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

Go to the SVN repository for this file.

1 /* $Id: tblastx_app.cpp 90590 2020-07-01 16:17:57Z fongah2 $
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: Christiam Camacho
27  *
28  */
29 
30 /** @file tblastx_app.cpp
31  * TBLASTX command line application
32  */
33 
34 #include <ncbi_pch.hpp>
35 #include <corelib/ncbiapp.hpp>
42 #include "blast_app_util.hpp"
43 
44 #ifndef SKIP_DOXYGEN_PROCESSING
46 USING_SCOPE(blast);
48 #endif
49 
51 {
52 public:
53  /** @inheritDoc */
56  version->SetVersionInfo(new CBlastVersion());
59  if (m_UsageReport.IsEnabled()) {
61  }
62  }
63 
66  }
67 private:
68  /** @inheritDoc */
69  virtual void Init();
70  /** @inheritDoc */
71  virtual int Run();
72 
73  /// This application's command line args
77 };
78 
80 {
81  // formulate command line arguments
82 
84 
85  // read the command line
86 
89 }
90 
92 {
93  int status = BLAST_EXIT_SUCCESS;
95 
96  try {
97 
98  // Allow the fasta reader to complain on invalid sequence input
100  SetDiagPostPrefix("tblastx");
101  SetDiagHandler(&bah, false);
102 
103  /*** Get the BLAST options ***/
104  const CArgs& args = GetArgs();
105  CRef<CBlastOptionsHandle> opts_hndl;
107  opts_hndl.Reset(&*m_CmdLineArgs->SetOptionsForSavedStrategy(args));
108  }
109  else {
110  opts_hndl.Reset(&*m_CmdLineArgs->SetOptions(args));
111  }
112  const CBlastOptions& opt = opts_hndl->GetOptions();
113 
114  /*** Initialize the database/subject ***/
116  CRef<CLocalDbAdapter> db_adapter;
117  CRef<CScope> scope;
118  InitializeSubject(db_args, opts_hndl, m_CmdLineArgs->ExecuteRemotely(),
119  db_adapter, scope);
120  _ASSERT(db_adapter && scope);
121 
122  /*** Get the query sequence(s) ***/
123  CRef<CQueryOptionsArgs> query_opts =
125  SDataLoaderConfig dlconfig =
127  db_adapter);
128  CBlastInputSourceConfig iconfig(dlconfig, query_opts->GetStrand(),
129  query_opts->UseLowercaseMasks(),
130  query_opts->GetParseDeflines(),
131  query_opts->GetRange());
133  ERR_POST(Warning << "Query is Empty!");
134  return BLAST_EXIT_SUCCESS;
135  }
138 
139  /*** Get the formatting options ***/
141  bool isArchiveFormat = fmt_args->ArchiveFormatRequested(args);
142  if(!isArchiveFormat) {
143  bah.DoNotSaveMessages();
144  }
145  CBlastFormat formatter(opt, *db_adapter,
146  fmt_args->GetFormattedOutputChoice(),
147  query_opts->GetParseDeflines(),
149  fmt_args->GetNumDescriptions(),
150  fmt_args->GetNumAlignments(),
151  *scope,
152  opt.GetMatrixName(),
153  fmt_args->ShowGis(),
154  fmt_args->DisplayHtmlOutput(),
155  opt.GetQueryGeneticCode(),
156  opt.GetDbGeneticCode(),
157  opt.GetSumStatisticsMode(),
159  db_adapter->GetFilteringAlgorithm(),
160  fmt_args->GetCustomOutputFormatSpec(),
161  false, false, NULL, NULL,
163  GetSubjectFile(args));
164 
165  formatter.SetQueryRange(query_opts->GetRange());
166  formatter.SetLineLength(fmt_args->GetLineLength());
167  formatter.SetHitsSortOption(fmt_args->GetHitsSortOption());
168  formatter.SetHspsSortOption(fmt_args->GetHspsSortOption());
169  formatter.SetCustomDelimiter(fmt_args->GetCustomDelimiter());
170  if(UseXInclude(*fmt_args, args[kArgOutput].AsString())) {
171  formatter.SetBaseFile(args[kArgOutput].AsString());
172  }
173  formatter.PrintProlog();
174 
175  /*** Process the input ***/
176  for (; !input.End(); formatter.ResetScopeHistory(), QueryBatchCleanup()) {
177 
178  CRef<CBlastQueryVector> query_batch(input.GetNextSeqBatch(*scope));
179  CRef<IQueryFactory> queries(new CObjMgr_QueryFactory(*query_batch));
180 
181  SaveSearchStrategy(args, m_CmdLineArgs, queries, opts_hndl);
182 
183  CRef<CSearchResultSet> results;
184 
186  CRef<CRemoteBlast> rmt_blast =
187  InitializeRemoteBlast(queries, db_args, opts_hndl,
190  results = rmt_blast->GetResultSet();
191  } else {
192  CLocalBlast lcl_blast(queries, opts_hndl, db_adapter);
194  results = lcl_blast.Run();
195  }
196 
197  if (fmt_args->ArchiveFormatRequested(args)) {
198  formatter.WriteArchive(*queries, *opts_hndl, *results, 0, bah.GetMessages());
199  bah.ResetMessages();
200  } else {
201  BlastFormatter_PreFetchSequenceData(*results, scope,
202  fmt_args->GetFormattedOutputChoice());
203  ITERATE(CSearchResultSet, result, *results) {
204  formatter.PrintOneResultSet(**result, query_batch);
205  }
206  }
207  }
208 
209  formatter.PrintEpilog(opt);
210 
212  opts_hndl->GetOptions().DebugDumpText(NcbiCerr, "BLAST options", 1);
213  }
214 
217  } CATCH_ALL(status)
218  if(!bah.GetMessages().empty()) {
219  const CArgs & a = GetArgs();
221  }
224  return status;
225 }
226 
227 #ifndef SKIP_DOXYGEN_PROCESSING
228 int NcbiSys_main(int argc, ncbi::TXChar* argv[])
229 {
230  return CTblastxApp().AppMain(argc, argv);
231 }
232 #endif /* SKIP_DOXYGEN_PROCESSING */
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 PrintErrorArchive(const CArgs &a, const list< CRef< CBlast4_error > > &msg)
Function to print blast archive with only error messages (search failed) to output stream.
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)
bool IsIStreamEmpty(CNcbiIstream &in)
void LogQueryInfo(CBlastUsageReport &report, const CBlastInput &q_info)
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.
virtual CNcbiIstream & GetInputStream()
Get the input stream.
size_t GetNumThreads() const
Get the number of threads to spawn.
CRef< CBlastOptionsHandle > SetOptions(const CArgs &args)
Extract the command line arguments into a CBlastOptionsHandle object.
CRef< CBlastDatabaseArgs > GetBlastDatabaseArgs() const
Get the BLAST database arguments.
CArgDescriptions * SetCommandLine()
Set the command line 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...
virtual CNcbiOstream & GetOutputStream()
Get the output stream.
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 LogBlastSearchInfo(blast::CBlastUsageReport &report)
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...
Encapsulates ALL the BLAST algorithm's options.
void AddParam(EUsageParams p, int val)
Keeps track of the version of the BLAST engine in the NCBI C++ toolkit.
Definition: version.hpp:53
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
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
CRef –.
Definition: ncbiobj.hpp:618
Search Results for All Queries.
CStopWatch –.
Definition: ncbitime.hpp:1937
Handles command line arguments for blastx binary.
virtual int GetQueryBatchSize() const
@inheritDoc
virtual int Run()
@inheritDoc
Definition: tblastx_app.cpp:91
CStopWatch m_StopWatch
Definition: tblastx_app.cpp:76
CBlastUsageReport m_UsageReport
Definition: tblastx_app.cpp:75
CRef< CTblastxAppArgs > m_CmdLineArgs
This application's command line args.
Definition: tblastx_app.cpp:74
CTblastxApp()
@inheritDoc
Definition: tblastx_app.cpp:54
virtual void Init()
@inheritDoc
Definition: tblastx_app.cpp:79
const string kArgOutput
Output file name.
void Print(const CCompactSAMApplication::AlignInfo &ai)
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
void SetFullVersion(CRef< CVersionAPI > version)
Set version data for the program.
Definition: ncbiapp.cpp:1174
void HideStdArgs(THideStdArgs hide_mask)
Set the hide mask for the Hide Std Flags.
Definition: ncbiapp.cpp:1312
virtual const CArgs & GetArgs(void) const
Get parsed command line arguments.
Definition: ncbiapp.cpp:305
int AppMain(int argc, const char *const *argv, const char *const *envp=0, EAppDiagStream diag=eDS_Default, const char *conf=NcbiEmptyCStr, const string &name=NcbiEmptyString)
Main function (entry point) for the NCBI application.
Definition: ncbiapp.cpp:819
CVersionInfo GetVersion(void) const
Get the program version information.
Definition: ncbiapp.cpp:1184
virtual void SetupArgDescriptions(CArgDescriptions *arg_desc)
Setup the command line argument descriptions.
Definition: ncbiapp.cpp:1195
#define ITERATE(Type, Var, Cont)
ITERATE macro to sequence through container elements.
Definition: ncbimisc.hpp:815
const CNcbiArguments & GetArguments(void) const
Get the application's cached unprocessed command-line arguments.
@ fHideXmlHelp
Hide XML help description.
@ fHideLogfile
Hide log file description.
@ fHideFullVersion
Hide full version description.
@ fHideDryRun
Hide dryrun description.
@ fHideConffile
Hide configuration file description.
#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
EDiagSev SetDiagPostLevel(EDiagSev post_sev=eDiag_Error)
Set the threshold severity for posting the messages.
Definition: ncbidiag.cpp:6129
#define ERR_POST(message)
Error posting with file, line number information but without error codes.
Definition: ncbidiag.hpp:186
void SetDiagHandler(CDiagHandler *handler, bool can_delete=true)
Set the diagnostic handler using the specified diagnostic handler class.
Definition: ncbidiag.cpp:6288
@ eDiag_Warning
Warning message.
Definition: ncbidiag.hpp:652
void Warning(CExceptionArgs_Base &args)
Definition: ncbiexpt.hpp:1191
void Reset(void)
Reset reference object.
Definition: ncbiobj.hpp:773
bool IsEnabled(void)
Indicates whether application usage statistics collection is enabled for a current reporter instance.
#define NcbiCerr
Definition: ncbistre.hpp:544
char TXChar
Definition: ncbistr.hpp:172
double Elapsed(void) const
Return time elapsed since first Start() or last Restart() call (in seconds).
Definition: ncbitime.hpp:2775
void Start(void)
Start the timer.
Definition: ncbitime.hpp:2764
#define CVersion
static int input()
Main class to perform a BLAST search on the local machine.
static int version
Definition: mdb_load.c:29
unsigned int a
Definition: ncbi_localip.c:102
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.
USING_SCOPE(blast)
int NcbiSys_main(int argc, ncbi::TXChar *argv[])
USING_NCBI_SCOPE
Definition: tblastx_app.cpp:45
Main argument class for TBLASTX application.
#define _ASSERT
else result
Definition: token2.c:20
Modified on Wed May 29 18:42:02 2024 by modify_doxy.py rev. 669887