private void init()
{
SearchTools.Init();
string[] srcRef = File.ReadAllLines(@"refs\prenom.ref");
srcRef = SearchTools.ArrayToLower(srcRef);
refs = SearchTools.buildArray(srcRef, 1);
}
public static int Search4(string txt, string[][] refs, bool allowPartofWord, int length)
{
if (txt == null)
{
return(0);
}
var note = 0;
Parallel.ForEach(refs, (refgroup) =>
{
var pref = refgroup[0].Substring(0, length);
int pos = txt.IndexOf(pref, StringComparison.InvariantCultureIgnoreCase);
while (pos > -1)
{
foreach (string s in refgroup)
{
var co = length;
while (pos + co < txt.Length &&
co < s.Length && (
txt[pos + co] == s[co] || //same caracter
txt[pos + co] + 32 == s[co] ||
SearchTools.includeAccent(txt[pos + co], s[co]))) //include different casse
{
co++;
}
if (co == s.Length)
{
bool isWord =
(pos == 0 || SearchTools.isWordSep(txt[pos - 1]) &&
(pos + s.Length >= txt.Length || SearchTools.isWordSep(txt[pos + s.Length])));
if (isWord || allowPartofWord)
{
note += s.Length;
}
break;
}
}
if (pos + 1 >= txt.Length)
{
break;
}
pos = txt.IndexOf(pref, pos + 1, StringComparison.InvariantCultureIgnoreCase);
}
});
return(note);
}
/// <summary>
/// Search for a list of string in a text
/// sample test 16.6763 second
/// </summary>
/// <param name="txt">text</param>
/// <param name="refs">array of string to search (grouped by first letter)</param>
/// <param name="allowPartofWord">if set allow word or part of word match</param>
/// <param name="length">number of character to use in group</param>
/// <param name="result">result list</param>
/// <returns></returns>
public static void Search4(string txt, string[][] refs, bool allowPartofWord, int length, List <SearchResult> result)
{
Parallel.ForEach(refs, (refgroup) =>
{
var pref = refgroup[0].Substring(0, length);
int pos = txt.IndexOf(pref, StringComparison.InvariantCultureIgnoreCase);
while (pos > -1)
{
foreach (string s in refgroup)
{
var co = length;
while (pos + co < txt.Length &&
co < s.Length && (
txt[pos + co] == s[co] || //same caracter
txt[pos + co] + 32 == s[co] ||
SearchTools.includeAccent(txt[pos + co], s[co]))) //include different casse
{
co++;
}
if (co == s.Length)
{
bool isWord =
(pos == 0 || SearchTools.isWordSep(txt[pos - 1]) &&
(pos + s.Length > txt.Length || SearchTools.isWordSep(txt[pos + s.Length])));
if (isWord || allowPartofWord)
{
result.Add(new SearchResult
{
Word = s,
NearNumber = pos > 2 && SearchTools.isNumber(txt[pos - 2]),
isPartOfWord = !isWord,
Position = pos
});
}
break;
}
}
if (pos + 1 >= txt.Length)
{
break;
}
pos = txt.IndexOf(pref, pos + 1, StringComparison.InvariantCultureIgnoreCase);
}
});
}
public static string GetContent(FileInfo fi)
{
if (!OfficeContentPath.ContainsKey(fi.Extension))
{
return(null);
}
if (fi.Name.StartsWith("~", StringComparison.InvariantCultureIgnoreCase) ||
fi.Name.StartsWith("$", StringComparison.InvariantCultureIgnoreCase))
{
return(null); //avoid temp file or in recycle bin
}
var result = new StringBuilder();
var targets = OfficeContentPath[fi.Extension].Split(',');
ZipFile zipFile;
using (zipFile = ZipFile.Read(fi.FullName))
{
foreach (string target in targets)
{
var filesList = new List <ZipEntry>();
if (target.EndsWith("*", StringComparison.InvariantCultureIgnoreCase))
{
filesList = zipFile.SelectEntries("*", target.Substring(0, target.Length - 1)).ToList();
}
else
{
filesList = zipFile.SelectEntries(target).ToList();
}
foreach (ZipEntry zEntry in filesList)
{
MemoryStream tempS = new MemoryStream();
zEntry.Extract(tempS);
result.Append(Encoding.UTF8.GetString(tempS.ToArray()));
tempS.Close();
}
}
return(SearchTools.XmlToText(result.ToString()));
}
}
/// <summary>
/// Search for a list of string in a text
/// Faster but full word only
/// Sample 14,496 for 20000 iteration
/// </summary>
/// <param name="txt">text</param>
/// <param name="refs">array of string to search (grouped by first letter)</param>
/// <param name="length">number of character to use in group</param>
/// <returns></returns>
public static void SearchWord(string txt, string[][] refs, int length, List <SearchResult> result)
{
//var refgroup = refs[2];
Parallel.ForEach(refs, (refgroup) =>
{
var pref = refgroup[0].Substring(0, length);
int pos = 0;
while (pos < txt.Length)
{
if (refgroup[0][0] == txt[pos] & (pos == 0 || SearchTools.isWordSep(txt[pos - 1])))//new word start and first char equal group letter
{
var txtLength = txt.Length;
var co = 1;
string word = txt[pos].ToString();
while ((pos + co < txt.Length && !SearchTools.isWordSep(txt[pos + co])))//search end of current word extract word
{
word += txt[pos + co];
co++;
}
foreach (string s in refgroup)
{
if (word.Length == s.Length && string.Equals(word, s, StringComparison.CurrentCultureIgnoreCase))
{
result.Add(new SearchResult
{
Word = s,
NearNumber = pos > 2 && SearchTools.isNumber(txt[pos - 2]),
Position = pos
});
pos += word.Length;
break;
}
}
}
pos++;
}
});
}
