OBConversion Class Reference

Class to convert from one format to another. More...

#include <openbabel/obconversion.h>

Inheritance diagram for OBConversion:


Option handling
enum	Option_type { INOPTIONS, OUTOPTIONS, GENOPTIONS }
const char *	IsOption (const char *opt, Option_type opttyp=OUTOPTIONS)
const std::map< std::string, std::string > *	GetOptions (Option_type opttyp)
void	AddOption (const char opt, Option_type opttyp, const char txt=NULL)
bool	RemoveOption (const char *opt, Option_type optype)
void	SetOptions (const char *options, Option_type opttyp)
static void	RegisterOptionParam (std::string name, OBFormat *pFormat, int numberParams=0, Option_type typ=OUTOPTIONS)
static int	GetOptionParams (std::string name, Option_type typ)
Convenience functions
bool	Write (OBBase pOb, std::ostream pout=NULL)
std::string	WriteString (OBBase *pOb, bool trimWhitespace=false)
bool	WriteFile (OBBase *pOb, std::string filePath)
void	CloseOutFile ()
bool	Read (OBBase pOb, std::istream pin=NULL)
bool	ReadString (OBBase *pOb, std::string input)
bool	ReadFile (OBBase *pOb, std::string filePath)
static OBFormat *	GetDefaultFormat ()
static std::string	BatchFileName (std::string &BaseName, std::string &InFile)
static std::string	IncrementedFileName (std::string &BaseName, const int Count)
static bool	CheckForUnintendedBatch (const std::string &infile, const std::string &outfile)
Public Member Functions
Construction
	OBConversion (std::istream is=NULL, std::ostream os=NULL)
	OBConversion (const OBConversion &o)
virtual	~OBConversion ()
Parameter get and set
std::istream *	GetInStream () const
std::ostream *	GetOutStream () const
void	SetInStream (std::istream *pIn)
void	SetOutStream (std::ostream *pOut)
bool	SetInAndOutFormats (const char inID, const char outID)
bool	SetInAndOutFormats (OBFormat pIn, OBFormat pOut)
bool	SetInFormat (const char *inID)
bool	SetInFormat (OBFormat *pIn)
bool	SetOutFormat (const char *outID)
bool	SetOutFormat (OBFormat *pOut)
OBFormat *	GetInFormat () const
OBFormat *	GetOutFormat () const
std::string	GetInFilename () const
std::streampos	GetInPos () const
size_t	GetInLen () const
const char *	GetTitle () const
OBConversion *	GetAuxConv () const
void	SetAuxConv (OBConversion *pConv)
Supported file format
std::vector< std::string >	GetSupportedInputFormat ()
std::vector< std::string >	GetSupportedOutputFormat ()
Conversion
int	Convert (std::istream is, std::ostream os)
int	Convert ()
int	FullConvert (std::vector< std::string > &FileList, std::string &OutputFileName, std::vector< std::string > &OutputFileList)
Conversion loop control
bool	AddChemObject (OBBase *pOb)
OBBase *	GetChemObject ()
bool	IsLast ()
bool	IsFirstInput ()
int	GetOutputIndex () const
void	SetOutputIndex (int indx)
void	SetMoreFilesToCome ()
void	SetOneObjectOnly (bool b=true)
void	SetLast (bool b)
Static Public Member Functions
Collection of formats
static int	RegisterFormat (const char ID, OBFormat pFormat, const char *MIME=NULL)
static OBFormat *	FindFormat (const char *ID)
static OBFormat *	FormatFromExt (const char *filename)
static OBFormat *	FormatFromMIME (const char *MIME)
static bool	GetNextFormat (Formatpos &itr, const char &str, OBFormat &pFormat)
Information
static const char *	Description ()
Protected Types
typedef std::map< std::string, int >	OPAMapType
Protected Member Functions
bool	SetStartAndEnd ()
bool	OpenAndSetFormat (bool SetFormat, std::ifstream *is)
Static Protected Member Functions
static FMapType &	FormatsMap ()
static FMapType &	FormatsMIMEMap ()
static OPAMapType &	OptionParamArray (Option_type typ)
static int	LoadFormatFiles ()
Protected Attributes
std::string	InFilename
std::istream *	pInStream
std::ostream *	pOutStream
OBFormat *	pInFormat
OBFormat *	pOutFormat
std::map< std::string, std::string >	OptionsArray [3]
int	Index
unsigned int	StartNumber
unsigned int	EndNumber
int	Count
bool	m_IsFirstInput
bool	m_IsLast
bool	MoreFilesToCome
bool	OneObjectOnly
bool	ReadyToInput
bool	CheckedForGzip
bool	NeedToFreeInStream
bool	NeedToFreeOutStream
OBBase *	pOb1
std::streampos	wInpos
std::streampos	rInpos
size_t	wInlen
size_t	rInlen
OBConversion *	pAuxConv
std::vector< std::string >	SupportedInputFormat
std::vector< std::string >	SupportedOutputFormat
Static Protected Attributes
static OBFormat *	pDefaultFormat = NULL
static int	FormatFilesLoaded = 0

Detailed Description

Class to convert from one format to another.

OBConversion maintains a list of the available formats, provides information on them, and controls the conversion process.

A conversion is carried out by the calling routine, usually in a user interface or an application program, making an instance of OBConversion. It is loaded with the in and out formats, any options and (usually) the default streams for input and output. Then either the Convert() function is called, which allows a single input file to be converted, or the extended functionality of FullConvert() is used. This allows multiple input and output files, allowing:

aggregation - the contents of many input files converted and sent to one output file;
splitting - the molecules from one input file sent to separate output files;
batch conversion - each input file converted to an output file.

These procedures constitute the "Convert" interface. OBConversion and the user interface or application program do not need to be aware of any other part of OpenBabel - mol.h is not #included. This allows any chemical object derived from OBBase to be converted; the type of object is decided by the input format class. However,currently, almost all the conversions are for molecules of class OBMol. / OBConversion can also be used with an "API" interface called from programs which manipulate chemical objects. Input/output is done with the Read() and Write() functions which work with any chemical object, but need to have its type specified. (The ReadMolecule() and WriteMolecule() functions of the format classes can also be used directly.)

Example code using OBConversion

To read in a molecule, manipulate it and write it out.

Set up an istream and an ostream, to and from files or elsewhere. (cin and cout are used in the example). Specify the file formats.

      OBConversion conv(&cin,&cout);
      if(conv.SetInAndOutFormats("SMI","MOL"))
      { 
         OBMol mol;
         if(conv.Read(&mol))
            // ...manipulate molecule 
                
         conv->Write(&mol);
      }

A two stage construction is used to allow error handling if the format ID is not recognized. This is necessary now that the formats are dynamic and errors are not caught at compile time. OBConversion::Read() is a templated function so that objects derived from OBBase can also be handled, in addition to OBMol, if the format routines are written appropriately.

To make a molecule from a SMILES string.

      std::string SmilesString;
      OBMol mol;
      stringstream ss(SmilesString)
      OBConversion conv(&ss);
      if(conv.SetInFormat("smi") && conv.Read(&mol))
         // ...

To do a file conversion without manipulating the molecule.

      #include <openbabel/obconversion.h> //mol.h is not needed
      ...set up an istream is and an ostream os 
      OBConversion conv(&is,&os);
      if(conv.SetInAndOutFormats("SMI","MOL"))
      {
         conv.SetOptions("h"); //Optional; (h adds expicit hydrogens)
         conv.Convert();
      }

To add automatic format conversion to an existing program.

The existing program inputs from the file identified by the const char* filename into the istream is. The file is assumed to have a format ORIG, but otherformats, identified by their file extensions, can now be used.

      ifstream ifs(filename); //Original code

      OBConversion conv;
      OBFormat* inFormat = conv.FormatFromExt(filename);
      OBFormat* outFormat = conv.GetFormat("ORIG");
      istream* pIn = &ifs; 
      stringstream newstream;
      if(inFormat && outFormat)
      {
         conv.SetInAndOutFormats(inFormat,outFormat);
         conv.Convert(pIn,&newstream);
         pIn=&newstream;
      }
      //else error; new features not available; fallback to original functionality 

      ...Carry on with original code using pIn

In Windows a degree of independence from OpenBabel can be achieved using DLLs. This code would be linked with obconv.lib. At runtime the following DLLs would be in the executable directory: obconv.dll, obdll.dll, one or more *.obf format files.

Member Typedef Documentation

typedef std::map<std::string,int> OPAMapType [protected]

Member Enumeration Documentation

enum Option_type

Three types of options set on the the command line by -a? , -x? , or -?

Enumerator:

INOPTIONS
OUTOPTIONS
GENOPTIONS

Constructor & Destructor Documentation

OBConversion	(	std::istream *	is = `NULL`,
		std::ostream *	os = `NULL`
	)

OBConversion ( const OBConversion & o )

Copy constructor.

~OBConversion ( ) [virtual]

Member Function Documentation

int RegisterFormat	(	const char *	ID,
		OBFormat *	pFormat,
		const char *	MIME = `NULL`
	)			`[static]`

Called once by each format class.

Class information on formats is collected by making an instance of the class derived from OBFormat(only one is usually required). RegisterFormat() is called from its constructor.

If the compiled format is stored separately, like in a DLL or shared library, the initialization code makes an instance of the imported OBFormat class.

OBFormat * FindFormat ( const char * ID ) [static]

Searches registered formats.

OBFormat * FormatFromExt ( const char * filename ) [static]

Searches registered formats for an ID the same as the file extension.

OBFormat * FormatFromMIME ( const char * MIME ) [static]

Searches registered formats for a MIME the same as the chemical MIME type passed.

bool GetNextFormat	(	Formatpos &	itr,
		const char *&	str,
		OBFormat *&	pFormat
	)			`[static]`

Repeatedly called to recover available Formats.

Returns the ID + the first line of the description in str and a pointer to the format in pFormat. If called with str==NULL the first format is returned; subsequent formats are returned by calling with str!=NULL and the previous value of itr returns false, and str and pFormat NULL, when there are no more formats. Use like:

        const char* str=NULL;
        Formatpos pos;
        OBConversion conv; // dummy to make sure static data is available
        while(OBConversion::GetNextFormat(pos,str,pFormat))
        {
                // use str and pFormat
        }
   *

NOTE: Because of dynamic loading problems, it is usually necessary to declare a "dummy" OBConversion object to access this static method. (Not elegant, but will hopefully be fixed in the future.)

const char * Description ( ) [static]

std::istream* GetInStream ( ) const [inline]

std::ostream* GetOutStream ( ) const [inline]

void SetInStream ( std::istream * pIn ) [inline]

void SetOutStream ( std::ostream * pOut ) [inline]

bool SetInAndOutFormats	(	const char *	inID,
		const char *	outID
	)

Sets the formats from their ids, e g CML.

Sets the formats from their ids, e g CML. If inID is NULL, the input format is left unchanged. Similarly for outID Returns true if both formats have been successfully set at sometime

bool SetInAndOutFormats	(	OBFormat *	pIn,
		OBFormat *	pOut
	)

bool SetInFormat ( const char * inID )

Sets the input format from an id e.g. CML.

bool SetInFormat ( OBFormat * pIn )

bool SetOutFormat ( const char * outID )

Sets the output format from an id e.g. CML.

bool SetOutFormat ( OBFormat * pOut )

OBFormat* GetInFormat ( ) const [inline]

OBFormat* GetOutFormat ( ) const [inline]

std::string GetInFilename ( ) const [inline]

std::streampos GetInPos ( ) const [inline]

Get the position in the input stream of the object being read.

size_t GetInLen ( ) const [inline]

Get the length in the input stream of the object being read.

const char * GetTitle ( ) const

Returns:: a default title which is the filename

OBConversion* GetAuxConv ( ) const [inline]

Extension method: deleted in ~OBConversion().

void SetAuxConv ( OBConversion * pConv ) [inline]

const char * IsOption	(	const char *	opt,
		Option_type	opttyp = `OUTOPTIONS`
	)

Determine whether an option is set.

Returns:: NULL if option not and a pointer to the associated text if it is

const std::map<std::string,std::string>* GetOptions ( Option_type opttyp ) [inline]

Access the map with option name as key and any associated text as value.

void AddOption	(	const char *	opt,
		Option_type	opttyp,
		const char *	txt = `NULL`
	)

Set an option of specified type, with optional text.

bool RemoveOption	(	const char *	opt,
		Option_type	optype
	)

void SetOptions	(	const char *	options,
		Option_type	opttyp
	)

Set several single character options of specified type from string like ab"btext"c"ctext".

void RegisterOptionParam	(	std::string	name,
		OBFormat *	pFormat,
		int	numberParams = `0`,
		Option_type	typ = `OUTOPTIONS`
	)			`[static]`

For example -h takes 0 parameters; -f takes 1. Call in a format constructor.

int GetOptionParams	(	std::string	name,
		Option_type	typ
	)			`[static]`

Returns:: the number of parameters registered for the option, or 0 if not found

std::vector< std::string > GetSupportedInputFormat ( )

Returns the list of supported input format

std::vector< std::string > GetSupportedOutputFormat ( )

Returns the list of supported output format

int Convert	(	std::istream *	is,
		std::ostream *	os
	)

Conversion for single input and output stream.

int Convert ( )

Conversion with existing streams.

Actions the "convert" interface. Calls the OBFormat class's ReadMolecule() which

makes a new chemical object of its chosen type (e.g. OBMol)
reads an object from the input file
subjects the chemical object to 'transformations' as specified by the Options
calls AddChemObject to add it to a buffer. The previous object is first output via the output Format's WriteMolecule(). During the output process calling IsFirst() and GetIndex() (the number of objects including the current one already output. allows more control, for instance writing <cml> and </cml> tags for multiple molecule outputs only.

AddChemObject does not save the object passed to it if it is NULL (as a result of a DoTransformation()) or if the number of the object is outside the range defined by StartNumber and EndNumber.This means the start and end counts apply to all chemical objects found whether or not they are output.

If ReadMolecule returns false the input conversion loop is exited.

int FullConvert	(	std::vector< std::string > &	FileList,
		std::string &	OutputFileName,
		std::vector< std::string > &	OutputFileList
	)

Conversion with multiple input/output files: makes input and output streams, and carries out normal, batch, aggregation, and splitting conversion.

Makes input and output streams, and carries out normal, batch, aggregation, and splitting conversion.

Normal Done if FileList contains a single file name and OutputFileName does not contain a *.

Aggregation Done if FileList has more than one file name and OutputFileName does not contain * . All the chemical objects are converted and sent to the single output file.

Splitting Done if FileList contains a single file name and OutputFileName contains a * . Each chemical object in the input file is converted and sent to a separate file whose name is OutputFileName with the replaced by 1, 2, 3, etc. OutputFileName must have at least one character other than the * before the extension. For example, if OutputFileName is NEW*.smi then the output files are NEW1.smi, NEW2.smi, etc.

Batch Conversion Done if FileList has more than one file name and contains a * . Each input file is converted to an output file whose name is OutputFileName with the * replaced by the inputfile name without its path and extension. So if the input files were inpath/First.cml, inpath/Second.cml and OutputFileName was NEW*.mol, the output files would be NEWFirst.mol, NEWSecond.mol.

If FileList is empty, the input stream that has already been set (usually in the constructor) is used. If OutputFileName is empty, the output stream already set is used.

On exit, OutputFileList contains the names of the output files.

Returns the number of Chemical objects converted.

bool AddChemObject ( OBBase * pOb )

Adds to internal array during input.

Called by ReadMolecule() to deliver an object it has read from an input stream. Used in two modes:

When Count is negative it is left negative and the routine is just a store for an OBBase object. The negative value returned tells the calling routine that no more objects are required.
When count is >=0, probably set by Convert(), it acts as a queue of 2: writing the currently stored value before accepting the supplied one. This delay allows output routines to respond differently when the written object is the last. Count is incremented with each call, even if pOb=NULL. Objects are not added to the queue if the count is outside the range StartNumber to EndNumber. There is no upper limit if EndNumber is zero. The return value is the return code from WriteChemObject - false on error.

OBBase * GetChemObject ( )

Retrieve from internal array during output.

Retrieves an object stored by AddChemObject() during output

bool IsLast ( )

True if no more objects to be output.

Reimplemented in XMLConversion.

bool IsFirstInput ( )

True if the first input object is being processed.

int GetOutputIndex ( ) const

Retrieves number of ChemObjects that have been actually output.

Returns the number of objects which have been output or are currently being output. The outputindex is incremented when an object for output is fetched by GetChemObject(). So the function will return 1 if called from WriteMolecule() during output of the first object.

void SetOutputIndex ( int indx )

Sets ouput index (maybe to control whether seen as first object).

void SetMoreFilesToCome ( )

Used with multiple input files. Off by default.

void SetOneObjectOnly ( bool b = true )

Used with multiple input files. Off by default.

void SetLast ( bool b ) [inline]

static OBFormat* GetDefaultFormat ( ) [inline, static]

The default format is set in a single OBFormat class (generally it is OBMol).

bool Write	(	OBBase *	pOb,
		std::ostream *	pout = `NULL`
	)

Outputs an object of a class derived from OBBase.

Part of "API" interface. The output stream can be specified and the change is retained in the OBConversion instance

std::string WriteString	(	OBBase *	pOb,
		bool	trimWhitespace = `false`
	)

Outputs an object of a class derived from OBBase as a string.

Part of "API" interface. The output stream is temporarily changed to the string and then restored This method is primarily intended for scripting languages without "stream" classes The optional "trimWhitespace" parameter allows trailing whitespace to be removed (e.g., in a SMILES string or InChI, etc.)

bool WriteFile	(	OBBase *	pOb,
		std::string	filePath
	)

Outputs an object of a class derived from OBBase as a file (with the supplied path).

Part of "API" interface. The output stream is changed to the supplied file and the change is retained in the OBConversion instance. This method is primarily intended for scripting languages without "stream" classes

void CloseOutFile ( )

Manually closes and deletes the output stream The file is closed anyway when in the OBConversion destructor or when WriteFile is called again.

Since:: version 2.1

bool Read	(	OBBase *	pOb,
		std::istream *	pin = `NULL`
	)

Reads an object of a class derived from OBBase into pOb.

Part of "API" interface. The input stream can be specified and the change is retained in the OBConversion instance

Returns:: false and pOb=NULL on error

bool ReadString	(	OBBase *	pOb,
		std::string	input
	)

Reads an object of a class derived from OBBase into pOb from the supplied string.

Part of "API" interface.

Returns:: false and pOb=NULL on error This method is primarily intended for scripting languages without "stream" classes

bool ReadFile	(	OBBase *	pOb,
		std::string	filePath
	)

Reads an object of a class derived from OBBase into pOb from the file specified.

Part of "API" interface. The output stream is changed to the supplied file and the change is retained in the OBConversion instance.

Returns:: false and pOb=NULL on error This method is primarily intended for scripting languages without "stream" classes

string BatchFileName	(	std::string &	BaseName,
		std::string &	InFile
	)			`[static, protected]`

Replaces * in BaseName by InFile without extension and path.

string IncrementedFileName	(	std::string &	BaseName,
		const int	Count
	)			`[static, protected]`

Replaces * in BaseName by Count.

bool CheckForUnintendedBatch	(	const std::string &	infile,
		const std::string &	outfile
	)			`[static, protected]`

Checks for misunderstandings when using the -m option.

bool SetStartAndEnd ( ) [protected]

FMapType & FormatsMap ( ) [static, protected]

contains ID and pointer to all OBFormat classes

This static function returns a reference to the FormatsMap which, because it is a static local variable is constructed only once. This fiddle is to avoid the "static initialization order fiasco" See Marshall Cline's C++ FAQ Lite document, www.parashift.com/c++-faq-lite/".

FMapType & FormatsMIMEMap ( ) [static, protected]

contains MIME and pointer to all OBFormat classes

This static function returns a reference to the FormatsMIMEMap which, because it is a static local variable is constructed only once. This fiddle is to avoid the "static initialization order fiasco" See Marshall Cline's C++ FAQ Lite document, www.parashift.com/c++-faq-lite/".

OBConversion::OPAMapType & OptionParamArray ( Option_type typ ) [static, protected]

int LoadFormatFiles ( ) [static, protected]

bool OpenAndSetFormat	(	bool	SetFormat,
		std::ifstream *	is
	)			`[protected]`

Member Data Documentation

std::string InFilename [protected]

std::istream* pInStream [protected]

std::ostream* pOutStream [protected]

OBFormat * pDefaultFormat = NULL [static, protected]

OBFormat* pInFormat [protected]

OBFormat* pOutFormat [protected]

std::map<std::string,std::string> OptionsArray[3] [protected]

int Index [protected]

unsigned int StartNumber [protected]

unsigned int EndNumber [protected]

int Count [protected]

bool m_IsFirstInput [protected]

bool m_IsLast [protected]

bool MoreFilesToCome [protected]

bool OneObjectOnly [protected]

bool ReadyToInput [protected]

bool CheckedForGzip [protected]

input stream is gzip-encoded

bool NeedToFreeInStream [protected]

bool NeedToFreeOutStream [protected]

int FormatFilesLoaded = 0 [static, protected]

OBBase* pOb1 [protected]

std::streampos wInpos [protected]

position in the input stream of the object being written

std::streampos rInpos [protected]

position in the input stream of the object being read

size_t wInlen [protected]

length in the input stream of the object being written

size_t rInlen [protected]

length in the input stream of the object being read

OBConversion* pAuxConv [protected]

Way to extend OBConversion.

std::vector<std::string> SupportedInputFormat [protected]

list of supported input format

std::vector<std::string> SupportedOutputFormat [protected]

list of supported output format

The documentation for this class was generated from the following files:

OBConversion Class Reference

Option handling

Convenience functions

Public Member Functions

Static Public Member Functions

Protected Types

Protected Member Functions

Static Protected Member Functions