using System; using System.Collections.Generic; using System.Diagnostics; using System.Linq; using System.Net; using System.Threading.Tasks; using FluentValidation.Results; using NLog; using NzbDrone.Common.Extensions; using NzbDrone.Common.Http; using NzbDrone.Core.Configuration; using NzbDrone.Core.Http.CloudFlare; using NzbDrone.Core.Indexers.Events; using NzbDrone.Core.Indexers.Exceptions; using NzbDrone.Core.IndexerSearch.Definitions; using NzbDrone.Core.Messaging.Events; using NzbDrone.Core.Parser.Model; using NzbDrone.Core.ThingiProvider; namespace NzbDrone.Core.Indexers { public abstract class HttpIndexerBase : IndexerBase where TSettings : IProviderConfig, new() { protected const int MaxNumResultsPerQuery = 1000; protected readonly IHttpClient _httpClient; protected readonly IEventAggregator _eventAggregator; public IDictionary Cookies { get; set; } public override bool SupportsRss => true; public override bool SupportsSearch => true; public override bool SupportsRedirect => false; public override bool FollowRedirect => false; public override IndexerCapabilities Capabilities { get; protected set; } public virtual int PageSize => 0; public virtual TimeSpan RateLimit => TimeSpan.FromSeconds(2); public abstract IIndexerRequestGenerator GetRequestGenerator(); public abstract IParseIndexerResponse GetParser(); public HttpIndexerBase(IHttpClient httpClient, IEventAggregator eventAggregator, IIndexerStatusService indexerStatusService, IConfigService configService, Logger logger) : base(indexerStatusService, configService, logger) { _httpClient = httpClient; _eventAggregator = eventAggregator; } public override Task Fetch(MovieSearchCriteria searchCriteria) { //TODO: Re-Enable when All Indexer Caps are fixed and tests don't fail //if (!SupportsSearch) //{ // return Task.FromResult(new Task()); //} return FetchReleases(g => SetCookieFunctions(g).GetSearchRequests(searchCriteria)); } public override Task Fetch(MusicSearchCriteria searchCriteria) { if (!SupportsSearch) { return Task.FromResult(new IndexerPageableQueryResult()); } return FetchReleases(g => SetCookieFunctions(g).GetSearchRequests(searchCriteria)); } public override Task Fetch(TvSearchCriteria searchCriteria) { if (!SupportsSearch) { return Task.FromResult(new IndexerPageableQueryResult()); } return FetchReleases(g => SetCookieFunctions(g).GetSearchRequests(searchCriteria)); } public override Task Fetch(BookSearchCriteria searchCriteria) { if (!SupportsSearch) { return Task.FromResult(new IndexerPageableQueryResult()); } return FetchReleases(g => SetCookieFunctions(g).GetSearchRequests(searchCriteria)); } public override Task Fetch(BasicSearchCriteria searchCriteria) { if (!SupportsSearch) { return Task.FromResult(new IndexerPageableQueryResult()); } return FetchReleases(g => SetCookieFunctions(g).GetSearchRequests(searchCriteria)); } public override async Task Download(HttpUri link) { Cookies = GetCookies(); var requestBuilder = new HttpRequestBuilder(link.FullUri); if (Cookies != null) { requestBuilder.SetCookies(Cookies); } var downloadBytes = Array.Empty(); try { var response = await _httpClient.ExecuteAsync(requestBuilder.Build()); downloadBytes = response.ResponseData; } catch (Exception) { _indexerStatusService.RecordFailure(Definition.Id); _logger.Error("Download failed"); } return downloadBytes; } protected IIndexerRequestGenerator SetCookieFunctions(IIndexerRequestGenerator generator) { //A func ensures cookies are always updated to the latest. This way, the first page could update the cookies and then can be reused by the second page. generator.GetCookies = () => { var cookies = _indexerStatusService.GetIndexerCookies(Definition.Id); var expiration = _indexerStatusService.GetIndexerCookiesExpirationDate(Definition.Id); if (expiration < DateTime.Now) { cookies = null; } return cookies; }; generator.CookiesUpdater = (cookies, expiration) => { UpdateCookies(cookies, expiration); }; return generator; } protected virtual IDictionary GetCookies() { var cookies = _indexerStatusService.GetIndexerCookies(Definition.Id); var expiration = _indexerStatusService.GetIndexerCookiesExpirationDate(Definition.Id); if (expiration < DateTime.Now) { cookies = null; } return cookies; } protected void UpdateCookies(IDictionary cookies, DateTime? expiration) { Cookies = cookies; _indexerStatusService.UpdateCookies(Definition.Id, cookies, expiration); } protected virtual async Task FetchReleases(Func pageableRequestChainSelector, bool isRecent = false) { var releases = new List(); var result = new IndexerPageableQueryResult(); var url = string.Empty; try { var generator = GetRequestGenerator(); var parser = GetParser(); parser.CookiesUpdater = (cookies, expiration) => { _indexerStatusService.UpdateCookies(Definition.Id, cookies, expiration); }; var pageableRequestChain = pageableRequestChainSelector(generator); var fullyUpdated = false; ReleaseInfo lastReleaseInfo = null; if (isRecent) { lastReleaseInfo = _indexerStatusService.GetLastRssSyncReleaseInfo(Definition.Id); } for (int i = 0; i < pageableRequestChain.Tiers; i++) { var pageableRequests = pageableRequestChain.GetTier(i); foreach (var pageableRequest in pageableRequests) { var pagedReleases = new List(); foreach (var request in pageableRequest) { url = request.Url.FullUri; var page = await FetchPage(request, parser); result.Queries.Add(page); pagedReleases.AddRange(page.Releases); if (isRecent && page.Releases.Any()) { if (lastReleaseInfo == null) { fullyUpdated = true; break; } var oldestReleaseDate = page.Releases.Select(v => v.PublishDate).Min(); if (oldestReleaseDate < lastReleaseInfo.PublishDate || page.Releases.Any(v => v.DownloadUrl == lastReleaseInfo.DownloadUrl)) { fullyUpdated = true; break; } if (pagedReleases.Count >= MaxNumResultsPerQuery && oldestReleaseDate < DateTime.UtcNow - TimeSpan.FromHours(24)) { fullyUpdated = false; break; } } else if (pagedReleases.Count >= MaxNumResultsPerQuery) { break; } if (!IsFullPage(page.Releases)) { break; } } releases.AddRange(pagedReleases.Where(IsValidRelease)); } if (releases.Any()) { break; } } if (isRecent && !releases.Empty()) { var ordered = releases.OrderByDescending(v => v.PublishDate).ToList(); if (!fullyUpdated && lastReleaseInfo != null) { var gapStart = lastReleaseInfo.PublishDate; var gapEnd = ordered.Last().PublishDate; _logger.Warn("Indexer {0} rss sync didn't cover the period between {1} and {2} UTC. Search may be required.", Definition.Name, gapStart, gapEnd); } lastReleaseInfo = ordered.First(); _indexerStatusService.UpdateRssSyncStatus(Definition.Id, lastReleaseInfo); } _indexerStatusService.RecordSuccess(Definition.Id); } catch (WebException webException) { if (webException.Status == WebExceptionStatus.NameResolutionFailure || webException.Status == WebExceptionStatus.ConnectFailure) { _indexerStatusService.RecordConnectionFailure(Definition.Id); } else { _indexerStatusService.RecordFailure(Definition.Id); } if (webException.Message.Contains("502") || webException.Message.Contains("503") || webException.Message.Contains("timed out")) { _logger.Warn("{0} server is currently unavailable. {1} {2}", this, url, webException.Message); } else { _logger.Warn("{0} {1} {2}", this, url, webException.Message); } } catch (TooManyRequestsException ex) { if (ex.RetryAfter != TimeSpan.Zero) { _indexerStatusService.RecordFailure(Definition.Id, ex.RetryAfter); } else { _indexerStatusService.RecordFailure(Definition.Id, TimeSpan.FromHours(1)); } _logger.Warn("API Request Limit reached for {0}", this); } catch (HttpException ex) { _indexerStatusService.RecordFailure(Definition.Id); _logger.Warn("{0} {1}", this, ex.Message); } catch (RequestLimitReachedException) { _indexerStatusService.RecordFailure(Definition.Id, TimeSpan.FromHours(1)); _logger.Warn("API Request Limit reached for {0}", this); } catch (IndexerAuthException) { _indexerStatusService.RecordFailure(Definition.Id); _logger.Warn("Invalid Credentials for {0} {1}", this, url); } catch (CloudFlareCaptchaException ex) { _indexerStatusService.RecordFailure(Definition.Id); ex.WithData("FeedUrl", url); if (ex.IsExpired) { _logger.Error(ex, "Expired CAPTCHA token for {0}, please refresh in indexer settings.", this); } else { _logger.Error(ex, "CAPTCHA token required for {0}, check indexer settings.", this); } } catch (IndexerException ex) { _indexerStatusService.RecordFailure(Definition.Id); _logger.Warn(ex, "{0}", url); } catch (Exception ex) { _indexerStatusService.RecordFailure(Definition.Id); ex.WithData("FeedUrl", url); _logger.Error(ex, "An error occurred while processing indexer feed. {0}", url); } result.Releases = CleanupReleases(releases); return result; } public override IndexerCapabilities GetCapabilities() { return Capabilities ?? ((IndexerDefinition)Definition).Capabilities; } protected virtual bool IsValidRelease(ReleaseInfo release) { if (release.DownloadUrl == null) { return false; } return true; } protected virtual bool IsFullPage(IList page) { return PageSize != 0 && page.Count >= PageSize; } protected virtual async Task FetchPage(IndexerRequest request, IParseIndexerResponse parser) { var response = await FetchIndexerResponse(request); try { var releases = parser.ParseResponse(response).ToList(); if (releases.Count == 0) { _logger.Trace(response.Content); } return new IndexerQueryResult { Releases = releases, ElapsedTime = response.ElapsedTime, StatusCode = (int)response.HttpResponse.StatusCode }; } catch (Exception ex) { ex.WithData(response.HttpResponse, 128 * 1024); _logger.Trace("Unexpected Response content ({0} bytes): {1}", response.HttpResponse.ResponseData.Length, response.HttpResponse.Content); throw; } } protected virtual bool CheckIfLoginNeeded(HttpResponse httpResponse) { if (httpResponse.StatusCode == HttpStatusCode.Unauthorized) { return true; } return false; } protected virtual Task DoLogin() { return Task.CompletedTask; } protected virtual void ModifyRequest(IndexerRequest request) { request.HttpRequest.Cookies.Clear(); if (Cookies != null) { foreach (var cookie in Cookies) { request.HttpRequest.Cookies.Add(cookie.Key, cookie.Value); } } } protected virtual async Task FetchIndexerResponse(IndexerRequest request) { _logger.Debug("Downloading Feed " + request.HttpRequest.ToString(false)); if (request.HttpRequest.RateLimit < RateLimit) { request.HttpRequest.RateLimit = RateLimit; } if (_configService.LogIndexerResponse) { request.HttpRequest.LogResponseContent = true; } request.HttpRequest.AllowAutoRedirect = FollowRedirect; Cookies = GetCookies(); if (Cookies != null) { foreach (var cookie in Cookies) { request.HttpRequest.Cookies.Add(cookie.Key, cookie.Value); } } var stopWatch = Stopwatch.StartNew(); request.HttpRequest.SuppressHttpError = true; var response = await _httpClient.ExecuteAsync(request.HttpRequest); stopWatch.Stop(); // Check reponse to see if auth is needed, if needed try again if (CheckIfLoginNeeded(response)) { _logger.Trace("Attempting to re-auth based on indexer search response"); await DoLogin(); ModifyRequest(request); response = await _httpClient.ExecuteAsync(request.HttpRequest); } // Throw any other http error we get after attempting auth if (response.HasHttpError) { _logger.Warn("HTTP Error - {0}", response); if ((int)response.StatusCode == 429) { throw new TooManyRequestsException(request.HttpRequest, response); } else { throw new HttpException(request.HttpRequest, response); } } UpdateCookies(Cookies, DateTime.Now + TimeSpan.FromDays(30)); return new IndexerResponse(request, response, stopWatch.ElapsedMilliseconds); } protected async Task ExecuteAuth(HttpRequest request) { var stopWatch = Stopwatch.StartNew(); var response = await _httpClient.ExecuteAsync(request); stopWatch.Stop(); _eventAggregator.PublishEvent(new IndexerAuthEvent(Definition.Id, !response.HasHttpError, stopWatch.ElapsedMilliseconds)); return response; } protected override async Task Test(List failures) { failures.AddIfNotNull(await TestConnection()); } protected virtual async Task TestConnection() { try { var parser = GetParser(); parser.CookiesUpdater = (cookies, expiration) => { _indexerStatusService.UpdateCookies(Definition.Id, cookies, expiration); }; var generator = GetRequestGenerator(); generator = SetCookieFunctions(generator); var firstRequest = generator.GetSearchRequests(new BasicSearchCriteria { SearchType = "search" }).GetAllTiers().FirstOrDefault()?.FirstOrDefault(); if (firstRequest == null) { return new ValidationFailure(string.Empty, "No rss feed query available. This may be an issue with the indexer or your indexer category settings."); } var releases = await FetchPage(firstRequest, parser); if (releases.Releases.Empty()) { return new ValidationFailure(string.Empty, "Query successful, but no results were returned from your indexer. This may be an issue with the indexer or your indexer category settings."); } } catch (IndexerAuthException ex) { _logger.Warn("Indexer returned result for RSS URL, Credentials appears to be invalid: " + ex.Message); return new ValidationFailure("ApiKey", "Invalid API Key"); } catch (RequestLimitReachedException ex) { _logger.Warn("Request limit reached: " + ex.Message); } catch (CloudFlareCaptchaException ex) { if (ex.IsExpired) { return new ValidationFailure("CaptchaToken", "CloudFlare CAPTCHA token expired, please Refresh."); } else { return new ValidationFailure("CaptchaToken", "Site protected by CloudFlare CAPTCHA. Valid CAPTCHA token required."); } } catch (UnsupportedFeedException ex) { _logger.Warn(ex, "Indexer feed is not supported"); return new ValidationFailure(string.Empty, "Indexer feed is not supported: " + ex.Message); } catch (IndexerException ex) { _logger.Warn(ex, "Unable to connect to indexer"); return new ValidationFailure(string.Empty, "Unable to connect to indexer. " + ex.Message); } catch (HttpException ex) { if (ex.Response.StatusCode == HttpStatusCode.BadRequest && ex.Response.Content.Contains("not support the requested query")) { _logger.Warn(ex, "Indexer does not support the query"); return new ValidationFailure(string.Empty, "Indexer does not support the current query. Check if the categories and or searching for movies are supported. Check the log for more details."); } else { _logger.Warn(ex, "Unable to connect to indexer"); return new ValidationFailure(string.Empty, "Unable to connect to indexer, check the log for more details"); } } catch (Exception ex) { _logger.Warn(ex, "Unable to connect to indexer"); return new ValidationFailure(string.Empty, "Unable to connect to indexer, check the log for more details"); } return null; } } }