private InfoPoco CatchInfoAndReturnInfoPoco(string word)
{
string template = @"http://tw.dictionary.search.yahoo.com/search?p={0}";
var url = string.Format(template, AppHelper.UrlEncode(word));
var html = "";
var pattern = "";
using (WebClient wc = new WebClient())
{
wc.Headers["Host"] = "tw.dictionary.search.yahoo.com";
wc.Headers["User-Agent"] =
"Mozilla/5.0 (Macintosh; Intel Mac OS X 10.10; rv:40.0) Gecko/20100101 Firefox/40.0";
wc.Headers["Accept"] = "*/*";
wc.Headers["Accept-Language"] = "zh-TW,zh;q=0.8,en-US;q=0.5,en;q=0.3";
wc.Headers["Referer"] = "http://tw.dictionary.search.yahoo.com/search";
wc.Headers["Origin"] = "http://tw.dictionary.search.yahoo.com";
byte[] myByte = wc.DownloadData(url);
var script = Encoding.GetEncoding((int)65001).GetString(myByte, 0, myByte.Length); //chinese
pattern = @"<ol class="".+?searchCenterMiddle.+?"">.+?</ol>";
html = Regex.Match(script, pattern).Value;
}
pattern = @"<div class=""dd algo.+?>(?<Content>.+?)</div></li>";
var matchArr = Regex.Matches(html, pattern).OfType<Match>()
.Skip(1)
.Reverse()
.Skip(1)
.Reverse();//trim head and tail
var mm = new InfoPoco(word);
var descriptionList = new List<string>();
foreach (Match m in matchArr)
{
var v = m.Groups["Content"].Value;
if (v.IndexOf("pronunciation_pos", StringComparison.Ordinal) != -1)
{
pattern = @"<span.+?id=""pronunciation_pos"".+?>(?<Phonetic>.+?)</span>";
var phonetic = Regex.Match(v, pattern).Groups["Phonetic"].Value;
mm.Phonetic = Regex.Replace(phonetic, @"<.+?>", "");
}
else
{
descriptionList.Add(AppHelper.GetClearAttributeHtml(v));
}
}
var description = string.Join("", descriptionList.ToArray());
if (!string.IsNullOrEmpty(description))
{
description = string.Format(Resource1.DescriptTemplate, description);
mm.Description = description;
}
return mm;
}
private void UpdateVocabularyInfo(Vocabulary vocabulary, InfoPoco infoPoco)
{
vocabulary.Phonetic = infoPoco.Phonetic;
vocabulary.Description = infoPoco.Description;
vocabulary.IsCatchInfo = true;
}
