/// <summary>
/// An article with given HTML extractor, title, HTML text and article type
/// </summary>
/// <param name="htmlPlainTextExtractor">The specified HTML extractor</param>
/// <param name="title">Title of article</param>
/// <param name="htmlText">HTML text of article</param>
/// <param name="titlePath">File path of title</param>
/// <param name="summaryPath">File path of summary</param>
/// <param name="typePath">File path of type</param>
/// <param name="type">Type of article</param>
public Article(IHtmlPlainTextExtractor htmlPlainTextExtractor, string title, string htmlText, string titlePath, string summaryPath, string typePath, ArticleType type = ArticleType.NotSpecified)
{
this.htmlPlainTextExtractor = htmlPlainTextExtractor;
this.Title = title;
this.HtmlText = htmlText;
this.TitlePath = titlePath;
this.SummaryPath = summaryPath;
this.TypePath = typePath;
this.Type = type;
}
/// <summary>
/// Query all articles in specific dataset, and use a given extractor as the HTML extractor
/// </summary>
/// <param name="extractor">The plain text extractor for article</param>
/// <returns>List of articles</returns>
public List<Article> Query(IHtmlPlainTextExtractor extractor)
{
List<Article> articles = new List<Article>();
// Get file IDs
// Must assume:
// * The filenames in title folder must be same as the ones in summary folder
// * Filename starts from 0, increase by 1
string[] files = Directory.GetFiles(titlePath);
System.Diagnostics.Debug.Assert(files.Length == Directory.GetFiles(summaryPath).Length);
// Query each article
for (int i = 0; i < files.Length; ++i)
{
articles.Add(Query(i, extractor));
}
return articles;
}
/// <summary>
/// Constructor with specified dataset and default HTML extractor
/// </summary>
/// <param name="dataset">The query dataset</param>
/// <param name="defaultExtractor">The HTML extractor for articles</param>
public ArticleQuerier(QueryDataset dataset, IHtmlPlainTextExtractor defaultExtractor)
{
// Generate dataset path
switch (dataset)
{
case QueryDataset.TrainDataset:
this.datasetPath = Path.Combine(General.Constants.DatasetPath, General.Constants.TrainDatasetDir);
break;
case QueryDataset.TestDataset:
this.datasetPath = Path.Combine(General.Constants.DatasetPath, General.Constants.TestDatasetDir);
break;
default:
throw new NotImplementedException();
}
this.titlePath = Path.Combine(datasetPath, Constants.TitleDir);
this.summaryPath = Path.Combine(datasetPath, Constants.SummaryDir);
this.typePath = Path.Combine(datasetPath, Constants.TypeDir);
this.defaultExtractor = defaultExtractor;
}
/// <summary>
/// Query an article with given ID in specific dataset, and use a given extractor as the HTML extractor
/// </summary>
/// <param name="id">The ID of article</param>
/// <param name="extractor">The HTML text extractor for article</param>
/// <returns>The article</returns>
public Article Query(int id, IHtmlPlainTextExtractor extractor)
{
string title;
string summary;
ArticleType type = ArticleType.NotSpecified;
string titleFilePath = Path.Combine(titlePath, id.ToString());
string summaryFilePath = Path.Combine(summaryPath, id.ToString());
string typeFilePath = Path.Combine(typePath, id.ToString());
// Read title and summary
try
{
FileStream titleFileStream = new FileStream(titleFilePath, FileMode.Open, FileAccess.Read);
StreamReader titleFileStreamReader = new StreamReader(titleFileStream);
title = titleFileStreamReader.ReadToEnd();
titleFileStreamReader.Close();
FileStream summaryFileStream = new FileStream(summaryFilePath, FileMode.Open, FileAccess.Read);
StreamReader summaryFileStreamReader = new StreamReader(summaryFileStream);
summary = summaryFileStreamReader.ReadToEnd();
summaryFileStreamReader.Close();
}
catch (IOException)
{
return null;
}
// Read marked type
try
{
if (File.Exists(typeFilePath))
{
FileStream typeFileStream = new FileStream(typeFilePath, FileMode.Open, FileAccess.Read);
StreamReader typeFileStreamReader = new StreamReader(typeFileStream);
string typeString = typeFileStreamReader.ReadToEnd();
typeFileStreamReader.Close();
int typeInt = Convert.ToInt32(typeString);
if (Enum.IsDefined(typeof(ArticleType), typeInt))
{
type = (ArticleType)typeInt;
}
else
{
throw new FormatException();
}
}
else
{
type = ArticleType.NotSpecified;
}
}
catch (Exception)
{
// Type not specified yet
type = ArticleType.NotSpecified;
}
// Construct article
Article article = new Article(extractor,
title, summary,
titleFilePath, summaryFilePath, typeFilePath,
type);
return article;
}
/// <summary>
/// An empty article with given HTML extractor
/// </summary>
/// <param name="htmlPlainTextExtractor">The specified HTML extractor</param>
protected Article(IHtmlPlainTextExtractor htmlPlainTextExtractor)
: this(htmlPlainTextExtractor, "", "")
{
}
/// <summary>
/// An article with given HTML extractor, title, HTML text and article type
/// </summary>
/// <param name="htmlPlainTextExtractor">The specified HTML extractor</param>
/// <param name="title">Title of article</param>
/// <param name="htmlText">HTML text of article</param>
/// <param name="type">Type of article</param>
public Article(IHtmlPlainTextExtractor htmlPlainTextExtractor, string title, string htmlText, ArticleType type = ArticleType.NotSpecified)
: this(htmlPlainTextExtractor, title, htmlText, "", "", "", type)
{
}
