/// <summary>
/// Parses a single biological sequence alignment text from a stream.
/// </summary>
/// <param name="reader">Reader</param>
/// <returns>Sequence</returns>
private ISequenceAlignment ParseOne(StreamReader reader)
{
// no empty files allowed
if (line == null)
ReadNextLine(reader);
if (line == null)
throw new InvalidDataException(Properties.Resource.IONoTextToParse);
if (!line.StartsWith("CLUSTAL", StringComparison.OrdinalIgnoreCase))
{
throw new InvalidDataException(
string.Format(CultureInfo.CurrentCulture, Properties.Resource.INVALID_INPUT_FILE, this.Name));
}
ReadNextLine(reader); // Skip blank lines until we get to the first block.
// Now that we're at the first block, one or more blank lines are the block separators, which we'll need.
skipBlankLines = false;
var mapIdToSequence = new Dictionary<string, Tuple<ISequence, List<byte>>>();
IAlphabet alignmentAlphabet = null;
bool isFirstBlock = true;
bool inBlock = false;
var endOfBlockSymbols = new HashSet<char> { '*', ' ', '.', '+', ':' };
while (reader.Peek() != -1)
{
// Blank line or consensus line signals end of block.
if (String.IsNullOrEmpty(line) || line.ToCharArray().All(endOfBlockSymbols.Contains))
{
if (inBlock)
{
// Blank line signifies end of block
inBlock = false;
isFirstBlock = false;
}
}
else // It's not a blank or consensus line.
{
// It's a data line in a block.
// Lines begin with sequence id, then the sequence segment, and optionally a number, which we will ignore
string[] tokens = line.Split((char[])null, StringSplitOptions.RemoveEmptyEntries); // (char[])null uses whitespace delimiters
string id = tokens[0];
string data = tokens[1].ToUpperInvariant();
byte[] byteData = Encoding.UTF8.GetBytes(data);
Tuple<ISequence, List<byte>> sequenceTuple;
IAlphabet alphabet = Alphabet;
inBlock = true;
if (isFirstBlock)
{
if (null == alphabet)
{
alphabet = Alphabets.AutoDetectAlphabet(byteData, 0, byteData.Length, alphabet);
if (null == alphabet)
{
throw new InvalidDataException(string.Format(
CultureInfo.InvariantCulture,
Properties.Resource.InvalidSymbolInString,
data));
}
if (null == alignmentAlphabet)
{
alignmentAlphabet = alphabet;
}
else
{
if (alignmentAlphabet != alphabet)
{
throw new InvalidDataException(string.Format(
CultureInfo.CurrentCulture,
Properties.Resource.SequenceAlphabetMismatch));
}
}
}
sequenceTuple = new Tuple<ISequence, List<byte>>(
new Sequence(alphabet, "") { ID = id },
new List<byte>());
sequenceTuple.Item2.AddRange(byteData);
mapIdToSequence.Add(id, sequenceTuple);
}
else
{
if (!mapIdToSequence.ContainsKey(id))
{
throw new InvalidDataException(string.Format(CultureInfo.CurrentCulture, Properties.Resource.ClustalUnknownSequence, id));
}
sequenceTuple = mapIdToSequence[id];
sequenceTuple.Item2.AddRange(byteData);
}
}
ReadNextLine(reader);
}
var sequenceAlignment = new SequenceAlignment();
var alignedSequence = new AlignedSequence();
sequenceAlignment.AlignedSequences.Add(alignedSequence);
foreach (var alignmentSequenceTuple in mapIdToSequence.Values)
{
alignedSequence.Sequences.Add(
new Sequence(alignmentSequenceTuple.Item1.Alphabet, alignmentSequenceTuple.Item2.ToArray())
{
ID = alignmentSequenceTuple.Item1.ID
});
}
return sequenceAlignment;
}
/// <summary>
/// Parses a single biological sequence alignment text from a reader.
/// </summary>
/// <param name="reader">A reader for a biological sequence alignment text.</param>
/// <returns>The parsed ISequenceAlignment object.</returns>
ISequenceAlignment ParseOne(TextReader reader)
{
ReadNextLine(reader);
if (line == null)
{
throw new Exception(Properties.Resource.INVALID_INPUT_FILE);
}
this.ParseHeader(reader);
var alignedSequence = new AlignedSequence();
IList<string> ids = null;
bool isInBlock = true;
if (this.line.StartsWith("begin", StringComparison.OrdinalIgnoreCase))
{
while (this.line != null && isInBlock)
{
if (string.IsNullOrEmpty(this.line.Trim()))
{
this.ReadNextLine(reader);
continue;
}
string blockName = GetTokens(this.line)[1];
switch (blockName.ToUpperInvariant())
{
case "TAXA":
case "TAXA;":
// This block contains the count of sequence & title of each sequence
ids = this.ParseTaxaBlock(reader);
break;
case "CHARACTERS":
case "CHARACTERS;":
// Block contains sequences
Dictionary<string, string> dataSet = this.ParseCharacterBlock(reader, ids);
IAlphabet alignmentAlphabet = null;
foreach (string id in ids)
{
IAlphabet alphabet = this.Alphabet;
string data = dataSet[id];
if (null == alphabet)
{
byte[] dataArray = data.ToByteArray();
alphabet = Alphabets.AutoDetectAlphabet(dataArray, 0, dataArray.Length, null);
if (null == alphabet)
{
throw new InvalidDataException(string.Format(
CultureInfo.InvariantCulture,
Properties.Resource.InvalidSymbolInString,
data));
}
if (null == alignmentAlphabet)
{
alignmentAlphabet = alphabet;
}
else
{
if (alignmentAlphabet != alphabet)
{
throw new InvalidDataException(string.Format(
CultureInfo.InvariantCulture,
Properties.Resource.SequenceAlphabetMismatch));
}
}
}
alignedSequence.Sequences.Add(new Sequence(alphabet, data) { ID = id });
}
break;
case "END":
case "END;":
// Have reached the end of block
isInBlock = false;
break;
default:
// skip this block
while (this.line != null)
{
this.ReadNextLine(reader);
if (0 == string.Compare(this.line, "end;", StringComparison.OrdinalIgnoreCase))
{
break;
}
}
break;
}
this.ReadNextLine(reader);
}
}
ISequenceAlignment sequenceAlignment = new SequenceAlignment();
sequenceAlignment.AlignedSequences.Add(alignedSequence);
return sequenceAlignment;
}
/// <summary>
/// Validate aligned sequence instance using different aligners
/// </summary>
/// <param name="nodeName">xml node name</param>
/// <param name="aligner">sw/nw/pw aligners</param>
private void ValidateAlignedSequenceCtor(string nodeName, ISequenceAligner aligner)
{
IAlphabet alphabet = Utility.GetAlphabet(this.utilityObj.xmlUtil.GetTextValue(nodeName,
Constants.AlphabetNameNode));
string origSequence1 = this.utilityObj.xmlUtil.GetTextValue(nodeName, Constants.SequenceNode1);
string origSequence2 = this.utilityObj.xmlUtil.GetTextValue(nodeName, Constants.SequenceNode2);
// Create input sequences
var inputSequences = new List<ISequence>();
inputSequences.Add(new Sequence(alphabet, origSequence1));
inputSequences.Add(new Sequence(alphabet, origSequence2));
// Get aligned sequences
IAlignedSequence alignedSequence = new AlignedSequence();
IList<ISequenceAlignment> alignment = aligner.Align(inputSequences);
// add aligned sequence and metadata information
for (int iseq = 0; iseq < alignment[0].AlignedSequences[0].Sequences.Count; iseq++)
{
alignedSequence.Sequences.Add(alignment[0].AlignedSequences[0].Sequences[iseq]);
}
foreach (string key in alignment[0].AlignedSequences[0].Metadata.Keys)
{
alignedSequence.Metadata.Add(key, alignment[0].AlignedSequences[0].Metadata[key]);
}
// Validate the alignedsequence properties
for (int index = 0; index < alignment[0].AlignedSequences[0].Sequences.Count; index++)
{
Assert.AreEqual(alignment[0].AlignedSequences[0].Sequences[index].ToString(),
alignedSequence.Sequences[index].ToString());
}
foreach (string key in alignment[0].AlignedSequences[0].Metadata.Keys)
{
Assert.AreEqual(alignment[0].AlignedSequences[0].Metadata[key],
alignedSequence.Metadata[key]);
}
ApplicationLog.WriteLine(@"Alignment BVT : Validation of aligned sequence ctor completed successfully");
}
