public static SrmDocument RecalculateAlignments(SrmDocument document, IProgressMonitor progressMonitor)
{
var newSources = ListAvailableRetentionTimeSources(document.Settings);
var newResultsSources = ListSourcesForResults(document.Settings.MeasuredResults, newSources);
var allLibraryRetentionTimes = ReadAllRetentionTimes(document, newSources);
var newFileAlignments = new List <FileRetentionTimeAlignments>();
IProgressStatus progressStatus = new ProgressStatus(@"Aligning retention times"); // CONSIDER: localize? Will users see this?
foreach (var retentionTimeSource in newResultsSources.Values)
{
progressStatus = progressStatus.ChangePercentComplete(100 * newFileAlignments.Count / newResultsSources.Count);
progressMonitor.UpdateProgress(progressStatus);
try
{
var fileAlignments = CalculateFileRetentionTimeAlignments(retentionTimeSource.Name, allLibraryRetentionTimes, progressMonitor);
newFileAlignments.Add(fileAlignments);
}
catch (OperationCanceledException)
{
progressMonitor.UpdateProgress(progressStatus.Cancel());
return(null);
}
}
var newDocRt = new DocumentRetentionTimes(newSources.Values, newFileAlignments);
var newDocument = document.ChangeSettings(document.Settings.ChangeDocumentRetentionTimes(newDocRt));
Debug.Assert(IsLoaded(newDocument));
progressMonitor.UpdateProgress(progressStatus.Complete());
return(newDocument);
}
public override bool Run(CancellationTokenSource cancelToken)
{
IProgressStatus status = new ProgressStatus();
//UpdateProgress(status);
bool success = true;
foreach (var spectrumFilename in SpectrumFileNames)
{
try
{
var pr = new ProcessRunner();
var psi = new ProcessStartInfo(@"java",
@"-Xmx8G -jar MSGFPlus.jar " +
$@"-s {spectrumFilename} -d {FastaFileNames[0]} -tda 1" +
$@"-t {precursorMzTolerance} -ti {isotopeErrorRange.Item1},{isotopeErrorRange.Item2} " +
$@"-m {fragmentationMethod} -inst {instrumentType} -e {enzyme} -ntt {ntt} -maxMissedCleavages {maxMissedCleavages}");
pr.Run(psi, string.Empty, this, ref status);
status = status.Complete().ChangeMessage(Resources.DDASearchControl_SearchProgress_Search_done);
}
catch (OperationCanceledException)
{
status = status.Cancel().ChangeMessage(Resources.DDASearchControl_SearchProgress_Search_canceled);
success = false;
}
catch (Exception ex)
{
status = status.ChangeErrorException(ex).ChangeMessage(string.Format(Resources.DdaSearch_Search_failed__0, ex.Message));
UpdateProgress(status);
success = false;
}
if (cancelToken.IsCancellationRequested)
{
success = false;
}
}
var deleteHelper = new DeleteTempHelper(modsFile);
deleteHelper.DeletePath();
return(success);
}
/// <summary>
/// Returns a string representation of the report based on the document.
/// </summary>
private string ReportToCsvString(SrmDocument doc, char separator, IProgressMonitor progressMonitor)
{
IProgressStatus status = new ProgressStatus(string.Format(Resources.ReportSpec_ReportToCsvString_Exporting__0__report, Name));
progressMonitor.UpdateProgress(status);
Report report = Report.Load(this);
StringWriter writer = new StringWriter();
using (Database database = new Database(doc.Settings)
{
ProgressMonitor = progressMonitor,
Status = status,
PercentOfWait = 80
})
{
database.AddSrmDocument(doc);
status = database.Status;
ResultSet resultSet;
try
{
resultSet = report.Execute(database);
}
catch (Exception)
{
progressMonitor.UpdateProgress(status.Cancel());
throw;
}
progressMonitor.UpdateProgress(status = status.ChangePercentComplete(95));
ResultSet.WriteReportHelper(resultSet, separator, writer, LocalizationHelper.CurrentCulture);
}
writer.Flush();
string csv = writer.ToString();
writer.Close();
progressMonitor.UpdateProgress(status.Complete());
return(csv);
}
// ReSharper restore UnusedMember.Local
private bool Load(ILoadMonitor loader)
{
ProgressStatus status =
new ProgressStatus(string.Format(Resources.BiblioSpecLibrary_Load_Loading__0__library,
Path.GetFileName(FilePath)));
loader.UpdateProgress(status);
long lenRead = 0;
// AdlerChecksum checksum = new AdlerChecksum();
try
{
// Use a buffered stream for initial read
BufferedStream stream = new BufferedStream(CreateStream(loader), 32 * 1024);
int countHeader = (int)LibHeaders.count * 4;
byte[] libHeader = new byte[countHeader];
if (stream.Read(libHeader, 0, countHeader) != countHeader)
{
throw new InvalidDataException(Resources.BiblioSpecLibrary_Load_Data_truncation_in_library_header_File_may_be_corrupted);
}
lenRead += countHeader;
// Check the first byte of the primary version number to determine
// whether the format is little- or big-endian. Little-endian will
// have the version number in this byte, while big-endian will have zero.
if (libHeader[(int)LibHeaders.version1 * 4] == 0)
{
_bigEndian = true;
}
int numSpectra = GetInt32(libHeader, (int)LibHeaders.num_spectra);
var dictLibrary = new Dictionary <LibKey, BiblioSpectrumInfo>(numSpectra);
var setSequences = new HashSet <LibSeqKey>();
string revStr = string.Format("{0}.{1}", // Not L10N
GetInt32(libHeader, (int)LibHeaders.version1),
GetInt32(libHeader, (int)LibHeaders.version2));
Revision = float.Parse(revStr, CultureInfo.InvariantCulture);
// checksum.MakeForBuff(libHeader, AdlerChecksum.ADLER_START);
countHeader = (int)SpectrumHeaders.count * 4;
byte[] specHeader = new byte[1024];
byte[] specSequence = new byte[1024];
for (int i = 0; i < numSpectra; i++)
{
int percent = i * 100 / numSpectra;
if (status.PercentComplete != percent)
{
// Check for cancellation after each integer change in percent loaded.
if (loader.IsCanceled)
{
loader.UpdateProgress(status.Cancel());
return(false);
}
// If not cancelled, update progress.
loader.UpdateProgress(status = status.ChangePercentComplete(percent));
}
// Read spectrum header
int bytesRead = stream.Read(specHeader, 0, countHeader);
if (bytesRead != countHeader)
{
throw new InvalidDataException(Resources.BiblioSpecLibrary_Load_Data_truncation_in_spectrum_header_File_may_be_corrupted);
}
// If this is the first header, and the sequence length is zero,
// then this is a Linux format library. Switch to linux format,
// and start over.
if (i == 0 && GetInt32(specHeader, (int)SpectrumHeaders.seq_len) == 0)
{
_linuxFormat = true;
stream.Seek(lenRead, SeekOrigin.Begin);
// Re-ead spectrum header
countHeader = (int)SpectrumHeadersLinux.count * 4;
bytesRead = stream.Read(specHeader, 0, countHeader);
if (bytesRead != countHeader)
{
throw new InvalidDataException(Resources.BiblioSpecLibrary_Load_Data_truncation_in_spectrum_header_File_may_be_corrupted);
}
}
lenRead += bytesRead;
// checksum.MakeForBuff(specHeader, checksum.ChecksumValue);
int charge = GetInt32(specHeader, (int)SpectrumHeaders.charge);
if (charge > TransitionGroup.MAX_PRECURSOR_CHARGE)
{
throw new InvalidDataException(Resources.BiblioSpecLibrary_Load_Invalid_precursor_charge_found_File_may_be_corrupted);
}
int numPeaks = GetInt32(specHeader, (int)SpectrumHeaders.num_peaks);
int seqLength = GetInt32(specHeader, (_linuxFormat ?
(int)SpectrumHeadersLinux.seq_len : (int)SpectrumHeaders.seq_len));
int copies = GetInt32(specHeader, (_linuxFormat ?
(int)SpectrumHeadersLinux.copies : (int)SpectrumHeaders.copies));
// Read sequence information
int countSeq = (seqLength + 1) * 2;
if (stream.Read(specSequence, 0, countSeq) != countSeq)
{
throw new InvalidDataException(Resources.BiblioSpecLibrary_Load_Data_truncation_in_spectrum_sequence_File_may_be_corrupted);
}
lenRead += countSeq;
// checksum.MakeForBuff(specSequence, checksum.ChecksumValue);
// Store in dictionary
if (IsUnmodified(specSequence, seqLength + 1, seqLength))
{
// These libraries should not have duplicates, but just in case.
// CONSIDER: Emit error about redundancy?
// These legacy libraries assume [+57.0] modified Cysteine
LibKey key = new LibKey(GetCModified(specSequence, ref seqLength), 0, seqLength, charge);
if (!dictLibrary.ContainsKey(key))
{
dictLibrary.Add(key, new BiblioSpectrumInfo((short)copies, (short)numPeaks, lenRead));
}
setSequences.Add(new LibSeqKey(key));
}
// Read over peaks
int countPeaks = 2 * sizeof(Single) * numPeaks;
stream.Seek(countPeaks, SeekOrigin.Current); // Skip spectrum
lenRead += countPeaks;
// checksum.MakeForBuff(specPeaks, checksum.ChecksumValue);
}
// Checksum = checksum.ChecksumValue;
_dictLibrary = dictLibrary;
_setSequences = setSequences;
loader.UpdateProgress(status.Complete());
return(true);
}
catch (InvalidDataException x)
{
loader.UpdateProgress(status.ChangeErrorException(x));
return(false);
}
catch (IOException x)
{
loader.UpdateProgress(status.ChangeErrorException(x));
return(false);
}
catch (Exception x)
{
x = new Exception(string.Format(Resources.BiblioSpecLibrary_Load_Failed_loading_library__0__, FilePath), x);
loader.UpdateProgress(status.ChangeErrorException(x));
return(false);
}
finally
{
if (ReadStream != null)
{
// Close the read stream to ensure we never leak it.
// This only costs on extra open, the first time the
// active document tries to read.
try { ReadStream.CloseStream(); }
catch (IOException) {}
}
}
}
public static void AddSpectra(MidasLibSpec libSpec, MsDataFilePath[] resultsFiles, SrmDocument doc, ILoadMonitor monitor, out List <MsDataFilePath> failedFiles)
{
// Get spectra from results files
var newSpectra = new List <DbSpectrum>();
var progress = new ProgressStatus(string.Empty).ChangeMessage(Resources.MidasLibrary_AddSpectra_Reading_MIDAS_spectra);
const int percentResultsFiles = 80;
failedFiles = new List <MsDataFilePath>();
for (var i = 0; i < resultsFiles.Length; i++)
{
var resultsFile = resultsFiles[i];
try
{
monitor.UpdateProgress(progress.ChangePercentComplete(i * percentResultsFiles / resultsFiles.Length));
var filePath = resultsFile.GetFilePath();
if (File.Exists(filePath))
{
var sampleIndex = resultsFile.GetSampleIndex();
using (var msd = new MsDataFileImpl(filePath, sampleIndex == -1 ? 0 : sampleIndex, resultsFile.GetLockMassParameters(), requireVendorCentroidedMS2: true))
{
if (ChromatogramDataProvider.HasChromatogramData(msd) && SpectraChromDataProvider.HasSpectrumData(msd))
{
var chromPrecursors = ReadChromPrecursorsFromMsd(msd, monitor).ToList();
newSpectra.AddRange(ReadDbSpectraFromMsd(msd, monitor));
MatchSpectraToChrom(newSpectra, chromPrecursors, monitor);
}
}
MatchSpectraToPeptides(newSpectra, doc, monitor);
}
else
{
failedFiles.Add(resultsFile);
}
}
catch (Exception x)
{
monitor.UpdateProgress(progress.ChangeErrorException(x));
failedFiles.Add(resultsFile);
}
if (monitor.IsCanceled)
{
monitor.UpdateProgress(progress.Cancel());
return;
}
}
if (!newSpectra.Any())
{
monitor.UpdateProgress(progress.Complete());
return;
}
progress = progress.ChangePercentComplete(percentResultsFiles);
monitor.UpdateProgress(progress);
// Add spectra to library
var midasLib = !File.Exists(libSpec.FilePath) ? Create(libSpec) : Load(libSpec, monitor);
if (midasLib == null)
{
monitor.UpdateProgress(progress.ChangeErrorException(new Exception(Resources.MidasLibrary_AddSpectra_Error_loading_MIDAS_library_for_adding_spectra_)));
return;
}
progress = progress.ChangeMessage(Resources.MidasLibrary_AddSpectra_Adding_spectra_to_MIDAS_library);
monitor.UpdateProgress(progress);
var results = new Dictionary <string, DbResultsFile>();
if (midasLib._spectra != null)
{
foreach (var kvp in midasLib._spectra)
{
results[kvp.Key.FilePath] = kvp.Key;
}
}
using (var sessionFactory = SessionFactoryFactory.CreateSessionFactory(libSpec.FilePath, typeof(MidasLibrary), false))
using (var session = new SessionWithLock(sessionFactory.OpenSession(), new ReaderWriterLock(), true))
using (var transaction = session.BeginTransaction())
{
for (var i = 0; i < newSpectra.Count; i++)
{
if (monitor.IsCanceled)
{
transaction.Rollback();
monitor.UpdateProgress(progress.Cancel());
return;
}
var spectrum = newSpectra[i];
monitor.UpdateProgress(progress.ChangePercentComplete(percentResultsFiles + (int)(100.0 * i / newSpectra.Count)));
DbResultsFile resultsFile;
if (!results.TryGetValue(spectrum.ResultsFile.FilePath, out resultsFile))
{
resultsFile = new DbResultsFile(spectrum.ResultsFile)
{
Id = null
};
results[spectrum.ResultsFile.FilePath] = resultsFile;
session.SaveOrUpdate(resultsFile);
}
else if (midasLib._spectra != null)
{
List <DbSpectrum> existingSpectra;
if (midasLib._spectra.TryGetValue(resultsFile, out existingSpectra) &&
existingSpectra.Any(x => Equals(x.ResultsFile.FilePath, spectrum.ResultsFile.FilePath) &&
Equals(x.PrecursorMz, spectrum.PrecursorMz) &&
Equals(x.RetentionTime, spectrum.RetentionTime)))
{
// This spectrum already exists in the library
continue;
}
}
var spectrumNewDisconnected = new DbSpectrum(spectrum)
{
Id = null, ResultsFile = resultsFile
};
session.SaveOrUpdate(spectrumNewDisconnected);
}
transaction.Commit();
monitor.UpdateProgress(progress.Complete());
}
}
private bool Load(IProgressMonitor monitor)
{
_spectra = null;
if (FilePath == null)
{
return(false);
}
var info = new FileInfo(FilePath);
if (!info.Exists || info.Length == 0)
{
return(false);
}
var progress = new ProgressStatus(string.Empty).ChangeMessage(Resources.MidasLibrary_Load_Loading_MIDAS_library);
monitor.UpdateProgress(progress);
var spectra = new Dictionary <DbResultsFile, List <DbSpectrum> >();
try
{
using (var sessionFactory = SessionFactoryFactory.CreateSessionFactory(FilePath, typeof(MidasLibrary), false))
using (var session = new SessionWithLock(sessionFactory.OpenSession(), new ReaderWriterLock(), false))
{
var libInfo = session.CreateCriteria(typeof(DbLibInfo)).List <DbLibInfo>();
if (libInfo.Count != 1)
{
throw new Exception(Resources.MidasLibrary_Load_Error_reading_LibInfo_from_MIDAS_library);
}
SchemaVersion = libInfo[0].SchemaVersion;
LibraryGuid = libInfo[0].Guid;
var readSpectra = session.CreateCriteria(typeof(DbSpectrum)).List <DbSpectrum>();
progress = progress.ChangeSegments(0, readSpectra.Count);
foreach (var spectrum in readSpectra)
{
if (monitor.IsCanceled)
{
monitor.UpdateProgress(progress.Cancel());
return(false);
}
progress = progress.NextSegment();
monitor.UpdateProgress(progress);
List <DbSpectrum> list;
if (!spectra.TryGetValue(spectrum.ResultsFile, out list))
{
list = new List <DbSpectrum>();
spectra[spectrum.ResultsFile] = list;
}
list.Add(spectrum);
}
}
}
catch
{
monitor.UpdateProgress(progress.Cancel());
return(false);
}
_spectra = spectra;
monitor.UpdateProgress(progress.Complete());
return(true);
}
private SrmDocument LookupProteinMetadata(SrmDocument docOrig, IProgressMonitor progressMonitor)
{
lock (_processedNodes)
{
// Check to make sure this operation was not canceled while this thread was
// waiting to acquire the lock. This also cleans up pending work.
if (progressMonitor.IsCanceled)
{
return(null);
}
IProgressStatus progressStatus = new ProgressStatus(Resources.ProteinMetadataManager_LookupProteinMetadata_resolving_protein_details);
int nResolved = 0;
int nUnresolved = docOrig.PeptideGroups.Select(pg => pg.ProteinMetadata.NeedsSearch()).Count();
if ((nUnresolved > 0) && !docOrig.Settings.PeptideSettings.BackgroundProteome.IsNone)
{
// Do a quick check to see if background proteome already has the info
if (!docOrig.Settings.PeptideSettings.BackgroundProteome.NeedsProteinMetadataSearch)
{
try
{
using (var proteomeDb = docOrig.Settings.PeptideSettings.BackgroundProteome.OpenProteomeDb())
{
foreach (PeptideGroupDocNode nodePepGroup in docOrig.PeptideGroups)
{
if (_processedNodes.ContainsKey(nodePepGroup.Id.GlobalIndex))
{
// We did this before we were interrupted
progressMonitor.UpdateProgress(progressStatus = progressStatus.ChangePercentComplete(100 * nResolved++ / nUnresolved));
}
else if (nodePepGroup.ProteinMetadata.NeedsSearch())
{
var proteinMetadata = proteomeDb.GetProteinMetadataByName(nodePepGroup.Name);
if ((proteinMetadata == null) && !Equals(nodePepGroup.Name, nodePepGroup.OriginalName))
{
proteinMetadata = proteomeDb.GetProteinMetadataByName(nodePepGroup.OriginalName); // Original name might hit
}
if ((proteinMetadata == null) && !String.IsNullOrEmpty(nodePepGroup.ProteinMetadata.Accession))
{
proteinMetadata = proteomeDb.GetProteinMetadataByName(nodePepGroup.ProteinMetadata.Accession); // Parsed accession might hit
}
if ((proteinMetadata != null) && !proteinMetadata.NeedsSearch())
{
// Background proteome has already resolved this
_processedNodes.Add(nodePepGroup.Id.GlobalIndex, proteinMetadata);
progressMonitor.UpdateProgress(
progressStatus =
progressStatus.ChangePercentComplete(100 * nResolved++ / nUnresolved));
}
}
if (progressMonitor.IsCanceled)
{
progressMonitor.UpdateProgress(progressStatus.Cancel());
return(null);
}
}
}
}
// ReSharper disable once EmptyGeneralCatchClause
catch
{
// The protDB file is busy, or some other issue - just go directly to web
}
}
}
if (nResolved != nUnresolved)
{
try
{
// Now go to the web for more protein metadata (or pretend to, depending on WebEnabledFastaImporter.DefaultWebAccessMode)
var docNodesWithUnresolvedProteinMetadata = new Dictionary <ProteinSearchInfo, PeptideGroupDocNode>();
var proteinsToSearch = new List <ProteinSearchInfo>();
foreach (PeptideGroupDocNode node in docOrig.PeptideGroups)
{
if (node.ProteinMetadata.NeedsSearch() && !_processedNodes.ContainsKey(node.Id.GlobalIndex)) // Did we already process this?
{
var proteinMetadata = node.ProteinMetadata;
if (proteinMetadata.WebSearchInfo.IsEmpty()) // Never even been hit with regex
{
// Use Regexes to get some metadata, and a search term
var parsedProteinMetaData = FastaImporter.ParseProteinMetaData(proteinMetadata);
if ((parsedProteinMetaData == null) || Equals(parsedProteinMetaData.Merge(proteinMetadata), proteinMetadata.SetWebSearchCompleted()))
{
// That didn't parse well enough to make a search term, or didn't add any new info - just set it as searched so we don't keep trying
_processedNodes.Add(node.Id.GlobalIndex, proteinMetadata.SetWebSearchCompleted());
if (progressMonitor.IsCanceled)
{
progressMonitor.UpdateProgress(progressStatus.Cancel());
return(null);
}
progressMonitor.UpdateProgress(progressStatus = progressStatus.ChangePercentComplete(100 * nResolved++ / nUnresolved));
proteinMetadata = null; // No search to be done
}
else
{
proteinMetadata = proteinMetadata.Merge(parsedProteinMetaData); // Fill in any gaps with parsed info
}
}
if (proteinMetadata != null)
{
// We note the sequence length because it's useful in disambiguating search results
proteinsToSearch.Add(new ProteinSearchInfo(new DbProteinName(null, proteinMetadata),
node.PeptideGroup.Sequence == null ? 0 : node.PeptideGroup.Sequence.Length));
docNodesWithUnresolvedProteinMetadata.Add(proteinsToSearch.Last(), node);
}
}
}
if (progressMonitor.IsCanceled)
{
progressMonitor.UpdateProgress(progressStatus.Cancel());
return(null);
}
progressMonitor.UpdateProgress(progressStatus = progressStatus.ChangePercentComplete(100 * nResolved / nUnresolved));
// Now we actually hit the internet
if (proteinsToSearch.Any())
{
foreach (var result in FastaImporter.DoWebserviceLookup(proteinsToSearch, progressMonitor, false)) // Resolve them all, now
{
Debug.Assert(!result.GetProteinMetadata().NeedsSearch());
_processedNodes.Add(docNodesWithUnresolvedProteinMetadata[result].Id.GlobalIndex, result.GetProteinMetadata());
if (progressMonitor.IsCanceled)
{
progressMonitor.UpdateProgress(progressStatus.Cancel());
return(null);
}
progressMonitor.UpdateProgress(progressStatus = progressStatus.ChangePercentComplete(100 * nResolved++ / nUnresolved));
}
}
}
catch (OperationCanceledException)
{
progressMonitor.UpdateProgress(progressStatus.Cancel());
return(null);
}
}
// And finally write back to the document
var listProteins = new List <PeptideGroupDocNode>();
foreach (PeptideGroupDocNode node in docOrig.MoleculeGroups)
{
if (_processedNodes.ContainsKey(node.Id.GlobalIndex))
{
listProteins.Add(node.ChangeProteinMetadata(_processedNodes[node.Id.GlobalIndex]));
}
else
{
listProteins.Add(node);
}
}
var docNew = docOrig.ChangeChildrenChecked(listProteins.Cast <DocNode>().ToArray());
progressMonitor.UpdateProgress(progressStatus.Complete());
return((SrmDocument)docNew);
}
}
private BackgroundProteome Load(IDocumentContainer container, PeptideSettings settings, SrmDocument docCurrent, bool isBackgroundLoad)
{
// Only allow one background proteome to load at a time. This can
// get tricky, if the user performs an undo and then a redo across
// a change in background proteome.
// Our only priority is accessing web services to add missing protein metadata.
// There may also be a load initiation by the Peptide Settings dialog as foreground task,
// it takes priority over the background task.
lock (_lockLoadBackgroundProteome)
{
BackgroundProteome originalBackgroundProteome = settings.BackgroundProteome;
BackgroundProteome validatedBackgroundProtome = originalBackgroundProteome.DatabaseValidated
? originalBackgroundProteome
: new BackgroundProteome(originalBackgroundProteome.BackgroundProteomeSpec);
if (IsNotLoadedExplained(settings, validatedBackgroundProtome, true) == null)
{
// protein metadata is resolved
CompleteProcessing(container, validatedBackgroundProtome);
Helpers.AssignIfEquals(ref validatedBackgroundProtome, originalBackgroundProteome);
return(validatedBackgroundProtome); // No change needed
}
// we are here to resolve the protein metadata
string name = originalBackgroundProteome.Name;
IProgressStatus progressStatus =
new ProgressStatus(string.Format(Resources.BackgroundProteomeManager_LoadBackground_Resolving_protein_details_for__0__proteome, name));
try
{
// The transaction commit for writing the digestion info can be very lengthy, avoid lock timeouts
// by doing that work in a tempfile that no other thread knows aboout
using (FileSaver fs = new FileSaver(originalBackgroundProteome.DatabasePath, StreamManager))
{
File.Copy(originalBackgroundProteome.DatabasePath, fs.SafeName, true);
var digestHelper = new DigestHelper(this, container, docCurrent, name, fs.SafeName, true);
bool success = digestHelper.LookupProteinMetadata(ref progressStatus);
if (digestHelper.IsCanceled || !success)
{
// Processing was canceled
if (docCurrent != null)
{
EndProcessing(docCurrent);
}
UpdateProgress(progressStatus.Cancel());
return(null);
}
using (var proteomeDb = ProteomeDb.OpenProteomeDb(originalBackgroundProteome.DatabasePath))
{
proteomeDb.DatabaseLock.AcquireWriterLock(int.MaxValue); // Wait for any existing readers to complete, prevent any new ones
try
{
if (File.GetLastWriteTime(fs.RealName) <= File.GetLastWriteTime(fs.SafeName)) // Don't overwrite if foreground task has already updated
{
proteomeDb.CloseDbConnection(); // Get rid of any file handles
if (!fs.Commit())
{
if (docCurrent != null)
{
EndProcessing(docCurrent);
}
throw new IOException(string.Format(Resources.BackgroundProteomeManager_LoadBackground_Unable_to_rename_temporary_file_to__0__,
fs.RealName));
}
}
}
finally
{
proteomeDb.DatabaseLock.ReleaseWriterLock();
}
}
var updatedProteome = new BackgroundProteome(originalBackgroundProteome);
using (var proteomeDb = originalBackgroundProteome.OpenProteomeDb())
{
proteomeDb.AnalyzeDb(); // Now it's safe to start this potentially lengthy indexing operation
}
CompleteProcessing(container, updatedProteome);
UpdateProgress(progressStatus.Complete());
return(updatedProteome);
}
}
catch (Exception x)
{
var message = new StringBuilder();
message.AppendLine(
string.Format(Resources.BackgroundProteomeManager_LoadBackground_Failed_updating_background_proteome__0__,
name));
message.Append(x.Message);
UpdateProgress(progressStatus.ChangeErrorException(new IOException(message.ToString(), x)));
return(null);
}
}
}
private SrmDocument LookupProteinMetadata(SrmDocument docOrig, IProgressMonitor progressMonitor)
{
lock (_processedNodes)
{
// Check to make sure this operation was not canceled while this thread was
// waiting to acquire the lock. This also cleans up pending work.
if (progressMonitor.IsCanceled)
{
return(null);
}
IProgressStatus progressStatus = new ProgressStatus(Resources.ProteinMetadataManager_LookupProteinMetadata_resolving_protein_details);
int nResolved = 0;
int nUnresolved = docOrig.PeptideGroups.Select(pg => pg.ProteinMetadata.NeedsSearch()).Count();
if ((nUnresolved > 0) && !docOrig.Settings.PeptideSettings.BackgroundProteome.IsNone)
{
// Do a quick check to see if background proteome already has the info
if (!docOrig.Settings.PeptideSettings.BackgroundProteome.NeedsProteinMetadataSearch)
{
try
{
using (var proteomeDb = docOrig.Settings.PeptideSettings.BackgroundProteome.OpenProteomeDb())
{
foreach (PeptideGroupDocNode nodePepGroup in docOrig.PeptideGroups)
{
if (_processedNodes.ContainsKey(nodePepGroup.Id.GlobalIndex))
{
// We did this before we were interrupted
nResolved++;
}
else if (nodePepGroup.ProteinMetadata.NeedsSearch())
{
var proteinMetadata = proteomeDb.GetProteinMetadataByName(nodePepGroup.Name);
if ((proteinMetadata == null) && !Equals(nodePepGroup.Name, nodePepGroup.OriginalName))
{
proteinMetadata = proteomeDb.GetProteinMetadataByName(nodePepGroup.OriginalName); // Original name might hit
}
if ((proteinMetadata == null) && !String.IsNullOrEmpty(nodePepGroup.ProteinMetadata.Accession))
{
proteinMetadata = proteomeDb.GetProteinMetadataByName(nodePepGroup.ProteinMetadata.Accession); // Parsed accession might hit
}
if ((proteinMetadata != null) && !proteinMetadata.NeedsSearch())
{
// Background proteome has already resolved this
_processedNodes.Add(nodePepGroup.Id.GlobalIndex, proteinMetadata);
nResolved++;
}
}
if (!UpdatePrecentComplete(progressMonitor, 100 * nResolved / nUnresolved, ref progressStatus))
{
return(null);
}
}
}
}
// ReSharper disable once EmptyGeneralCatchClause
catch
{
// The protDB file is busy, or some other issue - just go directly to web
}
}
}
if (nResolved != nUnresolved)
{
try
{
// Now go to the web for more protein metadata (or pretend to, depending on WebEnabledFastaImporter.DefaultWebAccessMode)
var docNodesWithUnresolvedProteinMetadata = new Dictionary <ProteinSearchInfo, PeptideGroupDocNode>();
var proteinsToSearch = new List <ProteinSearchInfo>();
foreach (PeptideGroupDocNode node in docOrig.PeptideGroups)
{
if (node.ProteinMetadata.NeedsSearch() && !_processedNodes.ContainsKey(node.Id.GlobalIndex)) // Did we already process this?
{
var proteinMetadata = node.ProteinMetadata;
if (proteinMetadata.WebSearchInfo.IsEmpty()) // Never even been hit with regex
{
// Use Regexes to get some metadata, and a search term
var parsedProteinMetaData = FastaImporter.ParseProteinMetaData(proteinMetadata);
if ((parsedProteinMetaData == null) || Equals(parsedProteinMetaData.Merge(proteinMetadata), proteinMetadata.SetWebSearchCompleted()))
{
// That didn't parse well enough to make a search term, or didn't add any new info - just set it as searched so we don't keep trying
_processedNodes.Add(node.Id.GlobalIndex, proteinMetadata.SetWebSearchCompleted());
if (!UpdatePrecentComplete(progressMonitor, 100 * nResolved++ / nUnresolved, ref progressStatus))
{
return(null);
}
proteinMetadata = null; // No search to be done
}
else
{
proteinMetadata = proteinMetadata.Merge(parsedProteinMetaData); // Fill in any gaps with parsed info
}
}
if (proteinMetadata != null)
{
// We note the sequence length because it's useful in disambiguating search results
proteinsToSearch.Add(new ProteinSearchInfo(new DbProteinName(null, proteinMetadata),
node.PeptideGroup.Sequence == null ? 0 : node.PeptideGroup.Sequence.Length));
docNodesWithUnresolvedProteinMetadata.Add(proteinsToSearch.Last(), node);
}
}
}
if (!UpdatePrecentComplete(progressMonitor, 100 * nResolved / nUnresolved, ref progressStatus))
{
return(null);
}
// Now we actually hit the internet
if (proteinsToSearch.Any())
{
foreach (var result in FastaImporter.DoWebserviceLookup(proteinsToSearch, progressMonitor, false)) // Resolve them all, now
{
Assume.IsTrue(!result.GetProteinMetadata().NeedsSearch());
_processedNodes.Add(docNodesWithUnresolvedProteinMetadata[result].Id.GlobalIndex, result.GetProteinMetadata());
if (!UpdatePrecentComplete(progressMonitor, 100 * nResolved++ / nUnresolved, ref progressStatus))
{
return(null);
}
}
}
}
catch (OperationCanceledException)
{
progressMonitor.UpdateProgress(progressStatus.Cancel());
return(null);
}
}
// And finally write back to the document
var listProteins = new List <PeptideGroupDocNode>();
foreach (PeptideGroupDocNode node in docOrig.MoleculeGroups)
{
if (_processedNodes.TryGetValue(node.Id.GlobalIndex, out var proteinMetadata))
{
// Compare existing and proposed metadata, ignoring name difference in case user changed
// the name manually in the Targets tree while a background metadata lookup was going on, and
// ignoring web search details since the existing node probably hasn't any yet.
//
// This fixes issue https://skyline.ms/announcements/home/support/thread.view?rowId=49107 in which:
// the user pasted a protein sequence into the Targets tree
// then tried to type in a name to replace the default assigned name "sequence1"
// after a few seconds the displayed name reverted to "sequence1" upon background protein metadata search completion
// N.B. as this is timing dependent, and our automated tests are mandated to not require internet
// access, writing a test for this fix (i.e. adding timings to the fake web lookup system) proved to
// be tricky and finally deemed not worth the effort for this fairly obscure problem.
if (!Equals(node.ProteinMetadata.Name, proteinMetadata.Name) && // Different name
Equals(node.ProteinMetadata.ChangeName(proteinMetadata.Name).ClearWebSearchInfo(), // But otherwise identical
proteinMetadata.ClearWebSearchInfo()))
{
// Leave (apparently user-renamed) node alone, and note the web search that was actually used.
listProteins.Add(node.ChangeProteinMetadata(node.ProteinMetadata.ChangeWebSearchInfo(proteinMetadata.WebSearchInfo)));
}
else
{
// Update the protein metadata for this node, if any
listProteins.Add(node.ChangeProteinMetadata(proteinMetadata));
}
}
else
{
// Not yet processed
listProteins.Add(node);
}
}
var docNew = docOrig.ChangeChildrenChecked(listProteins.Cast <DocNode>().ToArray());
progressMonitor.UpdateProgress(progressStatus.Complete());
return((SrmDocument)docNew);
}
}
// ReSharper restore NonLocalizedString
public static List <string> ConvertPilotFiles(IList <string> inputFiles, IProgressMonitor progress, ProgressStatus status)
{
string groupConverterExePath = null;
var inputFilesPilotConverted = new List <string>();
for (int index = 0; index < inputFiles.Count; index++)
{
string inputFile = inputFiles[index];
if (!inputFile.EndsWith(BiblioSpecLiteBuilder.EXT_PILOT))
{
inputFilesPilotConverted.Add(inputFile);
continue;
}
string outputFile = Path.ChangeExtension(inputFile, BiblioSpecLiteBuilder.EXT_PILOT_XML);
// Avoid re-converting files that have already been converted
if (File.Exists(outputFile))
{
// Avoid duplication, in case the user accidentally adds both .group and .group.xml files
// for the same results
if (!inputFiles.Contains(outputFile))
{
inputFilesPilotConverted.Add(outputFile);
}
continue;
}
string message = string.Format(Resources.VendorIssueHelper_ConvertPilotFiles_Converting__0__to_xml, Path.GetFileName(inputFile));
int percent = index * 100 / inputFiles.Count;
progress.UpdateProgress(status = status.ChangeMessage(message).ChangePercentComplete(percent));
if (groupConverterExePath == null)
{
var key = Registry.LocalMachine.OpenSubKey(KEY_PROTEIN_PILOT, false);
if (key != null)
{
string proteinPilotCommandWithArgs = (string)key.GetValue(string.Empty);
var proteinPilotCommandWithArgsSplit =
proteinPilotCommandWithArgs.Split(new[] { "\" \"" }, StringSplitOptions.RemoveEmptyEntries); // Remove " "%1" // Not L10N
string path = Path.GetDirectoryName(proteinPilotCommandWithArgsSplit[0].Trim(new[] { '\\', '\"' })); // Remove preceding "
if (path != null)
{
var groupFileExtractorPath = Path.Combine(path, EXE_GROUP_FILE_EXTRACTOR);
if (File.Exists(groupFileExtractorPath))
{
groupConverterExePath = groupFileExtractorPath;
}
else
{
var group2XmlPath = Path.Combine(path, EXE_GROUP2_XML);
if (File.Exists(group2XmlPath))
{
groupConverterExePath = group2XmlPath;
}
else
{
string errorMessage = string.Format(Resources.VendorIssueHelper_ConvertPilotFiles_Unable_to_find__0__or__1__in_directory__2____Please_reinstall_ProteinPilot_software_to_be_able_to_handle__group_files_,
EXE_GROUP_FILE_EXTRACTOR, EXE_GROUP2_XML, path);
throw new IOException(errorMessage);
}
}
}
}
if (groupConverterExePath == null)
{
throw new IOException(Resources.VendorIssueHelper_ConvertPilotFiles_ProteinPilot_software__trial_or_full_version__must_be_installed_to_convert___group__files_to_compatible___group_xml__files_);
}
}
// run group2xml
// ReSharper disable NonLocalizedString
var argv = new[]
{
"XML",
"\"" + inputFile + "\"",
"\"" + outputFile + "\""
};
// ReSharper restore NonLocalizedString
var psi = new ProcessStartInfo(groupConverterExePath)
{
CreateNoWindow = true,
UseShellExecute = false,
// Common directory includes the directory separator
WorkingDirectory = Path.GetDirectoryName(groupConverterExePath) ?? string.Empty,
Arguments = string.Join(" ", argv.ToArray()), // Not L10N
RedirectStandardError = true,
RedirectStandardOutput = true,
};
var sbOut = new StringBuilder();
var proc = new Process {
StartInfo = psi
};
proc.Start();
var reader = new ProcessStreamReader(proc);
string line;
while ((line = reader.ReadLine()) != null)
{
if (progress.IsCanceled)
{
proc.Kill();
throw new LoadCanceledException(status.Cancel());
}
sbOut.AppendLine(line);
}
while (!proc.WaitForExit(200))
{
if (progress.IsCanceled)
{
proc.Kill();
return(inputFilesPilotConverted);
}
}
if (proc.ExitCode != 0)
{
throw new IOException(TextUtil.LineSeparate(string.Format(Resources.VendorIssueHelper_ConvertPilotFiles_Failure_attempting_to_convert_file__0__to__group_xml_,
inputFile), string.Empty, sbOut.ToString()));
}
inputFilesPilotConverted.Add(outputFile);
}
progress.UpdateProgress(status.ChangePercentComplete(100));
return(inputFilesPilotConverted);
}
private static void ConvertBrukerToMzml(string filePathBruker,
string outputPath, IProgressMonitor monitor, ProgressStatus status)
{
// We use CompassXport, if it is installed, to convert a Bruker raw file to mzML. This solves two
// issues: the Bruker reader can't be called on any thread other than the main thread, and there
// is no 64-bit version of the reader. So we start CompassXport in its own 32-bit process,
// and use it to convert the raw data to mzML in a temporary file, which we read back afterwards.
var key = Registry.LocalMachine.OpenSubKey(KEY_COMPASSXPORT, false);
string compassXportExe = (key != null) ? (string)key.GetValue(string.Empty) : null;
if (compassXportExe == null)
{
throw new IOException(Resources.VendorIssueHelper_ConvertBrukerToMzml_CompassXport_software_must_be_installed_to_import_Bruker_raw_data_files_);
}
// CompassXport arguments
// ReSharper disable NonLocalizedString
var argv = new[]
{
"-a \"" + filePathBruker + "\"", // input file (directory)
"-o \"" + outputPath + "\"", // output file (directory)
"-mode 2", // mode 2 (mzML)
"-raw 0" // export line spectra (profile data is HUGE and SLOW!)
};
// ReSharper restore NonLocalizedString
// Start CompassXport in its own process.
var psi = new ProcessStartInfo(compassXportExe)
{
CreateNoWindow = true,
UseShellExecute = false,
// Common directory includes the directory separator
WorkingDirectory = Path.GetDirectoryName(Assembly.GetExecutingAssembly().Location) ?? string.Empty,
Arguments = string.Join(" ", argv), // Not L10N
RedirectStandardError = true,
RedirectStandardOutput = true,
};
var proc = new Process {
StartInfo = psi
};
proc.Start();
// CompassXport starts by calculating a hash of the input file. This takes a long time, and there is
// no intermediate output during this time. So we set the progress bar some fraction of the way and
// let it sit there and animate while we wait for the start of spectra processing.
const int hashPercent = 25; // percentage of import time allocated to calculating the input file hash
int spectrumCount = 0;
var sbOut = new StringBuilder();
var reader = new ProcessStreamReader(proc);
string line;
while ((line = reader.ReadLine()) != null)
{
if (monitor.IsCanceled)
{
proc.Kill();
throw new LoadCanceledException(status.Cancel());
}
sbOut.AppendLine(line);
line = line.Trim();
// The main part of conversion starts with the hash calculation.
if (line.StartsWith("Calculating hash")) // Not L10N
{
status = status.ChangeMessage(Resources.VendorIssueHelper_ConvertBrukerToMzml_Calculating_hash_of_input_file)
.ChangePercentComplete(hashPercent);
monitor.UpdateProgress(status);
continue;
}
// Determine how many spectra will be converted so we can track progress.
var match = Regex.Match(line, @"Converting (\d+) spectra"); // Not L10N
if (match.Success)
{
spectrumCount = int.Parse(match.Groups[1].Value);
continue;
}
// Update progress as each spectra batch is converted.
match = Regex.Match(line, @"Spectrum \d+ - (\d+)"); // Not L10N
if (match.Success)
{
var spectrumEnd = int.Parse(match.Groups[1].Value);
var percentComplete = hashPercent + (100 - hashPercent) * spectrumEnd / spectrumCount;
status = status.ChangeMessage(line).ChangePercentComplete(percentComplete);
monitor.UpdateProgress(status);
}
}
while (!proc.WaitForExit(200))
{
if (monitor.IsCanceled)
{
proc.Kill();
throw new LoadCanceledException(status.Cancel());
}
}
if (proc.ExitCode != 0)
{
throw new IOException(TextUtil.LineSeparate(string.Format(Resources.VendorIssueHelper_ConvertBrukerToMzml_Failure_attempting_to_convert__0__to_mzML_using_CompassXport_,
filePathBruker), string.Empty, sbOut.ToString()));
}
}
protected override bool LoadBackground(IDocumentContainer container, SrmDocument document, SrmDocument docCurrent)
{
// Only allow one background proteome to load at a time. This can
// get tricky, if the user performs an undo and then a redo across
// a change in background proteome.
// Our first priority is doing the digestions, the second is accessing web
// services to add missing protein metadata.
lock (_lockLoadBackgroundProteome)
{
BackgroundProteome originalBackgroundProteome = GetBackgroundProteome(docCurrent);
// Check to see whether the Digestion already exists but has not been queried yet.
BackgroundProteome backgroundProteomeWithDigestions = new BackgroundProteome(originalBackgroundProteome, true);
if (IsNotLoadedExplained(docCurrent, backgroundProteomeWithDigestions, true) == null)
{
// digest is ready, and protein metdata is resolved
CompleteProcessing(container, backgroundProteomeWithDigestions);
return(true);
}
// are we here to do the digest, or to resolve the protein metadata?
bool getMetadata = (IsNotLoadedExplained(docCurrent, backgroundProteomeWithDigestions, false) == null) &&
backgroundProteomeWithDigestions.NeedsProteinMetadataSearch;
string name = originalBackgroundProteome.Name;
ProgressStatus progressStatus =
new ProgressStatus(string.Format(getMetadata?Resources.BackgroundProteomeManager_LoadBackground_Resolving_protein_details_for__0__proteome:Resources.BackgroundProteomeManager_LoadBackground_Digesting__0__proteome, name));
try
{
using (FileSaver fs = new FileSaver(originalBackgroundProteome.DatabasePath, StreamManager))
{
File.Copy(originalBackgroundProteome.DatabasePath, fs.SafeName, true);
var digestHelper = new DigestHelper(this, container, docCurrent, name, fs.SafeName, true);
bool success;
if (getMetadata)
{
success = digestHelper.LookupProteinMetadata(ref progressStatus);
}
else
{
success = (digestHelper.Digest(ref progressStatus) != null);
}
if (!success)
{
// Processing was canceled
EndProcessing(docCurrent);
UpdateProgress(progressStatus.Cancel());
return(false);
}
using (var proteomeDb = ProteomeDb.OpenProteomeDb(originalBackgroundProteome.DatabasePath))
{
proteomeDb.DatabaseLock.AcquireWriterLock(int.MaxValue);
try
{
if (!fs.Commit())
{
EndProcessing(docCurrent);
throw new IOException(
string.Format(
Resources
.BackgroundProteomeManager_LoadBackground_Unable_to_rename_temporary_file_to__0__,
fs.RealName));
}
}
finally
{
proteomeDb.DatabaseLock.ReleaseWriterLock();
}
}
CompleteProcessing(container, new BackgroundProteome(originalBackgroundProteome, true));
UpdateProgress(progressStatus.Complete());
return(true);
}
}
catch (Exception x)
{
var message = new StringBuilder();
message.AppendLine(
string.Format(Resources.BackgroundProteomeManager_LoadBackground_Failed_updating_background_proteome__0__,
name));
message.Append(x.Message);
UpdateProgress(progressStatus.ChangeErrorException(new IOException(message.ToString(), x)));
return(false);
}
}
}
private bool Load(ILoadMonitor loader, ProgressStatus status, bool cached)
{
try
{
int loadPercent = 100;
if (!cached)
{
// Building the cache will take 95% of the load time.
loadPercent = 5;
status = status.ChangeMessage(string.Format(Resources.XHunterLibrary_Load_Building_binary_cache_for__0__library, Path.GetFileName(FilePath)));
status = status.ChangePercentComplete(0);
loader.UpdateProgress(status);
if (!CreateCache(loader, status, 100 - loadPercent))
{
return(false);
}
}
status = status.ChangeMessage(string.Format(Resources.XHunterLibrary_Load_Loading__0__library, Path.GetFileName(FilePath)));
loader.UpdateProgress(status);
var sm = loader.StreamManager;
using (Stream stream = sm.CreateStream(CachePath, FileMode.Open, true))
{
// Read library header from the end of the cache
int countHeader = (int)LibHeaders.count * sizeof(int);
stream.Seek(-countHeader, SeekOrigin.End);
byte[] libHeader = new byte[countHeader];
ReadComplete(stream, libHeader, countHeader);
int version = GetInt32(libHeader, (int)LibHeaders.format_version);
if (version != FORMAT_VERSION_CACHE)
{
return(false);
}
int countRevisionBytes = GetInt32(libHeader, (int)LibHeaders.revision_byte_count);
int countIdBytes = GetInt32(libHeader, (int)LibHeaders.id_byte_count);
stream.Seek(-countHeader - countRevisionBytes - countIdBytes, SeekOrigin.End);
Revision = ReadString(stream, countRevisionBytes);
Id = ReadString(stream, countIdBytes);
int numSpectra = GetInt32(libHeader, (int)LibHeaders.num_spectra);
var setSequences = new Dictionary <LibSeqKey, bool>(numSpectra);
var libraryEntries = new XHunterSpectrumInfo[numSpectra];
// Seek to beginning of spectrum headers
long locationHeaders = BitConverter.ToInt64(libHeader,
((int)LibHeaders.location_headers_lo) * sizeof(int));
stream.Seek(locationHeaders, SeekOrigin.Begin);
byte[] specSequence = new byte[1024];
byte[] specHeader = new byte[1024];
countHeader = (int)SpectrumCacheHeader.count * 4;
for (int i = 0; i < numSpectra; i++)
{
int percent = (100 - loadPercent) + (i * loadPercent / numSpectra);
if (status.PercentComplete != percent)
{
// Check for cancellation after each integer change in percent loaded.
if (loader.IsCanceled)
{
loader.UpdateProgress(status.Cancel());
return(false);
}
// If not cancelled, update progress.
loader.UpdateProgress(status = status.ChangePercentComplete(percent));
}
// Read spectrum header
ReadComplete(stream, specHeader, countHeader);
int seqKeyHash = GetInt32(specHeader, ((int)SpectrumCacheHeader.seq_key_hash));
int seqKeyLength = GetInt32(specHeader, ((int)SpectrumCacheHeader.seq_key_length));
int charge = GetInt32(specHeader, ((int)SpectrumCacheHeader.charge));
if (charge == 0 || charge > TransitionGroup.MAX_PRECURSOR_CHARGE)
{
throw new InvalidDataException(Resources.XHunterLibrary_Load_Invalid_precursor_charge_found_File_may_be_corrupted);
}
float i2 = BitConverter.ToSingle(specHeader, ((int)SpectrumCacheHeader.i2) * 4);
long location = BitConverter.ToInt64(specHeader, ((int)SpectrumCacheHeader.location_lo) * 4);
int numPeaks = GetInt32(specHeader, ((int)SpectrumCacheHeader.num_peaks));
float expect = BitConverter.ToSingle(specHeader, ((int)SpectrumCacheHeader.expect) * 4);
int seqLength = GetInt32(specHeader, (int)SpectrumCacheHeader.seq_len);
// Read sequence information
ReadComplete(stream, specSequence, seqLength);
LibKey key = new LibKey(specSequence, 0, seqLength, charge);
libraryEntries[i] = new XHunterSpectrumInfo(key, i2, expect, (short)numPeaks, location);
if (seqKeyLength > 0)
{
LibSeqKey seqKey = new LibSeqKey(key, seqKeyHash, seqKeyLength);
setSequences.Add(seqKey, true);
}
}
// Checksum = checksum.ChecksumValue;
_libraryEntries = libraryEntries;
_setSequences = setSequences;
loader.UpdateProgress(status.Complete());
// Create the stream from which the spectra will be read
CreateStream(loader);
}
return(true);
}
catch (InvalidDataException x)
{
if (!cached)
{
loader.UpdateProgress(status.ChangeErrorException(x));
}
return(false);
}
catch (IOException x)
{
if (!cached)
{
loader.UpdateProgress(status.ChangeErrorException(x));
}
return(false);
}
catch (Exception x)
{
if (!cached)
{
x = new Exception(string.Format(Resources.XHunterLibrary_Load_Failed_loading_library__0__, FilePath), x);
loader.UpdateProgress(status.ChangeErrorException(x));
}
return(false);
}
finally
{
if (ReadStream != null)
{
// Close the read stream to ensure we never leak it.
// This only costs on extra open, the first time the
// active document tries to read.
try { ReadStream.CloseStream(); }
catch (IOException) { }
}
}
}
// ReSharper restore UnusedMember.Local
private bool CreateCache(ILoadMonitor loader, ProgressStatus status, int percent)
{
var sm = loader.StreamManager;
BufferedStream stream = new BufferedStream(CreateStream(loader), 32 * 1024);
int version = 1;
string id = string.Empty, revision = string.Empty;
int size = ReadSize(stream);
int i;
if (size == 0)
{
version = 2;
size = ReadSize(stream);
const int countLibHeader = 256 - 8;
byte[] libHeader = new byte[countLibHeader];
if (stream.Read(libHeader, 0, libHeader.Length) != libHeader.Length)
{
throw new InvalidDataException(Resources.XHunterLibrary_CreateCache_Data_truncation_in_library_header_File_may_be_corrupted);
}
for (i = 0; i < libHeader.Length; i++)
{
if (libHeader[i] == 0)
{
break;
}
}
string header = Encoding.UTF8.GetString(libHeader, 0, i);
Match match = REGEX_HEADER.Match(header);
if (match.Success)
{
version = int.Parse(match.Groups[1].Value);
id = match.Groups[2].Value;
revision = match.Groups[3].Value;
}
}
var setLibKeys = new Dictionary <LibKey, bool>(size);
var setSequences = new Dictionary <LibSeqKey, bool>(size);
var libraryEntries = new List <XHunterSpectrumInfo>(size);
const int countHeader = ((int)SpectrumHeaders2.count) * sizeof(int);
byte[] specHeader = new byte[1024];
byte[] specSequence = new byte[1024];
i = 0;
while (stream.Read(specHeader, 0, countHeader) == countHeader)
{
int percentComplete = (i++ *percent / size);
if (status.PercentComplete != percentComplete)
{
// Check for cancellation after each integer change in percent loaded.
if (loader.IsCanceled)
{
loader.UpdateProgress(status.Cancel());
return(false);
}
// If not cancelled, update progress.
loader.UpdateProgress(status = status.ChangePercentComplete(percentComplete));
}
int charge = (version == 1
? GetInt32(specHeader, ((int)SpectrumHeaders1.charge))
: GetInt32(specHeader, ((int)SpectrumHeaders2.charge)));
float i2 = (version == 1
? GetSingle(specHeader, ((int)SpectrumHeaders1.i2))
: GetSingle(specHeader, ((int)SpectrumHeaders2.i2)));
int seqLength = (version == 1
? GetInt32(specHeader, ((int)SpectrumHeaders1.seq_len))
: GetInt32(specHeader, ((int)SpectrumHeaders2.seq_len)));
float expect = (version == 1 ? 0.001f : GetSingle(specHeader, (int)SpectrumHeaders2.expect));
// Read sequence information
ReadComplete(stream, specSequence, seqLength);
specSequence[seqLength] = 0;
short numPeaks = (short)ReadSize(stream);
// Save spectrum location
long location = stream.Position;
// Read over spectrum
int countPeaks = (sizeof(byte) + sizeof(float)) * numPeaks;
stream.Seek(countPeaks, SeekOrigin.Current); // Skip spectrum
// Read modifications
int numMods = ReadSize(stream);
byte[] sequence = specSequence;
if (numMods > 0)
{
StringBuilder sb = new StringBuilder();
ReadComplete(stream, specHeader, (4 + sizeof(double)) * numMods);
int iLast = 0;
double modTotal = 0;
for (int j = 0; j < numMods; j++)
{
int iPos = GetInt32(specHeader, j * 3);
double mod = BitConverter.ToDouble(specHeader, (j * 3 + 1) * 4);
// X! Hunter allows multiple modifications on the same
// residue. So, they need to be totaled, and assigned to a
// single residue to allow them to match Skyline modification
// settings.
if (iPos > iLast)
{
if (modTotal != 0)
{
sb.Append(SequenceMassCalc.GetModDiffDescription(modTotal));
}
sb.Append(Encoding.UTF8.GetString(specSequence, iLast, iPos - iLast));
modTotal = 0;
}
modTotal += mod;
iLast = iPos;
}
if (modTotal != 0)
{
sb.Append(SequenceMassCalc.GetModDiffDescription(modTotal));
}
sb.Append(Encoding.UTF8.GetString(specSequence, iLast, seqLength - iLast));
sequence = Encoding.UTF8.GetBytes(sb.ToString());
seqLength = sb.Length;
}
// Skip over homologs (list of protein IDs and start positions from a FASTA
// file used to generate the library)
int numHomologs = ReadSize(stream);
for (int j = 0; j < numHomologs; j++)
{
stream.Seek(ReadSize(stream) + 4, SeekOrigin.Current);
}
// These libraries should not have duplicates, but just in case.
// Apparently, GPM libraries do contain redundancies, as we found
// when a revision lost this test.
var key = new LibKey(sequence, 0, seqLength, charge);
if (!setLibKeys.ContainsKey(key))
{
setLibKeys.Add(key, true);
libraryEntries.Add(new XHunterSpectrumInfo(key, i2, expect, numPeaks, location));
}
}
libraryEntries.Sort(CompareSpectrumInfo);
using (FileSaver fs = new FileSaver(CachePath, sm))
using (Stream outStream = sm.CreateStream(fs.SafeName, FileMode.Create, true))
{
foreach (var info in libraryEntries)
{
LibSeqKey seqKey = new LibSeqKey(info.Key);
if (setSequences.ContainsKey(seqKey))
{
outStream.Write(BitConverter.GetBytes(0), 0, sizeof(int));
outStream.Write(BitConverter.GetBytes(-1), 0, sizeof(int));
}
else
{
// If it is unique, it will need to be added at cache load time.
setSequences.Add(seqKey, true);
outStream.Write(BitConverter.GetBytes(seqKey.GetHashCode()), 0, sizeof(int));
outStream.Write(BitConverter.GetBytes(seqKey.Length), 0, sizeof(int));
}
outStream.Write(BitConverter.GetBytes(info.Key.Charge), 0, sizeof(int));
outStream.Write(BitConverter.GetBytes(info.ProcessedIntensity), 0, sizeof(float));
outStream.Write(BitConverter.GetBytes(info.Location), 0, sizeof(long));
outStream.Write(BitConverter.GetBytes(info.NumPeaks), 0, sizeof(int));
outStream.Write(BitConverter.GetBytes(info.Expect), 0, sizeof(float));
info.Key.WriteSequence(outStream);
}
byte[] revisionBytes = Encoding.UTF8.GetBytes(revision);
outStream.Write(revisionBytes, 0, revisionBytes.Length);
byte[] idBytes = Encoding.UTF8.GetBytes(id);
outStream.Write(idBytes, 0, idBytes.Length);
outStream.Write(BitConverter.GetBytes(revisionBytes.Length), 0, sizeof(int));
outStream.Write(BitConverter.GetBytes(idBytes.Length), 0, sizeof(int));
outStream.Write(BitConverter.GetBytes(FORMAT_VERSION_CACHE), 0, sizeof(int));
outStream.Write(BitConverter.GetBytes(libraryEntries.Count), 0, sizeof(int));
outStream.Write(BitConverter.GetBytes((long)0), 0, sizeof(long));
sm.Finish(outStream);
fs.Commit();
sm.SetCache(FilePath, CachePath);
}
loader.UpdateProgress(status.Complete());
return(true);
}
