pub fn fulltext_allowlist() -> Vec<SourceAllowlist>Expand description
Always-compiled allowlist for the full-text extraction call path (ADR-0032).
Registers ar5iv.labs.arxiv.org under a dedicated "ar5iv" source key
so crate::paper_text::paper_text (GET /html/<arxiv-id>) can reach
the ar5iv LaTeXML-XHTML renderer in the default oa-only binary —
the same always-on posture as discovery_allowlist.
The host is an arXiv subdomain (*.arxiv.org already matches it under
the tier_1_allowlist "arxiv" key), so this adds no new
registrable domain to the network surface — it only registers the host
under a distinct source key so the provenance trail records that
extracted text came from the ar5iv HTML renderer, not the arXiv
PDF/Atom API (ADR-0032 D3). Full-text extraction is classified Tier-1
OA metadata (read-only, OA, never a PDF reinterpretation), so its
transport allowlist must exist regardless of any feature gate
(ADR-0032 D2). The CLI’s build_http_client extends the production
allowlist with this unconditionally.