Recognition: Advanced Topics

About Advanced Topics in Recognition

The user can enter unrecognized ink by enabling ink text or sketch ink. In this
mode, strokes appear as ink. To convert the ink to text, the user double-taps the ink
word; the user can cause multiple words to be recognized by selecting them
beforehand and then double-tapping the selection. The recognition system responds
by inverting the ink word or selection, as shown in Figure 10-2, and returning the
recognized text, which replaces the selection.

User interface to deferred recognition, with inverted ink

Programmer's Overview of Deferred Recognition

10

Deferred recognition is available in views based on the

clEditView

class or

clParagraphView

views that support ink text. This feature works with any

amount of input, from a single letter to a full page of text.

To initiate deferred recognition, the user double-taps the child views that display
the ink to be recognized. The recognized text is added to an edit view as if the user
had just written it. That is, a new

clParagraphView

child is added, or the

recognized text is appended to a nearby

clParagraphView

. After the recognized

text has been added, the original view containing the sketch ink or the ink text is
removed from its edit view parent.

Deferred recognition also invokes the

ViewAddChildScript

and

ViewDropChildScript

methods of the recognized text and unrecognized ink

views. Words added to nearby paragraphs invoke

ViewChangedScript

methods for those paragraphs, updating the

text

slot in those views; for some

paragraph views, the

viewBounds

slot is updated as well.

You can pass

recConfig

frames to the global functions

Recognize

,

RecognizePara

, and

RecognizePoly

to implement your own form of

deferred recognition. For more information, see "Deferred Recognition Functions"
(page 8-89) in Newton Programmer's Reference.