Tecack

Appearance

Stroke Data ​

Common Data Structure ​

In Tecack, we handle stroke data called "TecackStroke".

A TecackStroke represents the coordinate information of a single stroke, and any character is represented by an array of these strokes.

Specifically, it adopts the following structure:

go
// Note: a number is a 64-bit floating-point number

TecackStrokes     :=   [ { TecackStroke } ... ]
TecackStroke      :=   [ { Position } ... ]
Position          :=   X Y
X                 :=   number
Y                 :=   number
// Note: a number is a 64-bit floating-point number

TecackStrokes     :=   [ { TecackStroke } ... ]
TecackStroke      :=   [ { Position } ... ]
Position          :=   X Y
X                 :=   number
Y                 :=   number
TypeScript
ts
type TecackStrokes = Array<TecackStroke>;
type TecackStroke = Array<Position>;
type Position = [X, Y];
type X = number;
type Y = number;
type TecackStrokes = Array<TecackStroke>;
type TecackStroke = Array<Position>;
type Position = [X, Y];
type X = number;
type Y = number;

Data inputted through the canvas, and datasets carry this TecackStroke information.

The recognition algorithms in the backend also handle them.

Datasets ​

Datasets carry the following information:

go
// Note: char is a character represented by a single UTF-16 code unit

TecackDataset           :=   TargetChar TargetCharStrokeLength TecackStrokes
TargetChar              :=   char
TargetCharStrokeLength  :=   number
// Note: char is a character represented by a single UTF-16 code unit

TecackDataset           :=   TargetChar TargetCharStrokeLength TecackStrokes
TargetChar              :=   char
TargetCharStrokeLength  :=   number
TypeScript
ts
type TecackDataset = [TargetChar, TargetCharStrokeLength, Array<TecackStroke>];
type TargetChar = string;
type TargetCharStrokeLength = number;
type TecackDataset = [TargetChar, TargetCharStrokeLength, Array<TecackStroke>];
type TargetChar = string;
type TargetCharStrokeLength = number;

Creating Datasets ​

To create datasets, primarily two tools are used:

jTegaki ​

jTegaki allows you to create stroke data via GUI and outputs a language-independent TecackDataset. The format is XML.
For detailed usage, please refer to Tools > jTegaki.

codegen ​

codegen generates dataset codes based on XML.
Currently, only TypeScript is supported.
For detailed usage, please refer to Tools > codegen.