"""Extract a loss-inclusive, selected-PDF table directly from scores.json."""
import json
from pathlib import Path

LAB = Path(__file__).resolve().parent
SAMPLES = {"prose-single", "table-simple", "table-merged-header", "two-column",
           "scan-image", "mixed", "equation", "rotated-header", "columns-with-table",
           "wide-columns", "wide-columns-with-table"}
TOOLS = ["pdfplumber", "pdfplumber-columns", "markitdown", "pymupdf4llm", "markovo"]
KEYS = ["prose_recall", "reading_order", "table_rows_intact", "table_header_intact",
        "table_pipe_detected", "scan_text_recall", "pages_covered", "equation_recall", "elapsed_s"]
scores = json.loads((LAB / "scores.json").read_text())
cells = [c for c in scores["cells"] if c["tool"] in TOOLS and c["sample"] in SAMPLES]
cells.sort(key=lambda c: (TOOLS.index(c["tool"]), c["sample"]))


def display(value):
    if value is None:
        return "n/a"
    if isinstance(value, bool):
        return str(value).lower()
    return str(value)


lines = ["| Path | Fixture | P | O | R | H | Pipe | S | Pages | Math | Seconds |",
         "| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |"]
for cell in cells:
    lines.append("| " + " | ".join([cell["tool"], cell["sample"],
                                    *[display(cell.get(k)) for k in KEYS]]) + " |")
(LAB / "selected-scores.md").write_text("\n".join(lines) + "\n", encoding="utf-8")
print(f"Selected {len(cells)} cells; errors: {sum(bool(c.get('error')) for c in cells)}")