/// <summary>
/// Sets an EncodingCacheItem in the cache for the given key.
/// </summary>
/// <param name="key">Unique key</param>
/// <param name="item">EncodingCacheItem to add</param>
/// <history>
/// [Curtis_Beard] 05/28/2015 FIX: 69, Created for speed improvements for encoding detection
/// </history>
public void SetItem(string key, EncodingCacheItem item)
{
if (cache.ContainsKey(key))
{
// update item incase of change
cache[key] = item;
// Remove it from current position
lruList.Remove(key);
// Add it again, this will result in it being placed on top
lruList.AddLast(key);
}
else
{
cache.Add(key, item);
if (cache.Count > capacity)
{
// over capacity so remove least used item and key
cache.Remove(lruList.First.Value);
lruList.RemoveFirst();
}
lruList.AddLast(key);
}
}
/// <summary>
/// Search a given file for the searchText.
/// </summary>
/// <param name="file">FileInfo object for file to search for searchText</param>
/// <history>
/// [Curtis_Beard] 09/08/2005 Created
/// [Curtis_Beard] 11/21/2005 ADD: update hit count when actual line added
/// [Curtis_Beard] 12/02/2005 CHG: use SearchingFile instead of StatusMessage
/// [Curtis_Beard] 04/21/2006 CHG: use a regular expression match collection to get
/// correct count of hits in a line when using RegEx
/// [Curtis_Beard] 07/03/2006 FIX: 1500174, use a FileStream to open the files readonly
/// [Curtis_Beard] 07/07/2006 FIX: 1512029, RegEx use Case Sensitivity and WholeWords,
/// also use different whole word matching regex
/// [Curtis_Beard] 07/26/2006 ADD: 1512026, column position
/// [Curtis_Beard] 07/26/2006 FIX: 1530023, retrieve file with correct encoding
/// [Curtis_Beard] 09/12/2006 CHG: Converted to C#
/// [Curtis_Beard] 09/28/2006 FIX: check for any plugins before looping through them
/// [Curtis_Beard] 05/18/2006 FIX: 1723815, use correct whole word matching regex
/// [Curtis_Beard] 06/26/2007 FIX: correctly detect plugin extension support
/// [Curtis_Beard] 06/26/2007 FIX: 1779270, increase array size holding context lines
/// [Curtis_Beard] 10/09/2012 FIX: don't overwrite position when getting context lines
/// [Curtis_Beard] 10/12/2012 FIX: get correct position when using whole word option
/// [Curtis_Beard] 10/12/2012 CHG: 32, implement a hit count filter
/// [Curtis_Beard] 10/31/2012 CHG: renamed to SearchFileContents, remove parameter searchText
/// [Curtis_Beard] 08/19/2014 FIX: 57, escape search text when whole word is enabled but not regular expressions
/// [Curtis_Beard] 10/27/2014 CHG: 85, remove leading white space, remove use of newline so each line is in hit object
/// [Curtis_Beard] 02/09/2015 CHG: 92, support for specific file encodings
/// [Curtis_Beard] 03/05/2015 FIX: 64/35, if whole word doesn't pass our check but does pass regex, make it fail. Code cleanup.
/// [Curtis_Beard] 04/02/2015 CHG: remove line number logic and always include line number in MatchResultLine.
/// [Curtis_Beard] 05/18/2015 FIX: 72, don't grab file sample when detect encoding option is turned off.
/// [Curtis_Beard] 05/18/2015 FIX: 69, use same stream to detect encoding and grep contents
/// [Curtis_Beard] 05/26/2015 FIX: 69, add performance setting for file detection
/// [Curtis_Beard] 06/02/2015 FIX: 75, use sample size from performance setting
/// [theblackbunny] 06/25/2015 FIX: 39, remove context lines that intersect with each other in different MatchResults
/// </history>
private void SearchFileContents(FileInfo file)
{
// Raise SearchFile Event
OnSearchingFile(file);
FileStream _stream = null;
StreamReader _reader = null;
int _lineNumber = 0;
MatchResult match = null;
Regex _regularExp;
MatchCollection _regularExpCol = null;
bool _hitOccurred = false;
bool _fileNameDisplayed = false;
int _maxContextLines = 0;
var _context = new string[11];
int _contextIndex = -1;
int _lastHit = 0;
int userFilterCount = 0;
try
{
#region Plugin Processing
if (Plugins != null)
{
for (int i = 0; i < Plugins.Count; i++)
{
// find a valid plugin for this file type
if (Plugins[i].Enabled && Plugins[i].Plugin.IsAvailable)
{
// detect if plugin supports extension
bool isFound = Plugins[i].Plugin.IsFileSupported(file);
// if extension not supported try another plugin
if (!isFound)
continue;
Exception pluginEx = null;
// load plugin and perform grep
if (Plugins[i].Plugin.Load())
{
OnSearchingFileByPlugin(Plugins[i].Plugin.Name);
match = Plugins[i].Plugin.Grep(file, SearchSpec, ref pluginEx);
}
else
{
OnSearchError(file, new Exception(string.Format("Plugin {0} failed to load.", Plugins[i].Plugin.Name)));
}
Plugins[i].Plugin.Unload();
// if the plugin processed successfully
if (pluginEx == null)
{
// check for a hit
if (match != null)
{
match.FromPlugin = true;
// only perform is not using negation
if (!SearchSpec.UseNegation)
{
if (DoesPassHitCountCheck(match))
{
match.Index = MatchResults.Count;
MatchResults.Add(match);
OnFileHit(file, match.Index);
if (SearchSpec.ReturnOnlyFileNames)
match.SetHitCount();
OnLineHit(match, match.Index);
}
}
}
else if (SearchSpec.UseNegation)
{
// no hit but using negation so create one
match = new MatchResult(file) { Index = MatchResults.Count, FromPlugin = true };
MatchResults.Add(match);
OnFileHit(file, match.Index);
}
}
else
{
// the plugin had an error
OnSearchError(file, pluginEx);
}
return;
}
}
}
#endregion
// open stream to file to use in encoding detection if enabled and in grep logic
_stream = file.Open(FileMode.Open, FileAccess.Read, FileShare.ReadWrite);
#region Encoding Detection
string usedEncoder = string.Empty;
System.Text.Encoding encoding = null;
//
// User specified file encoding
//
FileEncoding fileEncoding = SearchSpec.FileEncodings != null && SearchSpec.FileEncodings.Count > 0 ?
(from f in SearchSpec.FileEncodings where f.FilePath.Equals(file.FullName, StringComparison.InvariantCultureIgnoreCase) && f.Enabled select f).ToList().FirstOrDefault()
: null;
if (fileEncoding != null)
{
usedEncoder = "User";
encoding = fileEncoding.Encoding;
}
else
{
//
// Detect file encoding if enabled
//
if (SearchSpec.EncodingDetectionOptions.DetectFileEncoding)
{
// encoding cache check
var key = file.FullName;
if (SearchSpec.EncodingDetectionOptions.UseEncodingCache &&
EncodingCache.Instance.ContainsKey(key))
{
var value = EncodingCache.Instance.GetItem(key);
usedEncoder = value.DetectorName;
encoding = System.Text.Encoding.GetEncoding(value.CodePage);
}
else
{
byte[] sampleBytes;
//Check if can read first
try
{
int sampleSize = EncodingOptions.GetSampleSizeByPerformance(SearchSpec.EncodingDetectionOptions != null ? SearchSpec.EncodingDetectionOptions.PerformanceSetting : EncodingOptions.Performance.Default);
sampleBytes = EncodingTools.ReadFileContentSample(_stream, sampleSize);
}
catch (Exception ex)
{
// can't read file for sample bytes
OnSearchError(file, ex);
return;
}
// detect encoding based on user set performance level that determines what detectors are used
encoding = EncodingDetector.Detect(sampleBytes,
out usedEncoder,
EncodingOptions.GetEncodingDetectorOptionsByPerformance(SearchSpec.EncodingDetectionOptions != null ? SearchSpec.EncodingDetectionOptions.PerformanceSetting : EncodingOptions.Performance.Default),
System.Text.Encoding.Default);
// add to cache if enabled
if (encoding != null && SearchSpec.EncodingDetectionOptions.UseEncodingCache)
{
var value = new EncodingCacheItem() { CodePage = encoding.CodePage, DetectorName = usedEncoder };
EncodingCache.Instance.SetItem(key, value);
}
}
}
else
{
// Use original encoding method before detect encoding option availalbe
usedEncoder = "Default";
encoding = System.Text.Encoding.Default;
}
}
if (encoding == null)
{
// Could not detect file encoding
OnSearchError(file, new Exception("Could not detect file encoding."));
return;
}
OnFileEncodingDetected(file, encoding, usedEncoder);
// process all encoding detectors and display results to output window
//var values = EncodingDetector.DetectAll(sampleBytes);
//if (values.Count > 0)
//{
// System.Diagnostics.Debug.WriteLine(string.Format("File: {0}", file.FullName));
// foreach (var value in values)
// {
// System.Diagnostics.Debug.WriteLine(string.Format("Encoding: {0} ({1})", value.Encoding != null ? value.Encoding.EncodingName : "None", value.Option.ToString()));
// }
// System.Diagnostics.Debug.WriteLine(Environment.NewLine);
//}
#endregion
// could have read some data for the encoding check, seek back to start of file
if (_stream.CanSeek)
{
_stream.Seek(0, SeekOrigin.Begin);
}
_reader = new StreamReader(_stream, encoding);
_maxContextLines = SearchSpec.ContextLines + 1;
do
{
string textLine = _reader.ReadLine();
if (textLine == null)
break;
else
{
_lineNumber += 1;
int _posInStr = -1;
if (SearchSpec.UseRegularExpressions)
{
if (textLine.Length > 0)
{
string pattern = string.Format("{0}{1}{0}", SearchSpec.UseWholeWordMatching ? "\\b" : string.Empty, SearchSpec.SearchText);
RegexOptions options = SearchSpec.UseCaseSensitivity ? RegexOptions.None : RegexOptions.IgnoreCase;
_regularExp = new Regex(pattern, options);
_regularExpCol = _regularExp.Matches(textLine);
if (_regularExpCol.Count > 0)
{
if (SearchSpec.UseNegation)
{
_hitOccurred = true;
}
_posInStr = 1;
}
}
}
else
{
// If we are looking for whole worlds only, perform the check.
if (SearchSpec.UseWholeWordMatching)
{
_regularExp = new Regex("\\b" + Regex.Escape(SearchSpec.SearchText) + "\\b", SearchSpec.UseCaseSensitivity ? RegexOptions.None : RegexOptions.IgnoreCase);
// if match is found, also check against our internal line hit count method to be sure they are in sync
Match mtc = _regularExp.Match(textLine);
if (mtc != null && mtc.Success && RetrieveLineMatches(textLine, SearchSpec).Count > 0)
{
if (SearchSpec.UseNegation)
{
_hitOccurred = true;
}
_posInStr = mtc.Index;
}
}
else
{
_posInStr = textLine.IndexOf(SearchSpec.SearchText, SearchSpec.UseCaseSensitivity ? StringComparison.Ordinal : StringComparison.OrdinalIgnoreCase);
if (SearchSpec.UseNegation && _posInStr > -1)
{
_hitOccurred = true;
}
}
}
//*******************************************
// We found an occurrence of our search text.
//*******************************************
if (_posInStr > -1)
{
//since we have a hit, check to see if negation is checked
if (SearchSpec.UseNegation)
break;
// create new hit and add to collection
if (match == null)
{
match = new MatchResult(file) { Index = MatchResults.Count, DetectedEncoding = encoding };
MatchResults.Add(match);
}
// don't show until passes count check
if (!_fileNameDisplayed && DoesPassHitCountCheck(match))
{
OnFileHit(file, match.Index);
_fileNameDisplayed = true;
}
// If we are only showing filenames, go to the next file.
if (SearchSpec.ReturnOnlyFileNames)
{
if (!_fileNameDisplayed)
{
OnFileHit(file, match.Index);
_fileNameDisplayed = true;
}
//notify that at least 1 hit is in file
match.SetHitCount();
OnLineHit(match, match.Index);
break;
}
// Display context lines if applicable.
if (SearchSpec.ContextLines > 0 && _lastHit <= 0)
{
if (match.Matches.Count > 0 && _lastHit < -_maxContextLines)
{
// Insert a blank space before the context lines.
var matchLine = new MatchResultLine() { Line = string.Empty, LineNumber = -1 };
match.Matches.Add(matchLine);
int _pos = match.Matches.Count - 1;
if (DoesPassHitCountCheck(match))
{
OnLineHit(match, _pos);
}
}
// Display preceding n context lines before the hit.
int tempContextLines = SearchSpec.ContextLines;
// But only output the context lines which are not part of the previous context
if(_lastHit >= -_maxContextLines)
{
tempContextLines = -_lastHit;
}
// Roll back the context index to get the first context line that needs to be displayed
_contextIndex = _contextIndex - tempContextLines;
if(_contextIndex < 0)
{
_contextIndex += _maxContextLines;
}
for (int tempPosInStr = tempContextLines; tempPosInStr >= 1; tempPosInStr--)
{
_contextIndex = _contextIndex + 1;
if (_contextIndex >= _maxContextLines)
_contextIndex = 0;
// If there is a match in the first one or two lines,
// the entire preceeding context may not be available.
if (_lineNumber > tempPosInStr)
{
// Add the context line.
var matchLine = new MatchResultLine() { Line = _context[_contextIndex], LineNumber = _lineNumber - tempPosInStr };
match.Matches.Add(matchLine);
int _pos = match.Matches.Count - 1;
if (DoesPassHitCountCheck(match))
{
OnLineHit(match, _pos);
}
}
}
}
_lastHit = SearchSpec.ContextLines;
//
// Add the actual "hit".
//
var matchLineFound = new MatchResultLine() { Line = textLine, LineNumber = _lineNumber, HasMatch = true };
if (SearchSpec.UseRegularExpressions)
{
_posInStr = _regularExpCol[0].Index;
match.SetHitCount(_regularExpCol.Count);
foreach (Match regExMatch in _regularExpCol)
{
matchLineFound.Matches.Add(new MatchResultLineMatch(regExMatch.Index, regExMatch.Length));
}
}
else
{
var lineMatches = RetrieveLineMatches(textLine, SearchSpec);
match.SetHitCount(lineMatches.Count);
matchLineFound.Matches = lineMatches;
}
matchLineFound.ColumnNumber = _posInStr + 1;
match.Matches.Add(matchLineFound);
int _index = match.Matches.Count - 1;
if (DoesPassHitCountCheck(match))
{
OnLineHit(match, _index);
}
}
else if (SearchSpec.ContextLines > 0)
{
if(_lastHit > 0)
{
//***************************************************
// We didn't find a hit, but since lastHit is > 0, we
// need to display this context line.
//***************************************************
var matchLine = new MatchResultLine() { Line = textLine, LineNumber = _lineNumber };
match.Matches.Add(matchLine);
int _index = match.Matches.Count - 1;
if (DoesPassHitCountCheck(match))
{
OnLineHit(match, _index);
}
}
if(_lastHit >= -_maxContextLines)
{
//*****************************************************
// We continue keeping track of the number of potential
// context lines since the last displayed context line
// until we pass (-_maxContextLines).
//*****************************************************
_lastHit -= 1;
}
} // Found a hit or not.
// If we are showing context lines, keep the last n+1 lines.
if (SearchSpec.ContextLines > 0)
{
_contextIndex += 1;
if (_contextIndex >= _maxContextLines)
_contextIndex = 0;
_context[_contextIndex] = textLine;
}
}
}
while (true);
// send event file/line hit if we haven't yet but it should be
if (!_fileNameDisplayed && match != null && DoesPassHitCountCheck(match))
{
// need to display it
OnFileHit(file, match.Index);
OnLineHit(match, match.Index);
}
// send event for file filtered if it fails the file hit count filter
if (!SearchSpec.UseNegation && !SearchSpec.ReturnOnlyFileNames && match != null && !DoesPassHitCountCheck(match))
{
// remove from grep collection only if
// not negation
// not filenames only
// actually have a hit
// doesn't pass the hit count filter
MatchResults.RemoveAt(MatchResults.Count - 1);
string filterValue = match.HitCount.ToString();
FilterItem filterItem = new FilterItem(new FilterType(FilterType.Categories.File, FilterType.SubCategories.MinimumHitCount), userFilterCount.ToString(), FilterType.ValueOptions.None, false, true);
OnFileFiltered(file, filterItem, filterValue);
}
//
// Check for no hits through out the file
//
if (SearchSpec.UseNegation && _hitOccurred == false)
{
//add the file to the hit list
if (!_fileNameDisplayed)
{
match = new MatchResult(file) { Index = MatchResults.Count, DetectedEncoding = encoding };
MatchResults.Add(match);
OnFileHit(file, match.Index);
}
}
}
finally
{
if (_reader != null)
_reader.Close();
if (_stream != null)
_stream.Close();
}
}
