Error : Pointer events are not supported for this type of part:Text Document
When I use the pointerEvents function on GetStarted, the processGestures is false, he appears a crash message and does not support the current part,
When using true, there is a confusion of recognition,
I want to pass in a collection of points to the iink SDK, and then the SDK will return the text I recognized correctly, but the api cannot recognize it correctly
I did not find your detailed explanation of the documentation of the pointerEvents function
What should I do to correctly and quickly identify a set of points?
Thank you for your question.
What is the exact crash message that you have?
The pointer events use is described in this part of the documentation:
We also have a "batch mode" sample available that is an example of how to integrate iink SDK off-screen, without any user interface. It consists in batch processing content, i.e. processing a series of pointer events corresponding to already collected ink strokes and exporting the recognition result, which seems to be what you would like to do.
You can find it on GitHub in this repository: https://github.com/MyScript/interactive-ink-additional-examples-android
Is there any detailed Api documentation for this project?
In addition, let me describe it more carefully, I only need such a scene to work:
1. I write text on the screen, collect all the point data at once, and transfer it to the SDK
2. After the SDK receives the data set, it recognizes all the points and returns the correct text language
The format of the point data is the entity class defined by you -PointerEvent
If there are other examples of this pertinence, please provide them.
I will debug the example of this link tomorrow.
I currently have a problem using batch processing to convert text.
When I need to convert a collection of points, it takes about ten seconds to generate text (json in the example).
However, a full page of paper may be twice as many points normally, which means that it will take about 0-30 seconds to recognize the time, or even longer.
But I see programs written in flutter language, using the recognition function to convert text quickly.
On the native Android platform, what adjustments should I make to get the recognition results quickly?
If I reduce the set of points uniformly, but save the points in the necessary three states, can I speed up the recognition process?
Because in some of our usage scenarios, a point set may be called for repeated recognition, and then recognized after editing.
This information is also very important!
Is there a corresponding example?
Thank you for your update.
In order to speed up the recognition, the recommandation is to use the incremental mode.
But may be in your case, it is not possible to collect and send the strokes in incremental mode?
Could you please, share with us a little more information so that we can better understand you use case?