"""What may leave the Space: identifiers only (ADR-0001 §6). Refusals are data."""

from __future__ import annotations

import pytest

from src.identifiers import MAX_IDENTIFIERS, parse_disease, parse_gene_list, parse_nct

# Real HGNC symbols chosen to exercise the edges of the pattern and of the sample-id prefixes.
VALID_SYMBOLS = [
    "KRAS",
    "TP53",
    "ELAVL1",
    "C1orf112",
    "HLA-DRB1",
    "NKX2-1",
    "IGHV1-69-2",
    "CPT1A",
    "ERRFI1",
    "DOK1",
    "SAA1",
    "SORD2P",
]


@pytest.mark.parametrize("symbol", VALID_SYMBOLS)
def test_real_symbols_pass(symbol):
    parsed = parse_gene_list(symbol)
    assert parsed.ok, parsed.message
    assert parsed.identifiers[0].kind == "symbol"


def test_a_list_splits_on_commas_spaces_and_semicolons_and_keeps_order():
    parsed = parse_gene_list("JAK1, JAK2;STAT1  jak1")
    assert parsed.ok
    assert [i.value for i in parsed.identifiers] == ["JAK1", "JAK2", "STAT1"]
    assert parsed.duplicates_dropped == 1


def test_ensembl_ids_are_genes_and_uniprot_only_when_allowed():
    assert parse_gene_list("ENSG00000133703").identifiers[0].kind == "ensembl"
    assert parse_gene_list("P05161", allow_uniprot=True).identifiers[0].kind == "uniprot"
    assert parse_gene_list("P05161").identifiers[0].kind == "symbol"


@pytest.mark.parametrize(
    "raw",
    [
        "what drugs target KRAS?",
        "TCGA-2J-AAB1-01A",
        "GSM1234567",
        "C3L-00017",
        "12.5",
        "KRAS 3.2 1.1",
        "<script>",
        "SELECT * FROM genes",
    ],
)
def test_free_text_sample_ids_and_numbers_are_refused(raw):
    parsed = parse_gene_list(raw)
    assert not parsed.ok
    assert parsed.reason == "invalid_identifier"
    assert "free text, sample ids and numbers are not" in parsed.message


def test_an_empty_list_is_refused():
    assert parse_gene_list("  ").reason == "invalid_identifier"
    assert parse_gene_list([]).reason == "invalid_identifier"


def test_more_than_the_cap_is_refused_with_the_cap():
    raw = ",".join(f"GENE{i}" for i in range(MAX_IDENTIFIERS + 1))
    parsed = parse_gene_list(raw)
    assert parsed.reason == "too_many_identifiers"
    assert str(MAX_IDENTIFIERS) in parsed.message
    assert parse_gene_list(",".join(f"GENE{i}" for i in range(MAX_IDENTIFIERS))).ok


def test_a_refusal_never_echoes_a_long_token():
    parsed = parse_gene_list("x" * 200 + "!")
    assert len(parsed.message) < 250
    assert "x" * 30 not in parsed.message


@pytest.mark.parametrize(
    "raw,kind",
    [
        ("EFO_0002618", "disease_id"),
        ("MONDO_0005192", "disease_id"),
        ("pancreatic ductal adenocarcinoma", "disease_name"),
        ("Crohn's disease", "disease_name"),
        ("type 2 diabetes mellitus", "disease_name"),
    ],
)
def test_diseases(raw, kind):
    parsed = parse_disease(raw)
    assert parsed.ok and parsed.identifiers[0].kind == kind


@pytest.mark.parametrize(
    "raw", ["", "a" * 81, "patient 12; DOB 1950-01-01?", "1234", "TCGA-2J-AAB1"]
)
def test_bad_diseases_are_refused(raw):
    assert parse_disease(raw).reason == "invalid_identifier"


def test_nct_ids():
    assert parse_nct("nct04635059").identifiers[0].value == "NCT04635059"
    assert parse_nct("NCT123").reason == "invalid_identifier"