Unimarc/unimarc/unimarc/SearchModel/MokpoLibSearcher.cs

using System;
using System.Net.Http;
using System.Threading.Tasks;
using System.Text.RegularExpressions;
using System.Web;
using UniMarc.SearchModel;
using System.Text;

namespace BokBonCheck
{
    public class MokpoLibSearcher : ILibrarySearcher
    {
        public string AreaCode { get; set; } = string.Empty;
        public string SiteName { get; protected set; }
        public string SiteUrl => "https://mokpolib.or.kr/dls_lt/index.php";
        public bool HttpApiMode { get; set; } = true;

        public int No { get; set; }

        private static readonly HttpClient _httpClient = new HttpClient()
        {
            DefaultRequestHeaders =
            {
                { "User-Agent", "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36" }
            }
        };

        public MokpoLibSearcher(int no, string areaCode, string areaName)
        {
            this.No = no;
            this.AreaCode = areaCode;
            this.SiteName = $"목포시립({areaName})";
        }

        public async Task StartDriver(bool showdriver = false)
        {
            // HTTP 클라이언트 사용으로 별도 드라이버 불필요
            await Task.CompletedTask;
        }

        public void StopDriver()
        {
            // HTTP 클라이언트 사용으로 별도 정리 불필요
        }

        public async Task<BookSearchResult> SearchAsync(string searchTerm)
        {
            var result = new BookSearchResult
            {
                SiteName = SiteName,
                SearchTerm = searchTerm,
                SearchTime = DateTime.Now
            };

            try
            {
                // 검색어 URL 인코딩
                var encodedSearchTerm = HttpUtility.UrlEncode(searchTerm, Encoding.UTF8);
                
                // 검색 URL 구성
                var searchUrl = $"{SiteUrl}?mod=wdDataSearch&act=searchResultList&manageSearch=&detailSearch=";
                
                // 도서관 코드가 있으면 추가
                if (!string.IsNullOrEmpty(AreaCode))
                {
                    searchUrl += $"&manageCode%5B{AreaCode}%5D={AreaCode}";
                }
                
                // 검색 방식을 제목으로 설정하고 검색어 추가
                searchUrl += $"&searchItem%5B%5D=title&searchWord%5B%5D={encodedSearchTerm}";

                Console.WriteLine($"목포시립도서관 검색 URL: {searchUrl}");

                // HTTP GET 요청 실행 (추가 헤더 포함)
                using (var request = new HttpRequestMessage(HttpMethod.Get, searchUrl))
                {
                    // 브라우저와 유사한 헤더 추가
                    request.Headers.Add("Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8");
                    request.Headers.Add("Accept-Language", "ko-KR,ko;q=0.8,en-US;q=0.5,en;q=0.3");
                    request.Headers.Add("Accept-Encoding", "gzip, deflate, br");
                    request.Headers.Add("Connection", "keep-alive");
                    request.Headers.Add("Upgrade-Insecure-Requests", "1");

                    var response = await _httpClient.SendAsync(request);

                    if (!response.IsSuccessStatusCode)
                    {
                        var errorContent = await response.Content.ReadAsStringAsync();
                        throw new HttpRequestException($"HTTP {(int)response.StatusCode} {response.StatusCode}: {errorContent}");
                    }

                    var htmlContent = await response.Content.ReadAsStringAsync();

                    // 검색 결과 수 추출
                    var resultCount = ExtractBookCount(htmlContent, out string errorMessage);

                    if (resultCount == -1)
                    {
                        result.BookCount = 0;
                        result.IsSuccess = false;
                        result.ErrorMessage = errorMessage;
                    }
                    else
                    {
                        result.BookCount = resultCount;
                        result.IsSuccess = true;
                        result.ErrorMessage = $"검색성공({resultCount}권)";
                    }
                }

            }
            catch (Exception ex)
            {
                result.IsSuccess = false;
                result.ErrorMessage = $"검색 오류: {ex.Message}";
                result.BookCount = 0;
                Console.WriteLine($"목포시립도서관 검색 오류: {ex.Message}");
            }

            return result;
        }

        private int ExtractBookCount(string htmlContent, out string errorMessage)
        {
            errorMessage = string.Empty;
            
            try
            {
                // HTML에서 "총 <strong class="cyan">N권(개)</strong>" 패턴 찾기
                var patterns = new[]
                {
                    @"총\s*<strong[^>]*class=""cyan""[^>]*>(\d+)권\(개\)</strong>",
                    @"총\s*<strong[^>]*>(\d+)권\(개\)</strong>",
                    @"총\s*(\d+)권\(개\)",
                    @"총\s*<strong[^>]*class=""cyan""[^>]*>(\d+)</strong>",
                    @"총\s*<strong[^>]*>(\d+)</strong>"
                };

                foreach (var pattern in patterns)
                {
                    var match = Regex.Match(htmlContent, pattern, RegexOptions.IgnoreCase);
                    if (match.Success)
                    {
                        if (int.TryParse(match.Groups[1].Value, out int count))
                        {
                            if (count == 0)
                            {
                                errorMessage = "검색결과없음";
                                return 0;
                            }
                            errorMessage = $"검색성공({count}권)";
                            return count;
                        }
                    }
                }

                // 검색 결과가 없다는 메시지 확인
                if (htmlContent.Contains("검색결과가 없습니다") ||
                    htmlContent.Contains("검색된 자료가 없습니다") ||
                    htmlContent.Contains("자료가 없습니다") ||
                    htmlContent.Contains("총 0권(개)"))
                {
                    errorMessage = "검색결과없음";
                    return 0;
                }

                // 더 넓은 범위로 숫자 찾기 시도
                var generalPatterns = new[]
                {
                    @"총\s*.*?(\d+)\s*권",
                    @"총\s*.*?(\d+)\s*개",
                    @"총\s*.*?(\d+)",
                    @"<strong[^>]*class=""cyan""[^>]*>(\d+)</strong>"
                };

                foreach (var pattern in generalPatterns)
                {
                    var match = Regex.Match(htmlContent, pattern, RegexOptions.IgnoreCase);
                    if (match.Success)
                    {
                        if (int.TryParse(match.Groups[1].Value, out int count))
                        {
                            if (count == 0)
                            {
                                errorMessage = "검색결과없음";
                                return 0;
                            }
                            errorMessage = $"검색성공({count}권)";
                            return count;
                        }
                    }
                }
                    
                errorMessage = "검색결과 패턴을 찾을 수 없음";
                return -1;
            }
            catch (Exception ex)
            {
                errorMessage = $"결과 분석 오류: {ex.Message}";
                return -1;
            }
        }

        public Task WaitForPageChange(OpenQA.Selenium.Support.UI.WebDriverWait wait)
        {
            throw new NotImplementedException();
        }
    }
}