public LuceneIndexingJob(TimeSpan frequence, Func <EntitiesContext> contextThunk, TimeSpan timeout, LuceneIndexLocation location)
: base("Lucene", frequence, timeout)
{
_updateIndex = () =>
{
using (var context = contextThunk())
{
var indexingService = new LuceneIndexingService(
new EntityRepository <Package>(context),
new EntityRepository <CuratedPackageVersion>(context),
LuceneCommon.GetDirectory(location),
null);
indexingService.UpdateIndex();
}
};
// Updates the index synchronously first time job is created.
// For startup code resiliency, we should handle exceptions for the database being down.
try
{
_updateIndex();
}
catch (SqlException e)
{
QuietLog.LogHandledException(e);
}
catch (DataException e)
{
QuietLog.LogHandledException(e);
}
}
public LuceneIndexingJob(TimeSpan frequence, Func<EntitiesContext> contextThunk, TimeSpan timeout)
: base("Lucene", frequence, timeout)
{
var context = contextThunk();
_indexingService = new LuceneIndexingService(
new EntityRepository<Package>(context),
new EntityRepository<CuratedPackage>(context),
LuceneCommon.GetDirectory(),
null);
// Updates the index synchronously first time job is created.
// For startup code resiliency, we should handle exceptions for the database being down.
try
{
_indexingService.UpdateIndex();
}
catch (SqlException e)
{
QuietLog.LogHandledException(e);
}
catch (DataException e)
{
QuietLog.LogHandledException(e);
}
}
public LuceneIndexingJob(TimeSpan frequence, TimeSpan timeout, LuceneIndexingService indexingService)
: base("Lucene", frequence, timeout)
{
_indexingService = indexingService;
// Updates the index synchronously first time job is created.
// For startup code resiliency, we should handle exceptions for the database being down.
try
{
_indexingService.UpdateIndex();
}
catch (SqlException e)
{
QuietLog.LogHandledException(e);
}
catch (DataException e)
{
QuietLog.LogHandledException(e);
}
}
public LuceneIndexingJob(TimeSpan frequence, TimeSpan timeout, LuceneIndexingService indexingService)
: base("Lucene", frequence, timeout)
{
_indexingService = indexingService;
// Updates the index synchronously first time job is created.
// For startup code resiliency, we should handle exceptions for the database being down.
try
{
_indexingService.UpdateIndex();
}
catch (SqlException e)
{
QuietLog.LogHandledException(e);
}
catch (DataException e)
{
QuietLog.LogHandledException(e);
}
}
public LuceneIndexingJob(TimeSpan frequence, TimeSpan timeout, LuceneIndexingService indexingService)
: base("Lucene", frequence, timeout)
{
_indexingService = indexingService;
try
{
_indexingService.UpdateIndex();
}
catch (SqlException e)
{
// Log but swallow the exception
ErrorSignal.FromCurrentContext().Raise(e);
}
catch (DataException e)
{
// Log but swallow the exception
ErrorSignal.FromCurrentContext().Raise(e);
}
}
public LuceneIndexingJob(TimeSpan frequence, TimeSpan timeout, LuceneIndexingService indexingService)
: base("Lucene", frequence, timeout)
{
_indexingService = indexingService;
try
{
_indexingService.UpdateIndex();
}
catch (SqlException e)
{
// Log but swallow the exception
ErrorSignal.FromCurrentContext().Raise(e);
}
catch (DataException e)
{
// Log but swallow the exception
ErrorSignal.FromCurrentContext().Raise(e);
}
}
public LuceneIndexingJob(TimeSpan frequence, Func <EntitiesContext> contextThunk, TimeSpan timeout)
: base("Lucene", frequence, timeout)
{
_indexingService = new LuceneIndexingService(
new PackageSource(contextThunk()),
LuceneCommon.GetDirectory());
// Updates the index synchronously first time job is created.
// For startup code resiliency, we should handle exceptions for the database being down.
try
{
_indexingService.UpdateIndex();
}
catch (SqlException e)
{
QuietLog.LogHandledException(e);
}
catch (DataException e)
{
QuietLog.LogHandledException(e);
}
}
public LuceneIndexingJob(TimeSpan frequence, TimeSpan timeout)
: base("Lucene", frequence, timeout)
{
indexingService = new LuceneIndexingService();
indexingService.UpdateIndex();
}
public LuceneIndexingJob(TimeSpan frequence, TimeSpan timeout)
: base("Lucene", frequence, timeout)
{
indexingService = new LuceneIndexingService();
indexingService.UpdateIndex();
}
static void Main(string[] args)
{
var packageDataJson = File.ReadAllText(@"samplePackageData.json");
var asArray = JArray.Parse(packageDataJson);
List<Package> packages = LoadPackages(asArray);
Lucene.Net.Store.Directory d = new Lucene.Net.Store.RAMDirectory();
var packageSource = new Mock<IPackageSource>();
packageSource
.Setup(ps => ps.GetPackagesForIndexing(It.IsAny<DateTime?>()))
.Returns(new PackageIndexEntity[0].AsQueryable());
var luceneIndexingService = new LuceneIndexingService(packageSource.Object, d);
luceneIndexingService.UpdateIndex(forceRefresh: true);
luceneIndexingService.AddPackages(packages.Select(p => new PackageIndexEntity(p)).ToList(), true);
luceneSearchService = new LuceneSearchService(d);
TestSearch("jQuery", "jQuery", 1);
TestSearch("jqueryui", "jQuery.UI.Combined", 4);
TestSearch("EntityFramework", "EntityFramework", 1);
TestSearch("Entity Framework", "EntityFramework", 1);
TestSearch("entity", "EntityFramework", 1);
//TestSearch("EF", "EntityFramework", 10); //Currently it's not high in the results AT ALL. Should it be?
TestSearch("Windows.Azure.Storage", "WindowsAzure.Storage", 1);
TestSearch("Windows Azure Storage", "WindowsAzure.Storage", 1);
TestSearch("Azure Storage", "WindowsAzure.Storage", 1);
TestSearch("Windows.Azure.Caching", "Microsoft.WindowsAzure.Caching", 1);
TestSearch("Windows Azure Caching", "Microsoft.WindowsAzure.Caching", 1);
TestSearch("Azure Caching", "Microsoft.WindowsAzure.Caching", 1);
TestSearch("Azure Cache", "Microsoft.WindowsAzure.Caching", 1);
TestSearch("Windows Azure Cache", "Microsoft.WindowsAzure.Caching", 1);
// These 3 popular 'service bus' packages are all in top 5 at time of building this corpus
// Actually NServiceBus has the most downloads, but is lowest ranked for some reason, well, that might change some day.
TestSearch("NServiceBus", "NServiceBus", 5);
TestSearch("Rhino.ServiceBus", "Rhino.ServiceBus", 5);
TestSearch("ServiceBus", "WindowsAzure.ServiceBus", 5);
// By FAR the most popular json package
TestSearch("json", "NewtonSoft.Json", 1);
TestSearch("Json.net", "NewtonSoft.Json", 1);
TestSearch("Newtonsoft", "NewtonSoft.Json", 1);
// A popular json javascript package
TestSearch("json javascript", "json2", 1);
TestSearch("json js", "json2", 1);
// Some other results with justifiably good relevance for json include the 'json' package, and ServiceStack.Text
TestSearch("json", "json", 10);
TestSearch("json", "ServiceStack.Text", 10);
TestSearch("Microsoft web api", "Microsoft.AspNet.WebApi", 3);
TestSearch("microsoft http", "Microsoft.AspNet.WebApi", 4);
TestSearch("microsoft http", "Microsoft.AspNet.WebApi.WebHost", 4);
TestSearch("microsoft http", "Microsoft.Net.Http", 4);
TestSearch("microsoft http", "Microsoft.Web.Infrastructure", 4);
// Finding some misc packages by their distinctive ids
TestSearch("Hammock", "Hammock", 1);
TestSearch("WebActivator", "WebActivatorEx", 1);
TestSearch("Modernizr", "Modernizr", 1);
TestSearch("SimpleInjector", "SimpleInjector", 1);
TestSearch("Simple Injector", "SimpleInjector", 1);
// Some more general term looking searches
TestSearch("Asp.net Mvc", "Microsoft.AspNet.Mvc", 2); //unbelievably it's losing to elmah...
TestSearch("asp.net web pages", "Microsoft.AspNet.WebPages", 1);
TestSearch("logging", "Elmah", 1);
TestSearch("search", "Lucene.Net", 1);
TestSearch("lucene", "Lucene.Net", 1);
TestSearch("search", "NHibernate.Search", 20);
TestSearch("hibernate", "NHibernate", 1);
TestSearch("hibernate", "FluentNHibernate", 2);
TestSearch("hibernate profiler", "NHibernateProfiler", 4);
TestSearch("sql profiler", "MiniProfiler", 1);
TestSearch("sql profiler", "LinqToSqlProfiler", 5);
TestSearch("haacked", "RouteMagic", 2);
TestSearch("haacked", "MVcHaack.Ajax", 4);
TestSearch("haacked", "WebBackgrounder", 5);
TestSearch("haacked", "routedebugger", 5);
TestSearch("NuGet", "NuGet.Core", 3);
TestSearch("NuGet", "NuGet.CommandLine", 2);
TestSearch("NuGet", "NuGet.Build", 3);
TestSearch("Mock", "RhinoMocks", 4);
TestSearch("Mock", "NUnit.Mocks", 4);
TestSearch("Mock", "Moq", 4);
TestSearch("Mock", "FakeItEasy", 20);
TestSearch("Mock", "Ninject.MockingKernel", 30);
TestSearch("Mock", "AutoWrockTestable", 30);
TestSearch("Mock", "Nukito", 40);
TestSearch("Mock", "MockJockey", 50);
TestSearch("Mock", "Machine.Specifications.AutoMocking", 50);
TestSearch("Mock", "WP-Fx.EasyMoq", 60);
TestSearch("razor", "Microsoft.AspNet.Razor", 1);
TestSearch("razor 2", "Microsoft.AspNet.Razor", 1);
TestSearch("memes", "FourOne.Memes", 1);
TestSearch("ninject", "Ninject", 1);
TestSearch("nunit", "NUnit", 1);
TestSearch("testing", "NUnit", 1);
TestSearch("testing", "xunit", 30); //It should probably be higher
TestSearch("Asp.net MVC scaffolding", "MvcScaffolding", 3);
TestSearch("Asp.net MVC scaffolding", "NLibScaffolding", 90);
TestSearch("Asp.net MVC scaffolding", "WijmoMvcScaffolding", 120);
TestSearch("Asp.net MVC scaffolding", "MvcScaffolding.Obsidian", 130);
TestSearch("mvc scaffold", "MvcScaffolding", 8);
TestSearch("mvc scaffold", "ModelScaffolding", 20);
TestSearch("Microsoft web infrastructure", "Microsoft.Web.Infrastructure", 1);
TestSearch("Dotnetopenauth", "DotNetOpenAuth", 1);
TestSearch("OpenID", "DotNetOpenAuth.OpenId.Core", 1);
TestSearch("image resizer", "ImageResizer", 1);
TestSearch("parsing", "CommandLineParser", 10);
TestSearch("knockoutjs", "knockoutjs", 1);
TestSearch("knockout js", "knockoutjs", 3); //it should probably be higher, i.e. beating json.net
// TestSearch("knockout.js", "knockoutjs", 1); // fails to find it
TestSearch("helpers", "microsoft-web-helpers", 1);
TestSearch("fluent mongo", "FluentMongo", 1);
TestSearch("fluent mongo", "MongoFluentUpdater", 15);
TestSearch("mongo", "mongocsharpdriver", 1);
TestSearch("mongo", "FluentMongo", 2);
TestSearch("mongo elmah", "elmah.mongodb", 15); //should be higher?
// These guys are by far the most popular DI packages
TestSearch("injection", "Unity", 2);
TestSearch("injection", "Ninject", 2);
TestSearch("dependency injection", "Unity", 2);
TestSearch("dependency injection", "Ninject", 2);
}
