from __future__ import annotations

import copy
import io
import zipfile

import pytest

from pallet.merge_appearance_classes import (EXPECTED_A_CLASSES, column_name, column_number,
    group_statistics, merge_target, parse_grouping, read_workbook, write_workbook)


def matrix():
    result={'B1':'Class 1','C1':'Class 2'}
    for i in range(44):result[f'A{i+2}']=f'A{i:02d}';result[f'{"B" if i%2==0 else "C"}{i+2}']=1
    return result


def test_every_a_class_has_one_group_and_order_is_stable():
    cells=matrix();before=copy.deepcopy(cells)
    groups,mapping=parse_grouping(cells)
    assert set(mapping)==EXPECTED_A_CLASSES and len(groups)==2
    assert groups[0]['original_appearance_classes']==[f'A{i:02d}' for i in range(0,44,2)]
    assert mapping['A43']['merged_class_id']==2 and mapping['A43']['indicator_cell']=='C45'
    assert cells==before


@pytest.mark.parametrize('change,error',[
    ({'C2':1},'exactly one'),({'B2':None},'exactly one'),({'A2':'A99'},'Unknown'),
    ({'A3':'A00'},'duplicate A'),({'B2':2},'Invalid grouping indicator'),
    ({'B2':'yes'},'Invalid grouping indicator'),({'D2':1},'Invalid grouping indicator'),
    ({'C1':'Class 1'},'duplicate new'),({'B1':'abc'},'Invalid class header'),
    ({'A2':None},'has no A class')])
def test_incomplete_ambiguous_or_illegal_spreadsheet_rejected(change,error):
    with pytest.raises(ValueError,match=error):parse_grouping(matrix()|change)


def test_unlisted_a_class_rejected():
    cells=matrix();cells.pop('A45');cells.pop('C45')
    with pytest.raises(ValueError,match='Uncovered'):parse_grouping(cells)


def test_zero_and_blank_indicators_are_not_assignments():
    cells=matrix()|{'C2':0,'B3':''}
    assert len(parse_grouping(cells)[1])==44


def test_targets_keep_original_clusters_scores_boxes_and_noise():
    _,mapping=parse_grouping(matrix())
    raw={'target_id':'frame::box:0000','appearance_cluster':2,'appearance_membership':.7,
         'geometry_cluster':9,'bbox':{'x':1.5},'split':'val'}
    before=copy.deepcopy(raw);result=merge_target(raw,mapping)
    assert result['original_appearance_class']=='A02' and result['merged_class_id']==1
    assert result['merged_class']=='Class 1' and raw==before
    assert all(result[k]==v for k,v in raw.items())
    noise=merge_target(raw|{'appearance_cluster':-1},mapping)
    assert noise['merged_class_id'] is None and noise['merged_class']=='unassigned'
    assert noise['class_merge_status']=='unassigned_appearance_noise'
    failed=merge_target(raw|{'appearance_cluster':-2},mapping)
    assert failed['merged_class_id'] is None and failed['class_merge_status']=='unassigned_processing_failure'
    with pytest.raises(ValueError,match='No mapping'):merge_target(raw|{'appearance_cluster':44},mapping)


def test_statistics_deduplicate_frames_and_groups_within_each_new_class():
    headers,mapping=parse_grouping(matrix())
    base={'frame_id':'f1','group_id':'g1','domain':'d1','split':'train'}
    rows=[merge_target(base|{'appearance_cluster':i},mapping) for i in (0,2,1)]
    rows.append(merge_target(base|{'appearance_cluster':-1},mapping))
    seed=[{'merged_class_id':1,'original_split':'val'},{'merged_class_id':1,'original_split':'train'}]
    groups=group_statistics(headers,rows,seed)
    assert groups[0]['targets']==2 and groups[0]['frames']==1 and groups[0]['acquisition_groups']==1
    assert groups[0]['splits']=={'train':2,'val':0,'test':0}
    assert groups[0]['seed_splits']=={'train':1,'val':1,'test':0} and groups[0]['seed_targets']==2
    assert groups[1]['frames']==1 and sum(g['targets'] for g in groups)==3


def test_excel_export_has_valid_xml_and_exact_rule_matrix(tmp_path):
    original=matrix();rows=[['原A类','Class 1','Class 2']]
    rows += [[original[f'A{i}'],original.get(f'B{i}'),original.get(f'C{i}')] for i in range(2,46)]
    data=write_workbook([('合并规则矩阵',rows),('中文统计',[['类别','目标数'],['Class 1',123]])])
    assert data==write_workbook([('合并规则矩阵',rows),('中文统计',[['类别','目标数'],['Class 1',123]])])
    file=tmp_path/'类别合并结果.xlsx';file.write_bytes(data)
    name,cells=read_workbook(file,'合并规则矩阵')
    assert name=='合并规则矩阵' and parse_grouping(cells)[1]==parse_grouping(original)[1]
    assert read_workbook(file,'中文统计')[1]['B2']==123
    with pytest.raises(ValueError,match='Multiple worksheets'):read_workbook(file)
    with pytest.raises(ValueError,match='not found'):read_workbook(file,'missing')
    import xml.etree.ElementTree as ET
    with zipfile.ZipFile(io.BytesIO(data)) as archive:
        assert archive.testzip() is None
        for name in archive.namelist():ET.fromstring(archive.read(name))


def test_formula_cells_rejected_even_if_cached_numeric_value_exists(tmp_path):
    data=write_workbook([('rules',[['A','Class 1'],['A00',1]])]);output=io.BytesIO()
    with zipfile.ZipFile(io.BytesIO(data)) as source,zipfile.ZipFile(output,'w') as target:
        for name in source.namelist():
            payload=source.read(name)
            if name=='xl/worksheets/sheet1.xml':payload=payload.replace(b'<v>1</v>',b'<f>SUM(1)</f><v>1</v>')
            target.writestr(name,payload)
    path=tmp_path/'formula.xlsx';path.write_bytes(output.getvalue())
    with pytest.raises(ValueError,match='Formula'):read_workbook(path)


@pytest.mark.parametrize('number',[1,26,27,52,53,703])
def test_excel_column_addresses(number):
    assert column_number(column_name(number))==number
