>jTSrSSKJrJr SSKJrJr SSKJrJ r SSK J r J r J r /SQrg)uFuse docling's spine with the Set-of-Mark/Camelot tables. docling and SoM/Camelot extract the same page and are each stronger at different things, so this package fuses them — each enriches the other. SoM/Camelot owns the vetted, text-grounded table bodies; docling owns the spine (reading order, headings, text), an independent table count, picture classification, and the OCR signal for regions read from the page image. `match_tables` pairs the two extractions by bounding-box overlap and classifies each region (see `models.RegionMatch`). The graft and annotation passes act on those records to produce the two corrected outputs: the SoM/Camelot tables and a unified `DoclingDocument` whose table bodies come from Camelot. `fuse_artifacts` is the standalone fusion step (already-produced data in, both outputs out); `DocumentFusion` is the local control flow that runs the document and table extractions in-process and then fuses them. )DocumentFusionfuse_artifacts)build_unified_documentmarkdown_to_table_data)MATCH_FRACTION match_tables) FusionResult MatchKind RegionMatch) rrr r r rrrrN)__doc__quber.core.fusion.fusionrrquber.core.fusion.graftrrquber.core.fusion.matchingrrquber.core.fusion.modelsr r r __all__8/home/mande/repo/quber/src/quber/core/fusion/__init__.pyrs#$DRCII r