I think to do this you can go two ways:

Render the text as an overlay, so that you can use model space (which then is screen space), and you can just use the text's bounding box.

Or, you can calculate the internal bounding box, although it uses some private vars:

from pygfx.objects._text import Rect

text_ob = text_bottom_right
text_blocks = text_ob._text_blocks
positions = text_ob.geometry.positions.data

total_rect = None
for i, text_block in enumerate(text_blocks):
    rect = Rect(text_block._rect.left, text_block._rect.right, text_block._rect.top, text_block._rect.bottom)
    dx, dy = positions[i,:2]
    rect.shift(dx, dy)
    if total_rect is None:
        total_rect = rect
    else:
        total_rect.left = min(total_rect.left, rect.left)
        total_rect.right = max(total_rect.right, rect.right)
        total_rect.top = max(total_rect.top, rect.top)
        total_rect.bottom = min(total_rect.bottom, rect.bottom)

print(total_rect)

From that, you should be able to use the camera's transform and projection matrix to map to world space.

Uh oh!

Map Text boundaries to world bounding box if Text is in screen space #967

Description

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

Map `Text` boundaries to world bounding box if `Text` is in screen space #967