document_analyzer/tests/test_detect_conflicts.py

"""Unit tests for pure functions in detect_conflicts.py — no LLM calls, no file I/O."""

import json
import sys
from pathlib import Path

import pytest

# Add detect_conflicts script to path
sys.path.insert(0, str(Path(__file__).parent.parent / "skills" / "conflict_detection_skill" / "scripts"))
import detect_conflicts as dc


# ── _is_nested_tree ──────────────────────────────────────────────────────────

def test_is_nested_tree_with_children_list():
    assert dc._is_nested_tree({"children": []}) is True


def test_is_nested_tree_without_children_list():
    assert dc._is_nested_tree({"nodes": []}) is False


def test_is_nested_tree_empty_dict():
    assert dc._is_nested_tree({}) is False


# ── _nested_tree_to_text ─────────────────────────────────────────────────────

def test_nested_tree_simple():
    tree = {
        "id": "N1",
        "name": "开始",
        "type": "start",
        "children": [
            {
                "id": "N2",
                "name": "判断条件",
                "type": "decision",
                "children": [
                    {
                        "condition": "是",
                        "node": {
                            "id": "N3",
                            "name": "执行动作",
                            "type": "action",
                        },
                    },
                    {
                        "condition": "否",
                        "node": {
                            "id": "N4",
                            "name": "结束",
                            "type": "end",
                        },
                    },
                ],
            },
        ],
    }
    result = dc._nested_tree_to_text(tree)
    assert "[起始] N1: 开始" in result
    assert "[判断] N2: 判断条件" in result
    assert '分支 "是":' in result
    assert "[动作] N3: 执行动作" in result
    assert '分支 "否":' in result
    assert "[结束] N4: 结束" in result


def test_nested_tree_empty_children():
    tree = {"id": "N1", "name": "结束", "type": "end"}
    result = dc._nested_tree_to_text(tree)
    assert "[结束] N1: 结束" in result


# ── _flat_tree_to_text ───────────────────────────────────────────────────────

def test_flat_tree_with_decision_and_action():
    lt = {
        "root": "START",
        "nodes": [
            {"id": "D1", "type": "decision", "condition": "车速>15",
             "branches": [{"value": "是", "target": "A1"}, {"value": "否", "target": "END"}]},
            {"id": "A1", "type": "action", "description": "禁止视频播放"},
            {"id": "END", "type": "end", "description": "结束"},
        ],
    }
    result = dc._flat_tree_to_text(lt)
    assert "根节点: START" in result
    assert '判断节点 D1: 条件="车速>15"' in result
    assert '分支 "是" → A1' in result
    assert "动作节点 A1: 禁止视频播放" in result


def test_flat_tree_empty():
    result = dc._flat_tree_to_text({})
    assert result == ""


# ── _logic_tree_to_text (dispatcher) ─────────────────────────────────────────

def test_logic_tree_to_text_nested():
    lt = {"children": [{"id": "N1", "name": "开始", "type": "start"}]}
    result = dc._logic_tree_to_text(lt)
    assert "[起始] N1: 开始" in result


def test_logic_tree_to_text_flat():
    lt = {"nodes": [{"id": "A1", "type": "action", "description": "测试"}]}
    result = dc._logic_tree_to_text(lt)
    assert "动作节点 A1: 测试" in result


# ── _parse_conflict_json ─────────────────────────────────────────────────────

def test_parse_no_conflict():
    assert dc._parse_conflict_json("[[NO_CONFLICT]]") == []
    assert dc._parse_conflict_json("some text [[NO_CONFLICT]] more") == []


def test_parse_conflict_valid_json():
    content = json.dumps([
        {"conflict_type": "condition_mismatch", "severity": "high",
         "section": "3.1", "image_snippet": "img", "text_snippet": "txt",
         "description": "冲突描述"}
    ], ensure_ascii=False)
    result = dc._parse_conflict_json(content)
    assert len(result) == 1
    assert result[0]["conflict_type"] == "condition_mismatch"


def test_parse_conflict_with_markdown_fence():
    content = '```json\n[{"conflict_type": "contradiction", "severity": "high", "section": "1.0", "image_snippet": "a", "text_snippet": "b", "description": "x"}]\n```'
    result = dc._parse_conflict_json(content)
    assert len(result) == 1
    assert result[0]["conflict_type"] == "contradiction"


def test_parse_conflict_unparseable():
    assert dc._parse_conflict_json("not valid json at all") == []
    assert dc._parse_conflict_json("") == []


def test_parse_conflict_mixed_content():
    content = 'some prefix text\n[\n  {"conflict_type": "scope_mismatch", "severity": "low", "section": "4.2", "image_snippet": "img", "text_snippet": "txt", "description": "d"}]\nsuffix'
    result = dc._parse_conflict_json(content)
    assert len(result) == 1
    assert result[0]["conflict_type"] == "scope_mismatch"


# ── _build_text_for_section ──────────────────────────────────────────────────

def test_build_text_para_blocks():
    sections = [
        {"source": "3.1 测试章节",
         "blocks": [
             {"type": "para", "text": "第一段文字"},
             {"type": "para", "text": "第二段文字"},
         ]},
    ]
    result = dc._build_text_for_section(sections, "3.1 测试章节")
    assert "第一段文字" in result
    assert "第二段文字" in result


def test_build_text_table_blocks():
    sections = [
        {"source": "4.0 功能列表",
         "blocks": [
             {"type": "table", "table": "功能表",
              "rows": [
                  {"columns": [{"name": "功能", "text": "视频播放"}, {"name": "状态", "text": "禁止"}]},
              ]},
         ]},
    ]
    result = dc._build_text_for_section(sections, "4.0 功能列表")
    assert "功能表" in result
    assert "视频播放" in result
    assert "禁止" in result


def test_build_text_section_not_found():
    sections = [{"source": "1.0 概述", "blocks": []}]
    result = dc._build_text_for_section(sections, "2.0 不存在的章节")
    assert result == ""


def test_build_text_mixed_blocks():
    sections = [
        {"source": "5.1 混合章节",
         "blocks": [
             {"type": "para", "text": "介绍文字"},
             {"type": "table", "table": "表1",
              "rows": [{"columns": [{"name": "A", "text": "值1"}]}]},
         ]},
    ]
    result = dc._build_text_for_section(sections, "5.1 混合章节")
    assert "介绍文字" in result
    assert "表1" in result
    assert "值1" in result