Character count of text (post-extraction, post-trim).
OptionalfailureBounded failure evidence, present only for status: "failed".
True when the page has no text; inspect status for the reason.
1-indexed page number.
Explicit extraction outcome. failed is never treated as a blank scan.
Extracted text content (may be empty for image-only pages).
One page of extracted PDF text.
is_emptytrue indicates either a blank page or — more commonly — an image-only page that needs OCR. Drives thehas_textaggregate on the envelope.