[WIP] Slurper: SageMath documentation pages

[katjabercic]

Index SageMath reference documentation pages as Items using the Sphinx objects.inv inventory file.

WIP: Do we want SageMath functions or documentation pages?

Data access

• GET https://doc.sagemath.org/html/en/reference/objects.inv — binary, no auth. Format: 4-line plain-text header, then zlib-compressed entries (one per line): name type priority uri dispname. Parse with zlib.decompress — no extra dependency needed.
• Index only std:doc entries (2,952 of 5,339 total). The py:module entries have dispname == "-" (no human title, only dotted code paths) — skip them.
• Sample entry: algebras/cubic_hecke_algebra std:doc -1 algebras/cubic_hecke_algebra.html Cubic Hecke Algebras

Item mapping

• identifier: name field (e.g. algebras/cubic_hecke_algebra)
• url: https://doc.sagemath.org/html/en/reference/ + uri field
• name: dispname field (e.g. Cubic Hecke Algebras)
• Source tag: Sg

Tasks

• web/slurper/source_sagemath.py with a slurper class + module-level instance.
• New Item.Source member (SAGEMATH = "Sg"); add it to Item.Source.key().
• management/commands/import_sagemath.py and clear_sagemath.py.
• Makefile target import-sagemath.
• make create-migrations for the new Source choice.
• make fix-files clean; tests pass.

Notes

• No Wikidata property for SageMath → index-only, links via NAME_EQ only.
• Docs are CC-BY-SA-4.0; credit SageMath with a link in the slurper docstring.
• Consider filtering out */index pages (section indexes like "Algebras") if
  they create noise in the concept graph.