private static string GetXrmlTextAll(IEnumerable <string> urls)
{
const string xpath = @"//div[@style='font-family: monospace;']";
var sb = new StringBuilder();
foreach (var url in urls)
{
try
{
var source = WebFetch.Fetch(url);
var doc = new HtmlDocument();
doc.LoadHtml(source);
var target = doc.DocumentNode.SelectSingleNode(xpath);
var text = HttpUtility.HtmlDecode(target.InnerText);
sb.Append(text);
}
catch (Exception ex)
{
// TODO: swallow it whole!
// yikes. or maybe log the issue, down the road....
}
}
return(sb.ToString());
}
private IEnumerable <string> GetRecentGutenbergBooks()
{
var outlinks = new List <string>();
// <div class="pgdbrecent">
//<h2><a name="a267"></a><a href="/browse/authors/a#a267">Appleton, Victor [pseud.]</a></h2>
//<ul>
// <li class="pgdbetext"><a href="/ebooks/952">Tom Swift and His Air Glider, or Seeking the Platinum Treasure</a> (English) (as Author)</li>
// <li class="pgdbetext"><a href="/ebooks/4711">Tom Swift in the City of Gold, or, Marvelous Adventures Underground</a> (English) (as Author)</li>
//</ul>
const string url = "http://www.gutenberg.org/browse/recent/last1";
const string xpath = "//div[@class='pgdbrecent']/ul/li/a";
var source = WebFetch.Fetch(url);
var doc = new HtmlDocument();
doc.LoadHtml(source);
var links = doc.DocumentNode.SelectNodes(xpath);
foreach (var l in links)
{
// TODO: this does not always exists
// so we will need to trap for missing files....
var link = "http://www.gutenberg.org" + l.Attributes["href"].Value + ".txt.utf8";
outlinks.Add(link);
}
return(outlinks);
}
// perhaps cache this list locally, and only regen as required?
// because it's a pain to pull....
// that should be true of all of of web-pulls
// cache locally, and only rebuild as required.
// hrm.... MORE COMPLICATED THAT I CURRENTLY REQUIRE
private List <string> GetXrmlPageLinks()
{
var source = WebFetch.Fetch(LibrarySource);
var doc = new HtmlDocument();
doc.LoadHtml(source);
var target = doc.DocumentNode.SelectNodes("//div[@class='mainlist']/div/ul/li/a");
return(target.Select(link => link.Attributes["href"].Value).ToList());
}
// TODO: build interfaces for getting all contents, random contents, and others
// return as Text sources
// which will have to be updated to use an optional URL instead of just file
private String GetGutenberg()
{
var urls = GetRecentGutenbergBooks();
var rnd = new Random();
var source = string.Empty;
// since a given text might not actually have a .txt.utf8 version
// pull another item if we've got an empty one
// TODO: we're redoing the randomization each time.
// UGH. however, this should all be temporary, not part of the final product....
while (source.Length == 0)
{
var subUrls = urls.OrderBy(x => rnd.Next()).Take(1).ToArray <string>();
source = WebFetch.Fetch(subUrls[0]);
}
return(source); // in this case , it's string data
}
public string Extract(string url)
{
var text = string.Empty;
try
{
const string xpath = @"//div[@style='font-family: monospace;']";
var source = WebFetch.Fetch(url);
var doc = new HtmlDocument();
doc.LoadHtml(source);
var target = doc.DocumentNode.SelectSingleNode(xpath);
text = HttpUtility.HtmlDecode(target.InnerText);
}
catch (Exception ex)
{
// TODO: log it, or something....
}
return(text);
}
