"""Modular's mojo-syntax skill: its table of removed syntax and replacements. The skill is Modular's correction layer for code written from old tutorials, which is exactly what students paste. Its "Removed syntax" table becomes the candidate list for Stage 2's quick fixes; each one is checked against the compiler's own fix-its (messages/) and against the pinned release notes (release/) before Programming+ offers it. The skill targets "latest" Mojo, which can be ahead of the pin; the dialect markers in mojo/pin.json and verify are what decide, not this table. Writes skill/removed-syntax.json. """ from __future__ import annotations import re from ..common import PipelineError, write_json from . import Context, fresh_dir NAME = "skill" SOURCE = "skills" INPUTS = ["mojo-syntax/SKILL.md"] def table_after(md: str, heading: str) -> list[list[str]]: m = re.search(rf"(?m)^##\s+{re.escape(heading)}\s*$", md) if not m: raise PipelineError(f"skill: no '## {heading}' section") rows = [] for line in md[m.end():].split("\n"): if line.startswith("## "): break if not line.startswith("|"): if rows: break continue cells = [c.strip() for c in line.strip().strip("|").split("|")] if all(re.fullmatch(r":?-+:?", c) for c in cells): continue rows.append(cells) return rows def code_spans(cell: str) -> list[str]: return re.findall(r"`([^`]+)`", cell) def run(ctx: Context) -> dict: out = fresh_dir(NAME) md = (ctx.root / "mojo-syntax" / "SKILL.md").read_text(encoding="utf-8") rows = table_after(md, "Removed syntax — DO NOT generate these") header, body = rows[0], rows[1:] entries = [{"removed": r[0], "replacement": r[1], "removedCode": code_spans(r[0]), "replacementCode": code_spans(r[1])} for r in body if len(r) >= 2] write_json(out / "removed-syntax.json", {"processor": NAME, "provenance": ctx.prov("mojo-syntax/SKILL.md"), "columns": header, "entries": entries}) return {"entries": len(entries)}