infospace-bench/src/infospace_bench/markdown_adapter.py

from __future__ import annotations

from dataclasses import dataclass, field
from pathlib import Path
from typing import Any

from markitect_tool import Heading, Section, parse_markdown_file
from markitect_tool.contract import check_markdown_file

from .errors import InfospaceError
from .lifecycle import load_infospace


@dataclass(frozen=True)
class MarkdownDiagnostic:
    severity: str
    code: str
    message: str
    source: dict[str, Any] | None = None
    contract: dict[str, Any] | None = None
    rule_id: str | None = None
    guidance: str | None = None
    details: dict[str, Any] = field(default_factory=dict)

    @classmethod
    def from_markitect(cls, diagnostic: Any) -> "MarkdownDiagnostic":
        data = diagnostic.to_dict()
        return cls(
            severity=str(data.get("severity") or ""),
            code=str(data.get("code") or ""),
            message=str(data.get("message") or ""),
            source=data.get("source"),
            contract=data.get("contract"),
            rule_id=data.get("rule_id"),
            guidance=data.get("guidance"),
            details=dict(data.get("details") or {}),
        )

    def to_dict(self) -> dict[str, Any]:
        data: dict[str, Any] = {
            "severity": self.severity,
            "code": self.code,
            "message": self.message,
        }
        if self.source:
            data["source"] = self.source
        if self.contract:
            data["contract"] = self.contract
        if self.rule_id:
            data["rule_id"] = self.rule_id
        if self.guidance:
            data["guidance"] = self.guidance
        if self.details:
            data["details"] = self.details
        return data


@dataclass(frozen=True)
class ParsedMarkdownArtifact:
    path: Path
    frontmatter: dict[str, Any]
    headings: list[Heading]
    sections: list[Section]

    def to_dict(self) -> dict[str, Any]:
        return {
            "path": str(self.path),
            "frontmatter": self.frontmatter,
            "headings": [heading.to_dict() for heading in self.headings],
            "sections": [section.to_dict() for section in self.sections],
        }


@dataclass(frozen=True)
class ArtifactValidationResult:
    artifact_id: str
    path: str
    contract_path: str
    valid: bool
    diagnostics: list[MarkdownDiagnostic] = field(default_factory=list)
    metrics: dict[str, Any] = field(default_factory=dict)

    def to_dict(self) -> dict[str, Any]:
        return {
            "artifact_id": self.artifact_id,
            "path": self.path,
            "contract_path": self.contract_path,
            "valid": self.valid,
            "diagnostics": [diagnostic.to_dict() for diagnostic in self.diagnostics],
            "metrics": self.metrics,
        }


def parse_markdown_artifact(path: str | Path) -> ParsedMarkdownArtifact:
    artifact_path = Path(path)
    document = parse_markdown_file(artifact_path)
    return ParsedMarkdownArtifact(
        path=artifact_path,
        frontmatter=document.frontmatter,
        headings=document.headings,
        sections=document.sections,
    )


def extract_section_text(
    parsed: ParsedMarkdownArtifact,
    heading: str,
) -> str:
    expected = _normalize_heading(heading)
    for section in parsed.sections:
        if _normalize_heading(section.heading.text) == expected:
            return "\n".join(block.text for block in section.blocks if block.text).strip()
    return ""


def validate_markdown_artifact(
    artifact_id: str,
    path: str | Path,
    contract_path: str | Path,
) -> ArtifactValidationResult:
    artifact_path = Path(path)
    contract = Path(contract_path)
    result = check_markdown_file(artifact_path, contract)
    return ArtifactValidationResult(
        artifact_id=artifact_id,
        path=str(artifact_path),
        contract_path=str(contract),
        valid=result.valid,
        diagnostics=[
            MarkdownDiagnostic.from_markitect(diagnostic)
            for diagnostic in result.diagnostics
        ],
        metrics=result.metrics,
    )


def validate_infospace_artifacts(root: str | Path) -> list[ArtifactValidationResult]:
    infospace = load_infospace(root)
    results: list[ArtifactValidationResult] = []
    for artifact in infospace.artifacts:
        contract_ref = (
            infospace.config.schemas.get(artifact.kind)
            or infospace.config.schemas.get("artifact")
        )
        if not contract_ref:
            continue
        artifact_path = infospace.root / artifact.path
        contract_path = infospace.root / contract_ref
        if not contract_path.is_file():
            raise InfospaceError(
                "missing_contract",
                f"Configured contract does not exist: {contract_path}",
                {
                    "artifact_id": artifact.id,
                    "contract_path": str(contract_path),
                },
            )
        results.append(
            validate_markdown_artifact(
                artifact.id,
                artifact_path,
                contract_path,
            )
        )
    return results


def _normalize_heading(value: str) -> str:
    return " ".join(value.strip().lower().split())