keyboard type it can find. That is done in same order as keyboard types are defined in PtiDefs.h. There is also another version this method, which gets keyboard type as a parameter.

Same as previous version of GetNumericModeKeysForQwertyL (see description above) but keyboard type is given as a parameter and mappings are return only for given keyboard type.

Returns phonetic spelling for current input.

Fills list with all the phonetic spellings for current input sequence.

Returns alternate spelling for given character.

Returns entry for given index in default user dictionary.

General command handling method. This method can be used for controlling core objects that require more information than just sequence of key presses.

Cycles to next tone mark in core related tone mark list.

Returns active input mode.

A convinience method for cheking qwerty based input mode.

Returns boolean value indicating whether current tone mark is valid for spelling.

Returns a boolean value indicating whether given scan code is allowed for current input mode.

Returns currently selected qwerty keyboard type.

Lists keyboard blocks available in keymapping data for given language.

Returns last entered key press.

Returns a list containing keyboard type values for all available physical keyboards connected to the device. Keyboard doesn't have to be active at calling time to be included in the output list.

Get hwr implementation uid list which support given language

Returns keymapping data for given key. Returned data depends on active language and input mode. Result string will be empty if there isn't key mapping adta available.

Some core objects may provide different set of results depending on whether the auto captitalization feature was used for entering the word or not. There is now way to tell on core level whether the word was auto-capitalizedby FEP or capitalized normally by the user. This method can be used to incicate core object that auto-capitalization was used for current input sequence. Clearing or commiting the word will cancel the effect of this method.

Returns a boolean value indicating whether there are more candidate pages available.

Two phase constructor.

Two phase constructor.

NOTE: THIS METHOD IS DEPRECATED AND WILL LEAVE WHEN CALLED.

Two phase constructor.

Returns pointer to next word in candidate list. FirstCandidate() must be called before calling this method. Returns pointer to empty descriptor if there isn't more candidates available.

Changes to next Chinese candidate page.

Returns the number of auto substitution entries in auto subst db.

Returns number of candidate words for current input sequence.

Returns number of entries in default user dictionary.

Returns number of available input languages.

Returns current observer.

Returns a value specifying how many phonetic spellings there is available for current input.

Returns pointer to previous word in candidate list. If there isn't previous candidate available (NextCandidate() wasn't succesfully called) then the return value will point to an empty string.

Changes to previous Chinese candidate page.

Returns a heap buffer containing data from given secondary data file. Returns null if file is not found.

Returns reading text for Japanese input.

Removes entry from default user dictionary of currently active core object.

Removes word from specific user dictionary.

Resets tone mark.

Selects given phonetic spelling for current input.

Returns the index of currently selected phonetic spelling.

Turns auto substitution feature on or off. Auto substitution feature replaces predefined strings with other strings. For example, if user types xmas, it could be auto substituted with Christmas.

Sets length of Chinese candidate page.

Sets text case.

Fills text buffer with given word, refreshes current input sequence and asks current core object to update candidate list accordingly.

Replaces key map for single key.

Activates requested input mode for active language.

Sets keyboard type for non-virtual keyboard. Keyboard type specifies which set of key mapping data is used.

Sets the maximum length for auto completed words. This method can be used when the client needs to be sure that all provided word completions will fit into remaining editor buffer. When the number of key presses in an input sequence exceeds the value given in aMaxLength, core will automatically reset this value to "no limit".

Sets a boolean value indicating whether number candidates are included to predictive candidate list. Number candidates are strings containing only digits. Number candidate feature must be supported by active prediction engine, otherwise this setting will be ignored.

Sets observer. See PtiObserver.h for observer API details.

Selects Chinese character (meaning that user has accepted character to be inserted into editor). Predictive candidate lists will be updated with Chinese characters associated to selected character. Associated charcaters can be accessed via ...CandidatePage() -methods. Return value can be ignored in current implementation.

Turns reordering feature on or off. This method can be used only if active core object supports reordering feature. It is also possible that core object supports reordering feature, but it can't be turned off. Reordeing feature keeps track of usage frequency for entered words and promotes most frequently used words in the candidate list. Details depend on underlying prediction engine.

Sets tone mark directly. This method is used if client wants to override default core dependant tone mark set or traditional cycle-through tone mark system doesn't suit its porposes.

Returns unicode value for current tone mark.

Creates secondary data file for given core object. Existing file will be overwritten. This data file may contain any additional data that the core object needs to store between sessions (for example used word dictionary, if the engine keeps reordering data in separate memory area).