lingyue-zhibao/test/test_upload_api.sh

#!/bin/bash

# ============================================
# 文件上传端到端测试脚本
# ============================================
# 测试流程: 上传 -> 解析等待 -> 向量提取 -> NER提取
# 使用方法: ./test_upload_api.sh [host] [port]
# 示例: ./test_upload_api.sh localhost 5232
# ============================================

# 配置参数
HOST=${1:-localhost}
PORT=${2:-5232}
BASE_URL="http://${HOST}:${PORT}"
UPLOAD_URL="${BASE_URL}/api/v1/parse/upload"
STATUS_URL="${BASE_URL}/parse/status"
REGISTER_URL="${BASE_URL}/auth/register"
TEXT_STORAGE_URL="${BASE_URL}/api/v1/graph/text-storage"
RAG_INDEX_URL="${BASE_URL}/api/rag/index"
NER_DOCUMENT_URL="${BASE_URL}/api/ner/document"

# 测试文件路径（相对于脚本所在目录）
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
TEST_FILE="${SCRIPT_DIR}/test.docx"

# 测试用户信息
TIMESTAMP=$(date +%s)
TEST_USERNAME="testuser_${TIMESTAMP}"
TEST_EMAIL="testuser_${TIMESTAMP}@test.com"
TEST_PASSWORD="Test123456!"
USER_ID=""
DOCUMENT_ID=""
DOCUMENT_TEXT=""

# 颜色定义
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
BLUE='\033[0;34m'
CYAN='\033[0;36m'
NC='\033[0m' # No Color

# 输出函数
print_header() {
    echo -e "\n${BLUE}============================================${NC}"
    echo -e "${BLUE}$1${NC}"
    echo -e "${BLUE}============================================${NC}"
}

print_step() {
    echo -e "\n${CYAN}>>> $1${NC}"
}

print_success() {
    echo -e "${GREEN}✓ $1${NC}"
}

print_error() {
    echo -e "${RED}✗ $1${NC}"
}

print_info() {
    echo -e "${YELLOW}➤ $1${NC}"
}

# 检查依赖
check_dependencies() {
    print_header "检查依赖"
    
    if ! command -v curl &> /dev/null; then
        print_error "curl 未安装"
        exit 1
    fi
    print_success "curl 已安装"
    
    if ! command -v jq &> /dev/null; then
        print_info "jq 未安装，JSON格式化将不可用"
        JQ_AVAILABLE=false
    else
        print_success "jq 已安装"
        JQ_AVAILABLE=true
    fi
}

# 检查测试文件
check_test_file() {
    print_header "检查测试文件"
    
    if [ ! -f "$TEST_FILE" ]; then
        print_error "测试文件不存在: $TEST_FILE"
        exit 1
    fi
    
    FILE_SIZE=$(stat -c%s "$TEST_FILE" 2>/dev/null || stat -f%z "$TEST_FILE" 2>/dev/null)
    print_success "测试文件存在: $TEST_FILE"
    print_info "文件大小: $FILE_SIZE bytes"
}

# 检查服务是否可用
check_service() {
    print_header "检查服务状态"
    
    print_info "测试服务: $BASE_URL"
    
    HTTP_CODE=$(curl -s -o /dev/null -w "%{http_code}" --connect-timeout 5 "${BASE_URL}/actuator/health" 2>/dev/null)
    
    if [ "$HTTP_CODE" = "200" ]; then
        print_success "服务正常运行 (HTTP $HTTP_CODE)"
    elif [ "$HTTP_CODE" = "000" ]; then
        print_error "无法连接到服务 $BASE_URL"
        print_info "请确保 parse-service 正在运行"
        exit 1
    else
        print_info "健康检查返回 HTTP $HTTP_CODE，继续测试..."
    fi
}

# 注册测试用户
register_test_user() {
    print_header "注册测试用户"
    
    print_info "用户名: $TEST_USERNAME"
    print_info "邮箱: $TEST_EMAIL"
    print_info "注册URL: $REGISTER_URL"
    
    RESPONSE=$(curl -s -w "\n%{http_code}" \
        -X POST "$REGISTER_URL" \
        -H "Content-Type: application/json" \
        -d "{\"username\":\"${TEST_USERNAME}\",\"email\":\"${TEST_EMAIL}\",\"password\":\"${TEST_PASSWORD}\",\"confirmPassword\":\"${TEST_PASSWORD}\"}" \
        --connect-timeout 10)
    
    HTTP_CODE=$(echo "$RESPONSE" | tail -n1)
    BODY=$(echo "$RESPONSE" | sed '$d')
    
    echo -e "\n${YELLOW}响应状态码:${NC} $HTTP_CODE"
    
    if [ "$JQ_AVAILABLE" = true ]; then
        echo "$BODY" | jq . 2>/dev/null || echo "$BODY"
    else
        echo "$BODY"
    fi
    
    # 解析用户ID
    if [ "$HTTP_CODE" = "200" ] || [ "$HTTP_CODE" = "201" ]; then
        print_success "用户注册成功!"
        
        if [ "$JQ_AVAILABLE" = true ]; then
            USER_ID=$(echo "$BODY" | jq -r '.data.user.id // .data.userId // .userId // empty' 2>/dev/null)
            if [ -z "$USER_ID" ] || [ "$USER_ID" = "null" ]; then
                # 尝试其他可能的字段
                USER_ID=$(echo "$BODY" | jq -r '.data.id // .id // empty' 2>/dev/null)
            fi
            
            if [ -n "$USER_ID" ] && [ "$USER_ID" != "null" ]; then
                print_info "用户ID: $USER_ID"
                echo "$USER_ID" > "${SCRIPT_DIR}/.last_user_id"
            else
                print_error "无法从响应中获取用户ID"
                echo "响应内容: $BODY"
                exit 1
            fi
        fi
    else
        print_error "用户注册失败 (HTTP $HTTP_CODE)"
        print_info "响应: $BODY"
        exit 1
    fi
}

# 测试文件上传
test_upload() {
    print_step "文件上传"
    
    print_info "上传URL: $UPLOAD_URL"
    print_info "用户ID: $USER_ID"
    print_info "文件: $TEST_FILE"
    
    echo -e "\n发送请求..."
    
    RESPONSE=$(curl -s -w "\n%{http_code}" \
        -X POST "$UPLOAD_URL" \
        -H "Content-Type: multipart/form-data" \
        -F "file=@${TEST_FILE}" \
        -F "userId=${USER_ID}" \
        --connect-timeout 10 \
        --max-time 300)
    
    HTTP_CODE=$(echo "$RESPONSE" | tail -n1)
    BODY=$(echo "$RESPONSE" | sed '$d')
    
    echo -e "\n${YELLOW}响应状态码:${NC} $HTTP_CODE"
    echo -e "${YELLOW}响应内容:${NC}"
    
    if [ "$JQ_AVAILABLE" = true ]; then
        echo "$BODY" | jq . 2>/dev/null || echo "$BODY"
    else
        echo "$BODY"
    fi
    
    # 解析响应
    if [ "$HTTP_CODE" = "200" ]; then
        print_success "文件上传成功!"
        
        # 提取documentId用于后续操作
        if [ "$JQ_AVAILABLE" = true ]; then
            DOCUMENT_ID=$(echo "$BODY" | jq -r '.data.documentId // .documentId // empty' 2>/dev/null)
            if [ -n "$DOCUMENT_ID" ] && [ "$DOCUMENT_ID" != "null" ]; then
                print_info "文档ID: $DOCUMENT_ID"
                echo "$DOCUMENT_ID" > "${SCRIPT_DIR}/.last_document_id"
            else
                print_error "无法从响应中获取文档ID"
                return 1
            fi
        fi
        return 0
    else
        print_error "文件上传失败 (HTTP $HTTP_CODE)"
        return 1
    fi
}

# 测试解析状态查询（单次）
test_parse_status() {
    local DOC_ID=$1
    
    print_info "文档ID: $DOC_ID"
    print_info "状态URL: ${STATUS_URL}/${DOC_ID}"
    
    RESPONSE=$(curl -s -w "\n%{http_code}" \
        -X GET "${STATUS_URL}/${DOC_ID}" \
        --connect-timeout 10)
    
    HTTP_CODE=$(echo "$RESPONSE" | tail -n1)
    BODY=$(echo "$RESPONSE" | sed '$d')
    
    echo -e "\n${YELLOW}响应状态码:${NC} $HTTP_CODE"
    echo -e "${YELLOW}响应内容:${NC}"
    
    if [ "$JQ_AVAILABLE" = true ]; then
        echo "$BODY" | jq . 2>/dev/null || echo "$BODY"
        
        # 显示关键状态信息
        if [ "$HTTP_CODE" = "200" ]; then
            STATUS=$(echo "$BODY" | jq -r '.data.status // empty' 2>/dev/null)
            PROGRESS=$(echo "$BODY" | jq -r '.data.progress // 0' 2>/dev/null)
            CURRENT_STEP=$(echo "$BODY" | jq -r '.data.currentStep // empty' 2>/dev/null)
            print_info "状态: $STATUS, 进度: ${PROGRESS}%, 当前步骤: $CURRENT_STEP"
        fi
    else
        echo "$BODY"
    fi
    
    if [ "$HTTP_CODE" = "200" ]; then
        print_success "状态查询成功!"
    else
        print_error "状态查询失败 (HTTP $HTTP_CODE)"
    fi
}

# 轮询解析状态直到完成
poll_parse_status() {
    local DOC_ID=$1
    local MAX_ATTEMPTS=${2:-60}
    local INTERVAL=${3:-3}
    
    print_step "轮询解析状态 (最多${MAX_ATTEMPTS}次, 间隔${INTERVAL}秒)"
    
    for ((i=1; i<=MAX_ATTEMPTS; i++)); do
        RESPONSE=$(curl -s "${STATUS_URL}/${DOC_ID}" --connect-timeout 10)
        
        if [ "$JQ_AVAILABLE" = true ]; then
            # 状态字段为 status，值为: pending/processing/completed/failed
            STATUS=$(echo "$RESPONSE" | jq -r '.data.status // .status // empty' 2>/dev/null)
            PROGRESS=$(echo "$RESPONSE" | jq -r '.data.progress // .progress // 0' 2>/dev/null)
            echo -ne "\r第 $i 次查询... 状态: $STATUS, 进度: ${PROGRESS}%    "
            
            if [ "$STATUS" = "completed" ] || [ "$STATUS" = "COMPLETED" ]; then
                echo ""
                print_success "解析完成!"
                return 0
            elif [ "$STATUS" = "failed" ] || [ "$STATUS" = "FAILED" ]; then
                echo ""
                print_error "解析失败!"
                echo "$RESPONSE" | jq .
                return 1
            fi
        else
            echo "$RESPONSE"
        fi
        
        sleep $INTERVAL
    done
    
    echo ""
    print_error "轮询超时，解析未完成"
    return 1
}

# 获取解析后的文本内容
get_document_text() {
    local DOC_ID=$1
    
    print_step "获取文档解析文本"
    
    print_info "文档ID: $DOC_ID"
    print_info "请求URL: ${TEXT_STORAGE_URL}/${DOC_ID}"
    
    RESPONSE=$(curl -s -w "\n%{http_code}" \
        -X GET "${TEXT_STORAGE_URL}/${DOC_ID}" \
        --connect-timeout 10)
    
    HTTP_CODE=$(echo "$RESPONSE" | tail -n1)
    BODY=$(echo "$RESPONSE" | sed '$d')
    
    echo -e "${YELLOW}响应状态码:${NC} $HTTP_CODE"
    
    if [ "$HTTP_CODE" = "200" ]; then
        if [ "$JQ_AVAILABLE" = true ]; then
            # 获取文件路径
            FILE_PATH=$(echo "$BODY" | jq -r '.data.filePath // empty' 2>/dev/null)
            if [ -n "$FILE_PATH" ] && [ "$FILE_PATH" != "null" ]; then
                print_success "获取文本存储记录成功!"
                print_info "文件路径: $FILE_PATH"
                
                # 读取文件内容
                if [ -f "$FILE_PATH" ]; then
                    DOCUMENT_TEXT=$(cat "$FILE_PATH" 2>/dev/null)
                    TEXT_LENGTH=${#DOCUMENT_TEXT}
                    print_success "读取文本成功 (长度: $TEXT_LENGTH 字符)"
                    
                    # 显示前200个字符
                    echo -e "${YELLOW}文本预览:${NC}"
                    echo "${DOCUMENT_TEXT:0:200}..."
                    return 0
                else
                    print_error "文件不存在: $FILE_PATH"
                    return 1
                fi
            else
                print_error "响应中无文件路径"
                echo "$BODY" | jq . 2>/dev/null
                return 1
            fi
        fi
    else
        print_error "获取文本存储失败 (HTTP $HTTP_CODE)"
        echo "$BODY"
        return 1
    fi
}

# 向量提取（RAG 索引）
test_vector_extraction() {
    local DOC_ID=$1
    local TEXT=$2
    
    print_step "向量提取 (RAG 索引)"
    
    print_info "文档ID: $DOC_ID"
    print_info "文本长度: ${#TEXT} 字符"
    print_info "请求URL: $RAG_INDEX_URL"
    
    # 构建请求JSON（需要转义文本中的特殊字符）
    if [ "$JQ_AVAILABLE" = true ]; then
        REQUEST_BODY=$(jq -n \
            --arg docId "$DOC_ID" \
            --arg text "$TEXT" \
            '{documentId: $docId, text: $text}')
    else
        # 简单转义
        ESCAPED_TEXT=$(echo "$TEXT" | sed 's/\\/\\\\/g' | sed 's/"/\\"/g' | tr '\n' ' ')
        REQUEST_BODY="{\"documentId\":\"${DOC_ID}\",\"text\":\"${ESCAPED_TEXT}\"}"
    fi
    
    RESPONSE=$(curl -s -w "\n%{http_code}" \
        -X POST "$RAG_INDEX_URL" \
        -H "Content-Type: application/json" \
        -d "$REQUEST_BODY" \
        --connect-timeout 30 \
        --max-time 300)
    
    HTTP_CODE=$(echo "$RESPONSE" | tail -n1)
    BODY=$(echo "$RESPONSE" | sed '$d')
    
    echo -e "${YELLOW}响应状态码:${NC} $HTTP_CODE"
    echo -e "${YELLOW}响应内容:${NC}"
    
    if [ "$JQ_AVAILABLE" = true ]; then
        echo "$BODY" | jq . 2>/dev/null || echo "$BODY"
    else
        echo "$BODY"
    fi
    
    if [ "$HTTP_CODE" = "200" ]; then
        if [ "$JQ_AVAILABLE" = true ]; then
            CHUNK_COUNT=$(echo "$BODY" | jq -r '.data.chunkCount // empty' 2>/dev/null)
            if [ -n "$CHUNK_COUNT" ] && [ "$CHUNK_COUNT" != "null" ]; then
                print_success "向量提取成功! 生成 $CHUNK_COUNT 个分块"
            else
                print_success "向量提取成功!"
            fi
        else
            print_success "向量提取成功!"
        fi
        return 0
    else
        print_error "向量提取失败 (HTTP $HTTP_CODE)"
        return 1
    fi
}

# NER 提取
test_ner_extraction() {
    local DOC_ID=$1
    
    print_step "NER 提取 (命名实体识别)"
    
    print_info "文档ID: $DOC_ID"
    print_info "请求URL: ${NER_DOCUMENT_URL}/${DOC_ID}"
    
    RESPONSE=$(curl -s -w "\n%{http_code}" \
        -X POST "${NER_DOCUMENT_URL}/${DOC_ID}" \
        -H "Content-Type: application/json" \
        --connect-timeout 30 \
        --max-time 300)
    
    HTTP_CODE=$(echo "$RESPONSE" | tail -n1)
    BODY=$(echo "$RESPONSE" | sed '$d')
    
    echo -e "${YELLOW}响应状态码:${NC} $HTTP_CODE"
    echo -e "${YELLOW}响应内容:${NC}"
    
    if [ "$JQ_AVAILABLE" = true ]; then
        echo "$BODY" | jq . 2>/dev/null || echo "$BODY"
    else
        echo "$BODY"
    fi
    
    if [ "$HTTP_CODE" = "200" ]; then
        if [ "$JQ_AVAILABLE" = true ]; then
            ENTITY_COUNT=$(echo "$BODY" | jq -r '.data.entityCount // empty' 2>/dev/null)
            RELATION_COUNT=$(echo "$BODY" | jq -r '.data.relationCount // empty' 2>/dev/null)
            if [ -n "$ENTITY_COUNT" ] && [ "$ENTITY_COUNT" != "null" ]; then
                print_success "NER 提取成功! 实体: $ENTITY_COUNT, 关系: $RELATION_COUNT"
            else
                print_success "NER 提取成功!"
            fi
        else
            print_success "NER 提取成功!"
        fi
        return 0
    else
        print_error "NER 提取失败 (HTTP $HTTP_CODE)"
        return 1
    fi
}

# 显示使用帮助
show_help() {
    echo "使用方法: $0 [选项] [host] [port]"
    echo ""
    echo "端到端测试流程: 上传文件 -> 等待解析 -> 向量提取 -> NER提取"
    echo ""
    echo "选项:"
    echo "  -h, --help        显示帮助信息"
    echo "  -e, --e2e         执行完整端到端测试 (默认)"
    echo "  -u, --upload-only 仅执行上传测试"
    echo "  -s, --status      仅查询上次上传的文档状态"
    echo "  -v, --vector      仅执行向量提取（使用上次的文档）"
    echo "  -n, --ner         仅执行NER提取（使用上次的文档）"
    echo ""
    echo "示例:"
    echo "  $0                      # 使用默认配置执行完整端到端测试"
    echo "  $0 192.168.1.100 5232   # 指定服务器地址"
    echo "  $0 -u                   # 仅上传文件"
    echo "  $0 -s                   # 查询上次上传的状态"
    echo "  $0 -v                   # 对上次文档执行向量提取"
    echo "  $0 -n                   # 对上次文档执行NER提取"
}

# 主函数
main() {
    local MODE="e2e"  # 默认执行完整端到端测试
    
    # 解析参数
    while [[ $# -gt 0 ]]; do
        case $1 in
            -h|--help)
                show_help
                exit 0
                ;;
            -e|--e2e)
                MODE="e2e"
                shift
                ;;
            -u|--upload-only)
                MODE="upload"
                shift
                ;;
            -s|--status)
                MODE="status"
                shift
                ;;
            -v|--vector)
                MODE="vector"
                shift
                ;;
            -n|--ner)
                MODE="ner"
                shift
                ;;
            -p|--poll)
                # 兼容旧参数，等同于e2e
                MODE="e2e"
                shift
                ;;
            *)
                if [[ ! "$1" =~ ^- ]]; then
                    if [[ -z "$HOST_SET" ]]; then
                        HOST=$1
                        HOST_SET=true
                    else
                        PORT=$1
                    fi
                fi
                shift
                ;;
        esac
    done
    
    # 更新URL
    BASE_URL="http://${HOST}:${PORT}"
    UPLOAD_URL="${BASE_URL}/api/v1/parse/upload"
    STATUS_URL="${BASE_URL}/parse/status"
    REGISTER_URL="${BASE_URL}/auth/register"
    TEXT_STORAGE_URL="${BASE_URL}/api/v1/graph/text-storage"
    RAG_INDEX_URL="${BASE_URL}/api/rag/index"
    NER_DOCUMENT_URL="${BASE_URL}/api/ner/document"
    
    print_header "文件上传端到端测试"
    echo "目标服务: $BASE_URL"
    echo "测试模式: $MODE"
    echo "时间: $(date '+%Y-%m-%d %H:%M:%S')"
    
    check_dependencies
    
    # 根据模式执行不同操作
    case $MODE in
        status)
            # 仅查询状态
            if [ -f "${SCRIPT_DIR}/.last_document_id" ]; then
                DOCUMENT_ID=$(cat "${SCRIPT_DIR}/.last_document_id")
                print_header "查询解析状态"
                test_parse_status "$DOCUMENT_ID"
            else
                print_error "未找到上次上传的文档ID"
                exit 1
            fi
            ;;
        
        vector)
            # 仅执行向量提取
            if [ -f "${SCRIPT_DIR}/.last_document_id" ]; then
                DOCUMENT_ID=$(cat "${SCRIPT_DIR}/.last_document_id")
                print_header "向量提取测试"
                if get_document_text "$DOCUMENT_ID"; then
                    test_vector_extraction "$DOCUMENT_ID" "$DOCUMENT_TEXT"
                fi
            else
                print_error "未找到上次上传的文档ID"
                exit 1
            fi
            ;;
        
        ner)
            # 仅执行NER提取
            if [ -f "${SCRIPT_DIR}/.last_document_id" ]; then
                DOCUMENT_ID=$(cat "${SCRIPT_DIR}/.last_document_id")
                print_header "NER 提取测试"
                test_ner_extraction "$DOCUMENT_ID"
            else
                print_error "未找到上次上传的文档ID"
                exit 1
            fi
            ;;
        
        upload)
            # 仅上传
            check_test_file
            check_service
            register_test_user
            test_upload
            ;;
        
        e2e)
            # 完整端到端测试
            check_test_file
            check_service
            register_test_user
            
            print_header "步骤 1/4: 文件上传"
            test_upload
            
            if [ -z "$DOCUMENT_ID" ] && [ -f "${SCRIPT_DIR}/.last_document_id" ]; then
                DOCUMENT_ID=$(cat "${SCRIPT_DIR}/.last_document_id")
            fi
            
            if [ -z "$DOCUMENT_ID" ]; then
                print_error "无法获取文档ID，终止测试"
                exit 1
            fi
            
            print_header "步骤 2/4: 等待解析完成"
            if ! poll_parse_status "$DOCUMENT_ID" 60 3; then
                print_error "解析未完成，终止测试"
                exit 1
            fi
            
            print_header "步骤 3/4: 向量提取"
            if get_document_text "$DOCUMENT_ID"; then
                test_vector_extraction "$DOCUMENT_ID" "$DOCUMENT_TEXT"
            else
                print_info "跳过向量提取（无法获取文本）"
            fi
            
            print_header "步骤 4/4: NER 提取"
            test_ner_extraction "$DOCUMENT_ID"
            ;;
    esac
    
    print_header "测试完成"
    echo -e "${GREEN}文档ID: $DOCUMENT_ID${NC}"
    echo "可使用以下命令进行后续操作:"
    echo "  $0 -s            # 查询状态"
    echo "  $0 -v            # 重新向量提取"
    echo "  $0 -n            # 重新NER提取"
}

# 运行主函数
main "$@"