Integrating your own format

The API of Chemaxon IO allows you to implement your own format and to add it to the existing framework. The following steps are needed: A complete example code is available to demonstrate this:

Creating file format metadata for your own format

Creating the metadata includes specifying file extension, importer, exporter, record reader, format recognizer and some format features like atomic coordinates, multiple records storage possibility. 
	
	MYFORMAT = new MFileFormat(
	    "My Format", // format description
	    "myformat", // format name
	    null, // no MIME type
	    "myformat myf", // file extension
	    "MyFormatRecordReader", // record reader class
	    "MyFormatImport",  // import module implementation
	    "MyFormatExport", // export module implementation
	    "MyFormatRecognizer 10", // recognition class with priority specification
	    MFileFormat.F_IMPORT | MFileFormat.F_EXPORT | MFileFormat.F_RECOGNIZER
	    | MFileFormat.F_MOLECULE | MFileFormat.F_COORDS
	    | MFileFormat.F_MULTIPLE_RECORDS_LEGAL
	    | MFileFormat.F_MULTIPLE_RECORDS_POSSIBLE);

For a complete source code, please see Init.java.

After this step the format is registered by: 
	
	MFileFormatUtil.registerFormat(Init.MYFORMAT);

Creating exporter for your own format

In order to create your own format exporter you need to implement the convert() method of chemaxon.marvin.io.MolExportModule which is the abstract base class of molecule export modules. The convert() might return String or byte array or image object. Optionally, you can implement the open(String fmtopts) and close() methods also. The overriding open(String fmtopts) method opens the exporter stream and should call super.open(fmtopts) at the beginning. In case of some multi-molecule formats such as RDfile, files begin with a header. This header must be returned by open(), either as a String object or a byte[] array. The close() method is called after the last convert() to close the stream if needed.

For a complete source code, please see MyFormatExport.java.



Creating importer for your own format

The MolImportModule is the base class of Molecule import modules.
The two basic abstract methods are the createMol() which creates a new target molecule object and the readMol(Molecule mol) which reads a molecule from the file.

For a complete source code, please see MyFormatImport.java.





Creating record reader for your own format

Record is a string (or byte representation) of a single molecule with properties in a multi-molecule file.
Record reading is faster than reading into molecule objects and makes property pre-reading possible.
If you would like to use record reading you have to implement the MRecordReader interface. MPropHandler helps the property pre-reading.



For a complete source code, please see MyFormatRecordReader.java.



Creating format recognizer for your own format

Sometimes it is useful to implement the format recognition 
for your own format to detect the input format from the content of the structure file 
in order to import the structure file without file format specification parameter.
The implementation of the tryToRecognize() and needsMore() from the Recognizer interface is needed. This functionality is just optional.

For a complete source code, please see MyFormatRecognizer.java.






This picture illustrates how the classes related to your own format connect to the existing framework:



General overview of IO classes in the existing framework:

 

Behavior of IO classes in the existing framework:





   

      
       

        Molecule converter  

      		  Previous chapter
       
   
     		
     
      

        Table of contents