NCBI C++ ToolKit
blastp_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 blastp_node.cpp
31  * blastp node api
32  */
33 
34 #include <ncbi_pch.hpp>
35 #include <corelib/ncbiapp.hpp>
41 #include "blast_app_util.hpp"
42 #include "blastp_node.hpp"
43 
44 #ifndef SKIP_DOXYGEN_PROCESSING
46 USING_SCOPE(blast);
48 #endif
49 
50 CBlastpNode::CBlastpNode (int node_num, const CNcbiArguments & ncbi_args, const CArgs& args,
51  CBlastAppDiagHandler & bah, string & input,
52  int query_index, int num_queries, CBlastNodeMailbox * mailbox):
53  CBlastNode(node_num, ncbi_args, args, bah, query_index, num_queries, mailbox), m_Input(kEmptyStr)
54 {
55  m_Input.swap(input);
58  SendMsg(CBlastNodeMsg::eRunRequest, (void*) this);
59 }
60 
62 {
63  if(GetState() == eDone) {
64  if (m_CmdLineArgs->GetOutputStrStream().rdbuf()->in_avail() > 0) {
65  os << m_CmdLineArgs->GetOutputStrStream().rdbuf();
66  }
67  return GetStatus();
68  }
69  return -1;
70 }
71 
73 {
75 }
76 
77 void *
79 {
80  int status = BLAST_EXIT_SUCCESS;
83 
86  try {
87  /*** Get the BLAST options ***/
88  const CArgs& args = GetArgs();
89  CRef<CBlastOptionsHandle> opts_hndl;
92  }
93  else {
94  opts_hndl.Reset(&*m_CmdLineArgs->SetOptions(args));
95  }
96 
97  const CBlastOptions& opt = opts_hndl->GetOptions();
98 
99  /*** Initialize the database/subject ***/
101  CRef<CLocalDbAdapter> db_adapter;
102  CRef<CScope> scope;
103  InitializeSubject(db_args, opts_hndl, m_CmdLineArgs->ExecuteRemotely(), db_adapter, scope);
104  _ASSERT(db_adapter && scope);
105 
106  /*** Get the query sequence(s) ***/
107  CRef<CQueryOptionsArgs> query_opts =
109  SDataLoaderConfig dlconfig =
111  db_adapter);
112  CBlastInputSourceConfig iconfig(dlconfig, query_opts->GetStrand(),
113  query_opts->UseLowercaseMasks(),
114  query_opts->GetParseDeflines(),
115  query_opts->GetRange());
117  CBlastInput input(&fasta);
118 
119 
120  /*** Get the formatting options ***/
122  bool isArchiveFormat = fmt_args->ArchiveFormatRequested(args);
123  if(!isArchiveFormat) {
124  bah.DoNotSaveMessages();
125  }
126  CBlastFormat formatter(opt, *db_adapter,
127  fmt_args->GetFormattedOutputChoice(),
128  query_opts->GetParseDeflines(),
130  fmt_args->GetNumDescriptions(),
131  fmt_args->GetNumAlignments(),
132  *scope,
133  opt.GetMatrixName(),
134  fmt_args->ShowGis(),
135  fmt_args->DisplayHtmlOutput(),
136  opt.GetQueryGeneticCode(),
137  opt.GetDbGeneticCode(),
138  opt.GetSumStatisticsMode(),
140  db_adapter->GetFilteringAlgorithm(),
141  fmt_args->GetCustomOutputFormatSpec(),
142  false, false, NULL, NULL,
144  GetSubjectFile(args));
145 
146  formatter.SetQueryRange(query_opts->GetRange());
147  formatter.SetLineLength(fmt_args->GetLineLength());
148  formatter.SetHitsSortOption(fmt_args->GetHitsSortOption());
149  formatter.SetHspsSortOption(fmt_args->GetHspsSortOption());
150  formatter.SetCustomDelimiter(fmt_args->GetCustomDelimiter());
151  if(UseXInclude(*fmt_args, args[kArgOutput].AsString())) {
152  formatter.SetBaseFile(args[kArgOutput].AsString() + "_" + NStr::IntToString(GetNodeNum()));
153  }
154  formatter.PrintProlog();
155 
156  /*** Process the input ***/
157  for (; !input.End(); formatter.ResetScopeHistory(), QueryBatchCleanup()) {
158 
159  CRef<CBlastQueryVector> query_batch(input.GetNextSeqBatch(*scope));
160  CRef<IQueryFactory> queries(new CObjMgr_QueryFactory(*query_batch));
161 
162  SaveSearchStrategy(args, m_CmdLineArgs, queries, opts_hndl);
163 
164  CRef<CSearchResultSet> results;
165 
167  CRef<CRemoteBlast> rmt_blast =
168  InitializeRemoteBlast(queries, db_args, opts_hndl,
171  results = rmt_blast->GetResultSet();
172  } else {
173  CLocalBlast lcl_blast(queries, opts_hndl, db_adapter);
174  lcl_blast.SetNumberOfThreads(1);
175  results = lcl_blast.Run();
176  }
177 
178  if (fmt_args->ArchiveFormatRequested(args)) {
179  formatter.WriteArchive(*queries, *opts_hndl, *results, 0, bah.GetMessages());
180  bah.ResetMessages();
181  } else {
182  BlastFormatter_PreFetchSequenceData(*results, scope,
183  fmt_args->GetFormattedOutputChoice());
184  ITERATE(CSearchResultSet, result, *results) {
185  formatter.PrintOneResultSet(**result, query_batch);
186  }
187  }
188  }
189 
190  formatter.PrintEpilog(opt);
191 
193  opts_hndl->GetOptions().DebugDumpText(NcbiCerr, "BLAST options", 1);
194  }
195 
196  SetQueriesLength(input.GetTotalLengthProcessed());
197 
198  } CATCH_ALL(status)
199 
200  SetStatus(status);
201  if (status == BLAST_EXIT_SUCCESS) {
202  SetState(eDone);
203  SendMsg(CBlastNodeMsg::ePostResult, (void *) this);
204 
205  }
206  else {
207  SetState(eError);
208  SendMsg(CBlastNodeMsg::eErrorExit, (void *) this);
209  }
210 
211  return NULL;
212 }
213 
Produce formatted blast output for command line applications.
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.
string GetSubjectFile(const CArgs &args)
Get name of subject file @parameter args arguments class [in].
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.
USING_SCOPE(blast)
USING_NCBI_SCOPE
Definition: blastp_node.cpp:45
blastn node api
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
Class representing a text file containing sequences in fasta format.
This class formats the BLAST results for command line applications.
void SetHitsSortOption(int hitsSortOption)
void SetHspsSortOption(int hspsSortOption)
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 SetCustomDelimiter(string customDelim)
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
void SetDataLoaderPrefix()
Definition: blast_node.cpp:112
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.
virtual CNcbiOstream & GetOutputStream()
Get the output stream.
CNcbiStrstream & GetOutputStrStream()
Definition: blastp_args.hpp:75
virtual CNcbiIstream & GetInputStream()
Get the input stream.
virtual ~CBlastpNode(void)
Definition: blastp_node.cpp:72
virtual void * Main(void)
Derived (user-created) class must provide a real thread function.
Definition: blastp_node.cpp:78
CRef< CBlastpNodeArgs > m_CmdLineArgs
Definition: blastp_node.hpp:56
string m_Input
Definition: blastp_node.hpp:55
virtual int GetBlastResults(CNcbiOstream &os)
Definition: blastp_node.cpp:61
CBlastpNode(int check_num, const CNcbiArguments &ncbi_args, const CArgs &args, CBlastAppDiagHandler &bah, string &input, int query_index, int num_queries, CBlastNodeMailbox *mailbox=NULL)
Definition: blastp_node.cpp:50
void DebugDumpText(ostream &out, const string &bundle, unsigned int depth) const
Definition: ddumpable.cpp:56
int GetHitsSortOption() const
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.
int GetHspsSortOption() const
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?
string GetCustomDelimiter()
Class to perform a BLAST search on local BLAST databases Note that PHI-BLAST can be run using this cl...
Definition: local_blast.hpp:62
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
Search Results for All Queries.
const string kArgOutput
Output file name.
int GetDbGeneticCode() const
virtual void SetNumberOfThreads(size_t nthreads)
Mutator for the number of threads.
void ResetMessages(void)
Reset messgae buffer, erase all saved message.
Definition: blast_aux.cpp:1174
CRef< CSearchResultSet > Run()
Executes the search.
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
#define kEmptyStr
Definition: ncbistr.hpp:123
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.
Configuration structure for the CBlastScopeSource.
#define _ASSERT
else result
Definition: token2.c:20
Modified on Tue Apr 23 07:38:35 2024 by modify_doxy.py rev. 669887