merge master 2.3.0

[bamtools.git] / src / toolkit / bamtools_random.cpp

diff --git a/src/toolkit/bamtools_random.cpp b/src/toolkit/bamtools_random.cpp
index e28ea70962088578ddb27989c25f996925185f11..367ac58a475882862c3a678200027b9fb30b1a35 100644 (file)
--- a/src/toolkit/bamtools_random.cpp
+++ b/src/toolkit/bamtools_random.cpp
@@ -2,7 +2,7 @@
 // bamtools_random.cpp (c) 2010 Derek Barnett, Erik Garrison
 // Marth Lab, Department of Biology, Boston College
 // ---------------------------------------------------------------------------
-// Last modified: 7 April 2011 (DB)
+// Last modified: 24 July 2013 (DB)
 // ---------------------------------------------------------------------------
 // Grab a random subset of alignments (testing tool)
 // ***************************************************************************
@@ -17,6 +17,7 @@ using namespace BamTools;
 
 #include <ctime>
 #include <cstdlib>
+#include <fstream>
 #include <iostream>
 #include <string>
 #include <vector>
@@ -43,25 +44,32 @@ struct RandomTool::RandomSettings {
     // flags
     bool HasAlignmentCount;
     bool HasInput;
+    bool HasInputFilelist;
     bool HasOutput;
+    bool HasRandomNumberSeed;
     bool HasRegion;
     bool IsForceCompression;
 
     // parameters
     unsigned int AlignmentCount;
     vector<string> InputFiles;
+    string InputFilelist;
     string OutputFilename;
+    unsigned int RandomNumberSeed;
     string Region;
     
     // constructor
     RandomSettings(void)
         : HasAlignmentCount(false)
         , HasInput(false)
+        , HasInputFilelist(false)
         , HasOutput(false)
+        , HasRandomNumberSeed(false)
         , HasRegion(false)
         , IsForceCompression(false)
         , AlignmentCount(RANDOM_MAX_ALIGNMENT_COUNT)
         , OutputFilename(Options::StandardOut())
+        , RandomNumberSeed(0)
     { }  
 };  
 
@@ -90,9 +98,23 @@ struct RandomTool::RandomToolPrivate {
 bool RandomTool::RandomToolPrivate::Run(void) {
 
     // set to default stdin if no input files provided
-    if ( !m_settings->HasInput )
+    if ( !m_settings->HasInput && !m_settings->HasInputFilelist )
         m_settings->InputFiles.push_back(Options::StandardIn());
 
+    // add files in the filelist to the input file list
+    if ( m_settings->HasInputFilelist ) {
+
+        ifstream filelist(m_settings->InputFilelist.c_str(), ios::in);
+        if ( !filelist.is_open() ) {
+            cerr << "bamtools random ERROR: could not open input BAM file list... Aborting." << endl;
+            return false;
+        }
+
+        string line;
+        while ( getline(filelist, line) )
+            m_settings->InputFiles.push_back(line);
+    }
+
     // open our reader
     BamMultiReader reader;
     if ( !reader.Open(m_settings->InputFiles) ) {
@@ -147,7 +169,10 @@ bool RandomTool::RandomToolPrivate::Run(void) {
     }
 
     // seed our random number generator
-    srand( time(NULL) );
+    if ( m_settings->HasRandomNumberSeed )
+        srand( m_settings->RandomNumberSeed );
+    else
+        srand( time(NULL) );
 
     // grab random alignments
     BamAlignment al;
@@ -212,17 +237,22 @@ RandomTool::RandomTool(void)
     , m_impl(0)
 { 
     // set program details
-    Options::SetProgramInfo("bamtools random", "grab a random subset of alignments", "[-in <filename> -in <filename> ...] [-out <filename>] [-forceCompression] [-n] [-region <REGION>]");
+    Options::SetProgramInfo("bamtools random", "grab a random subset of alignments",
+                            "[-in <filename> -in <filename> ... | -list <filelist>] [-out <filename>] [-forceCompression] [-n] [-region <REGION>]");
     
     // set up options 
     OptionGroup* IO_Opts = Options::CreateOptionGroup("Input & Output");
-    Options::AddValueOption("-in",  "BAM filename", "the input BAM file",  "", m_settings->HasInput,  m_settings->InputFiles,     IO_Opts, Options::StandardIn());
-    Options::AddValueOption("-out", "BAM filename", "the output BAM file", "", m_settings->HasOutput, m_settings->OutputFilename, IO_Opts, Options::StandardOut());
+    Options::AddValueOption("-in",     "BAM filename", "the input BAM file",                         "", m_settings->HasInput,          m_settings->InputFiles,     IO_Opts, Options::StandardIn());
+    Options::AddValueOption("-list",   "filename",     "the input BAM file list, one line per file", "", m_settings->HasInputFilelist,  m_settings->InputFilelist,  IO_Opts);
+    Options::AddValueOption("-out",    "BAM filename", "the output BAM file",                        "", m_settings->HasOutput,         m_settings->OutputFilename, IO_Opts, Options::StandardOut());
+    Options::AddValueOption("-region", "REGION",       "only pull random alignments from within this genomic region. Index file is recommended for better performance, and is used automatically if it exists. See \'bamtools help index\' for more details on creating one", "", m_settings->HasRegion, m_settings->Region, IO_Opts);
     Options::AddOption("-forceCompression", "if results are sent to stdout (like when piping to another tool), default behavior is to leave output uncompressed. Use this flag to override and force compression", m_settings->IsForceCompression, IO_Opts);
-    Options::AddValueOption("-region", "REGION", "only pull random alignments from within this genomic region. Index file is recommended for better performance, and is used automatically if it exists. See \'bamtools help index\' for more details on creating one", "", m_settings->HasRegion, m_settings->Region, IO_Opts);
     
     OptionGroup* SettingsOpts = Options::CreateOptionGroup("Settings");
-    Options::AddValueOption("-n", "count", "number of alignments to grab. Note - no duplicate checking is performed", "", m_settings->HasAlignmentCount, m_settings->AlignmentCount, SettingsOpts, RANDOM_MAX_ALIGNMENT_COUNT);
+    Options::AddValueOption("-n", "count", "number of alignments to grab. Note - no duplicate checking is performed", "",
+                            m_settings->HasAlignmentCount, m_settings->AlignmentCount, SettingsOpts, RANDOM_MAX_ALIGNMENT_COUNT);
+    Options::AddValueOption("-seed", "unsigned integer", "random number generator seed (for repeatable results). Current time is used if no seed value is provided.", "",
+                            m_settings->HasRandomNumberSeed, m_settings->RandomNumberSeed, SettingsOpts);
 }
 
 RandomTool::~RandomTool(void) {