NCBI C++ ToolKit
rpstblastn_node.cpp
Go to the documentation of this file.

Go to the SVN repository for this file.

1 /* $Id:
2  * ===========================================================================
3  *
4  * PUBLIC DOMAIN NOTICE
5  * National Center for Biotechnology Information
6  *
7  * This software/database is a "United States Government Work" under the
8  * terms of the United States Copyright Act. It was written as part of
9  * the author's official duties as a United States Government employee and
10  * thus cannot be copyrighted. This software/database is freely available
11  * to the public for use. The National Library of Medicine and the U.S.
12  * Government have not placed any restriction on its use or reproduction.
13  *
14  * Although all reasonable efforts have been taken to ensure the accuracy
15  * and reliability of the software and data, the NLM and the U.S.
16  * Government do not and cannot warrant the performance or results that
17  * may be obtained by using this software or data. The NLM and the U.S.
18  * Government disclaim all warranties, express or implied, including
19  * warranties of performance, merchantability or fitness for any particular
20  * purpose.
21  *
22  * Please cite the author in any work or product based on this material.
23  *
24  * ===========================================================================
25  *
26  * Authors: Amelia Fong
27  *
28  */
29 
30 /** @file rpstblastn_node.cpp
31  * RPSTBLASTN MT command line application
32  */
33 
34 #include <ncbi_pch.hpp>
35 #include <corelib/ncbiapp.hpp>
42 #include "blast_app_util.hpp"
43 #include "rpstblastn_node.hpp"
45 
46 #ifndef SKIP_DOXYGEN_PROCESSING
48 USING_SCOPE(blast);
50 #endif
51 
52 CRPSTBlastnNode::CRPSTBlastnNode (int node_num, const CNcbiArguments & ncbi_args, const CArgs& args,
53  CBlastAppDiagHandler & bah, const string & input,
54  int query_index, int num_queries,CBlastNodeMailbox * mailbox):
55  CBlastNode(node_num, ncbi_args, args, bah, query_index, num_queries, mailbox), m_Input(input)
56 {
59  SendMsg(CBlastNodeMsg::eRunRequest, (void*) this);
60 }
61 
63 {
64  if(GetState() == eDone) {
65  if (m_CmdLineArgs->GetOutputStrStream().rdbuf()->in_avail() > 0) {
66  os << m_CmdLineArgs->GetOutputStrStream().rdbuf();
67  }
68  return GetStatus();
69  }
70  return -1;
71 }
72 
74 {
76 }
77 
78 void *
80 {
81  int status = BLAST_EXIT_SUCCESS;
84 
86  try {
87  const CArgs& args = GetArgs();
88  CRef<CBlastOptionsHandle> opts_hndl;
91  }
92  else {
93  opts_hndl.Reset(&*m_CmdLineArgs->SetOptions(args));
94  }
95 
97  opts_hndl, true);
98  const CBlastOptions& opt = opts_hndl->GetOptions();
99 
100  /*** Initialize the database ***/
102  CRef<CLocalDbAdapter> db_adapter;
103  CRef<CScope> scope;
104  InitializeSubject(db_args, opts_hndl, m_CmdLineArgs->ExecuteRemotely(),
105  db_adapter, scope);
106  _ASSERT(db_adapter && scope);
107 
108  /*** Get the query sequence(s) ***/
109  CRef<CQueryOptionsArgs> query_opts =
111  SDataLoaderConfig dlconfig =
113  db_adapter);
114  CBlastInputSourceConfig iconfig(dlconfig, query_opts->GetStrand(),
115  query_opts->UseLowercaseMasks(),
116  query_opts->GetParseDeflines(),
117  query_opts->GetRange());
120 
121  /*** Get the formatting options ***/
123  bool isArchiveFormat = fmt_args->ArchiveFormatRequested(args);
124  if(!isArchiveFormat) {
125  bah.DoNotSaveMessages();
126  }
127  CBlastFormat formatter(opt, *db_adapter,
128  fmt_args->GetFormattedOutputChoice(),
129  query_opts->GetParseDeflines(),
131  fmt_args->GetNumDescriptions(),
132  fmt_args->GetNumAlignments(),
133  *scope,
134  opt.GetMatrixName(),
135  fmt_args->ShowGis(),
136  fmt_args->DisplayHtmlOutput(),
137  opt.GetQueryGeneticCode(),
138  opt.GetDbGeneticCode(),
139  opt.GetSumStatisticsMode(),
141  db_adapter->GetFilteringAlgorithm(),
142  fmt_args->GetCustomOutputFormatSpec(),
143  false, false, NULL, NULL,
145 
146  formatter.SetQueryRange(query_opts->GetRange());
147  formatter.SetLineLength(fmt_args->GetLineLength());
148  if(UseXInclude(*fmt_args, args[kArgOutput].AsString())) {
149  formatter.SetBaseFile(args[kArgOutput].AsString() + "_" + NStr::IntToString(GetNodeNum()));
150  }
151  formatter.PrintProlog();
152 
153  /*** Process the input ***/
154  for (; !input.End(); formatter.ResetScopeHistory(), QueryBatchCleanup()) {
155 
156  CRef<CBlastQueryVector> query_batch(input.GetNextSeqBatch(*scope));
157  CRef<IQueryFactory> queries(new CObjMgr_QueryFactory(*query_batch));
158 
159  SaveSearchStrategy(args, m_CmdLineArgs, queries, opts_hndl);
160 
161  CRef<CSearchResultSet> results;
162 
164  {
165  CRef<CRemoteBlast> rmt_blast =
166  InitializeRemoteBlast(queries, db_args, opts_hndl,
169  results = rmt_blast->GetResultSet();
170  }
171  else
172  {
173  CLocalRPSBlast local_search (query_batch, db_args->GetDatabaseName(), opts_hndl, 1);
174  results = local_search.Run();
175  }
176 
177  if (fmt_args->ArchiveFormatRequested(args)) {
178  formatter.WriteArchive(*queries, *opts_hndl, *results, 0, bah.GetMessages());
179  bah.ResetMessages();
180  } else {
181  BlastFormatter_PreFetchSequenceData(*results, scope,
182  fmt_args->GetFormattedOutputChoice());
183  ITERATE(CSearchResultSet, result, *results) {
184  formatter.PrintOneResultSet(**result, query_batch);
185  }
186  }
187  }
188 
189  formatter.PrintEpilog(opt);
190 
192  opts_hndl->GetOptions().DebugDumpText(NcbiCerr, "BLAST options", 1);
193  }
194 
195  SetQueriesLength(input.GetTotalLengthProcessed());
196  } CATCH_ALL(status)
197 
198  SetStatus(status);
199  if (status == BLAST_EXIT_SUCCESS) {
200  SetState(eDone);
201  SendMsg(CBlastNodeMsg::ePostResult, (void *) this);
202 
203  }
204  else {
205  SetState(eError);
206  SendMsg(CBlastNodeMsg::eErrorExit, (void *) this);
207 
208  }
209 
210  return NULL;
211 }
Produce formatted blast output for command line applications.
void CheckForFreqRatioFile(const string &rps_dbname, CRef< CBlastOptionsHandle > &opt_handle, bool isRpsblast)
CRef< blast::CRemoteBlast > InitializeRemoteBlast(CRef< blast::IQueryFactory > queries, CRef< blast::CBlastDatabaseArgs > db_args, CRef< blast::CBlastOptionsHandle > opts_hndl, bool verbose_output, const string &client_id, CRef< objects::CPssmWithParameters > pssm)
Initializes a CRemoteBlast instance for usage by command line BLAST binaries.
blast::SDataLoaderConfig InitializeQueryDataLoaderConfiguration(bool query_is_protein, CRef< blast::CLocalDbAdapter > db_adapter)
Initialize the data loader configuration for the query.
void SaveSearchStrategy(const CArgs &args, blast::CBlastAppArgs *cmdline_args, CRef< blast::IQueryFactory > queries, CRef< blast::CBlastOptionsHandle > opts_hndl, CRef< objects::CPssmWithParameters > pssm, unsigned int num_iters)
Save the search strategy corresponding to the current command line search.
void QueryBatchCleanup()
Clean up formatter scope and release.
bool RecoverSearchStrategy(const CArgs &args, blast::CBlastAppArgs *cmdline_args)
Recover search strategy from input file.
void InitializeSubject(CRef< blast::CBlastDatabaseArgs > db_args, CRef< blast::CBlastOptionsHandle > opts_hndl, bool is_remote_search, CRef< blast::CLocalDbAdapter > &db_adapter, CRef< objects::CScope > &scope)
Initializes the subject/database as well as its scope.
string GetCmdlineArgs(const CNcbiArguments &a)
void BlastFormatter_PreFetchSequenceData(const blast::CSearchResultSet &results, CRef< CScope > scope, blast::CFormattingArgs::EOutputFormat format_type)
This method optimize the retrieval of sequence data to scope.
bool UseXInclude(const CFormattingArgs &f, const string &s)
Utility functions for BLAST command line applications.
#define CATCH_ALL(exit_code)
Standard catch statement for all BLAST command line programs.
#define BLAST_EXIT_SUCCESS
Command line binary exit code: success.
Interface for reading SRA sequences into blast input.
CArgs –.
Definition: ncbiargs.hpp:379
CRef< CBlastOptionsHandle > SetOptionsForSavedStrategy(const CArgs &args)
Combine the command line arguments into a CBlastOptions object recovered from saved search strategy.
CRef< CBlastOptionsHandle > SetOptions(const CArgs &args)
Extract the command line arguments into a CBlastOptionsHandle object.
CRef< CBlastDatabaseArgs > GetBlastDatabaseArgs() const
Get the BLAST database arguments.
bool ExecuteRemotely() const
Determine whether the search should be executed remotely or not.
bool ProduceDebugRemoteOutput() const
Return whether debug (verbose) output should be produced on remote searches (only available when comp...
CRef< CQueryOptionsArgs > GetQueryOptionsArgs() const
Get the options for the query sequence(s)
string GetClientId() const
Retrieve the client ID for remote requests.
CRef< CFormattingArgs > GetFormattingArgs() const
Get the formatting options.
bool ProduceDebugOutput() const
Return whether debug (verbose) output should be produced on remote searches (only available when comp...
Class to capture message from diag handler.
Definition: blast_aux.hpp:249
string GetDatabaseName() const
Get the BLAST database name.
Definition: blast_args.hpp:931
Class representing a text file containing sequences in fasta format.
This class formats the BLAST results for command line applications.
void PrintOneResultSet(const blast::CSearchResults &results, CConstRef< blast::CBlastQueryVector > queries, unsigned int itr_num=numeric_limits< unsigned int >::max(), blast::CPsiBlastIterationState::TSeqIds prev_seqids=blast::CPsiBlastIterationState::TSeqIds(), bool is_deltablast_domain_result=false)
Print all alignment information for a single query sequence along with any errors or warnings (errors...
void PrintEpilog(const blast::CBlastOptions &options)
Print the footer of the blast report.
void SetBaseFile(string base)
For use by XML2 only.
void ResetScopeHistory()
Resets the scope history for some output formats.
void SetLineLength(size_t len)
Set Alignment Length.
void WriteArchive(blast::IQueryFactory &queries, blast::CBlastOptionsHandle &options_handle, const blast::CSearchResultSet &results, unsigned int num_iters=0, const list< CRef< objects::CBlast4_error > > &msg=list< CRef< objects::CBlast4_error > >())
Writes out the query and results as an "archive" format.
void PrintProlog()
Print the header of the blast report.
void SetQueryRange(const TSeqRange &query_range)
Set query range.
Class that centralizes the configuration data for sequences to be converted.
Definition: blast_input.hpp:48
Generalized converter from an abstract source of biological sequence data to collections of blast inp...
EState GetState()
Definition: blast_node.hpp:100
void SendMsg(CBlastNodeMsg::EMsgType msg_type, void *ptr=NULL)
Definition: blast_node.cpp:103
const CArgs & GetArgs()
Definition: blast_node.hpp:102
CBlastAppDiagHandler & GetDiagHandler()
Definition: blast_node.hpp:103
int GetNodeNum()
Definition: blast_node.hpp:99
int GetStatus()
Definition: blast_node.hpp:101
const CNcbiArguments & GetArguments()
Definition: blast_node.hpp:104
void SetState(EState state)
Definition: blast_node.hpp:112
void SetQueriesLength(int l)
Definition: blast_node.hpp:114
void SetStatus(int status)
Definition: blast_node.hpp:113
string & GetNodeIdStr()
Definition: blast_node.hpp:106
Encapsulates ALL the BLAST algorithm's options.
void DebugDumpText(ostream &out, const string &bundle, unsigned int depth) const
Definition: ddumpable.cpp:56
virtual bool ArchiveFormatRequested(const CArgs &args) const
string GetCustomOutputFormatSpec() const
Retrieve for string that specifies the custom output format for tabular and comma-separated value.
EOutputFormat GetFormattedOutputChoice() const
Get the choice of formatted output.
TSeqPos GetNumAlignments() const
Number of alignments to show in traditional BLAST output.
bool ShowGis() const
Display the NCBI GIs in formatted output?
TSeqPos GetNumDescriptions() const
Number of one-line descriptions to show in traditional BLAST output.
size_t GetLineLength() const
bool DisplayHtmlOutput() const
Display HTML output?
CRef< CSearchResultSet > Run()
CNcbiArguments –.
Definition: ncbienv.hpp:236
NCBI C++ Object Manager dependant implementation of IQueryFactory.
objects::ENa_strand GetStrand() const
Get strand to search in query sequence(s)
Definition: blast_args.hpp:800
bool GetParseDeflines() const
Should the defline be parsed?
Definition: blast_args.hpp:804
bool QueryIsProtein() const
Is the query sequence protein?
Definition: blast_args.hpp:807
TSeqRange GetRange() const
Get query sequence range restriction.
Definition: blast_args.hpp:796
bool UseLowercaseMasks() const
Use lowercase masking in FASTA input?
Definition: blast_args.hpp:802
virtual CNcbiOstream & GetOutputStream()
Get the output stream.
virtual CNcbiIstream & GetInputStream()
Get the input stream.
CNcbiStrstream & GetOutputStrStream()
virtual int GetQueryBatchSize() const
@inheritDoc
CRPSTBlastnNode(int check_num, const CNcbiArguments &ncbi_args, const CArgs &args, CBlastAppDiagHandler &bah, const string &input, int query_index, int num_queries, CBlastNodeMailbox *mailbox=NULL)
virtual void * Main(void)
Derived (user-created) class must provide a real thread function.
virtual ~CRPSTBlastnNode(void)
CRef< CRPSTBlastnNodeArgs > m_CmdLineArgs
virtual int GetBlastResults(CNcbiOstream &os)
Search Results for All Queries.
const string kArgOutput
Output file name.
int GetDbGeneticCode() const
void ResetMessages(void)
Reset messgae buffer, erase all saved message.
Definition: blast_aux.cpp:1174
int GetFilteringAlgorithm()
Retrieve the database filtering algorithm.
const CBlastOptions & GetOptions() const
Return the object which this object is a handle for.
int GetQueryGeneticCode() const
bool GetSumStatisticsMode() const
Sum statistics options.
CRef< CSearchResultSet > GetResultSet()
Submit the search (if necessary) and return the results.
void DoNotSaveMessages(void)
Call to turn off saving diag message, discard all saved message.
Definition: blast_aux.cpp:1189
const char * GetMatrixName() const
list< CRef< objects::CBlast4_error > > & GetMessages(void)
Return list of saved diag messages.
Definition: blast_aux.hpp:262
#define ITERATE(Type, Var, Cont)
ITERATE macro to sequence through container elements.
Definition: ncbimisc.hpp:815
#define NULL
Definition: ncbistd.hpp:225
void SetDiagPostPrefix(const char *prefix)
Specify a string to prefix all subsequent error postings with.
Definition: ncbidiag.cpp:6097
void Reset(void)
Reset reference object.
Definition: ncbiobj.hpp:773
IO_PREFIX::ostream CNcbiOstream
Portable alias for ostream.
Definition: ncbistre.hpp:149
#define NcbiCerr
Definition: ncbistre.hpp:544
static string IntToString(int value, TNumToStringFlags flags=0, int base=10)
Convert int to string.
Definition: ncbistr.hpp:5084
static int input()
Main class to perform a BLAST search on the local machine.
Defines the CNcbiApplication and CAppException classes for creating NCBI applications.
NOTE: This file contains work in progress and the APIs are likely to change, please do not rely on th...
Declares the CRemoteBlast class.
Declares the CLocalRPSBlast class.
Main argument class for RPS TBLASTN application.
USING_SCOPE(blast)
USING_NCBI_SCOPE
RPSTBLASTN node api.
Configuration structure for the CBlastScopeSource.
#define _ASSERT
else result
Definition: token2.c:20
Modified on Sat Apr 13 11:46:49 2024 by modify_doxy.py rev. 669887