ÞĐƑ (ÞĐƑîüḿ) ƒöŕḿàţ (.þđƒ)

Text + geometry extraction via PDFium.

ÎĐpdf

ŠöüŕçéPlugin

Éẋţéñšîöñš.pdf

ḾÎḾÉ Ţýþéšapplication/pdf

ÇàþàƃîļîţîéšRead

Þàŕàḿéţéŕš

Þàŕàḿéţéŕ	Ţýþé	Đéƒàüļţ	Đéšçŕîþţîöñ
geometry	boolean	false	Emit one block per positioned text rectangle, each carrying a bounding box (for the visual layout view), instead of one plain-text block per page.
glyphs	boolean	false	Additionally attach per-character bounding boxes to each text block (implies geometry). Enables character-precise highlighting; slightly slower.
tier3	boolean	false	Render each page and run the kapi-vision layout model over it for authoritative structure (multi-column reading order, borderless tables, figure/caption regions) beyond what page geometry alone recovers. Requires the kapi-vision plugin; falls back to geometric structure (tier 2) when it is not installed. Implies geometry.

Çöñƒîĝüŕé îţ ļîṽé

Çöñƒîĝüŕàţîöñ

Ļöàđîñĝ ƒöŕḿ…

Çöñƒîĝüŕàţîöñ öüţþüţ

ṽš defaults ñö çĥàñĝéš

apiVersion: neokapi/v1
kind: PdfFormatConfig
metadata:
  name: pdf
spec:
  # (default configuration)

← Ƃàçķ ţö ţĥé Ƒöŕḿàţ Ŕéƒéŕéñçé