XML Converter is an advanced COBOL-XML(SOAP) conversion tool supporting many latest XML features and designed for speed and reliability. It is written in COBOL and distributed in source code.
It uses a regular COBOL record definition to describe both the COBOL data and the structure of the XML document.
In its simplest form the COBOL record layout can be used directly, without any modification and this still would allow to create or to read a well formed XML document.
For more complex XML documents comment lines with specific information are added, but because they do not change the original COBOL layout the same copybook can be shared between the existing applications and the XML Converter.
Features:
runs on all COBOL platforms
handles documents of unlimited size and format
converts in both directions: COBOL to and from XML (or SOAP)
supports choices and sequences including nested and repeating
supports unbounded elements, nillable, extensible, namespaces, substitution group names, CDATA, optional elements
handles mixed content elements and non-tagged group level fields
runs in a validating or non-validating mode
uses COBOL record definition to describe both the COBOL data and the XML document
converts from EBCDIC to ASCII and vice versa
translates binary fields between big-endian and little-endian
does not use any external work-files (I-O is costly)
has a built-in protection for non-numeric data in COMP-3 and zoned decimal (PIC 9) fields
supports multi dimensional arrays inside the records (COBOL limits applies)
has a large internal record size limit of 1GB (the actual limit depends on the host operating system)
parses COBOL copybooks at run-time, but only once, during the first call, to achieve maximum performance
it is distributed in source code allowing you to change program name, data structure sizes, and to have as many copies as you want (say with different access/execution rights or transaction names in CICS)
Cobol Converter
Cobol Converter is a Mainframe, Unix, iSeries, and Windows data manipulation tool that can transfer information between different record layouts and field types or quickly generate an XML document (or SOAP message) without the need for a schema, DTD, or even a sample XML file. It does that by utilizing the regular COBOL record definition (that you might already have). It is written in COBOL and distributed in source code.
It can transform numeric fields to all other numeric, numeric to text, but also text to numeric - assuming input holds a valid number or spaces. It can convert the whole record from binary to text format (and then convert it back to binary on the other end of transmission channel) just by specifying one option.
It can also parse an XML document or SOAP message and transfer information to a corresponding COBOL structure (limited to fixed-length XML documents with fixed-position nodes, basically meaning it can only read its own output - still useful to create a coding/decoding pair of applications to process, for example, MQ messages). This is also done by specifying just one option.
Features:
runs on all COBOL platforms
works at high logical level, meaning you just map structures and fields in source and destination records without specifying the byte offset and field length, like in File-Aid or many other tools
converts alphabetic fields from EBCDIC to ASCII and vice versa
translates binary fields between big-endian and little-endian
does not use any external work-files (I-O is costly)
has a built-in protection for non-numeric data in COMP-3 and zoned decimal (PIC 9) fields
supports multi dimensional arrays inside the records (COBOL limits applies)
supports redefinitions (no limits)
properly initializes all elementary and group items (including the repeating ones) with default values but also the VALUE values (surprisingly large number of compilers have problem doing that)
confirms to most of ANSI85 COBOL data features
has a large internal record size limit of 1GB (actual limit depends on the host operating system limit for the COBOL 01 level: Mainframe - 16MB, Unix - 1GB, others - depending on compiler)
parses COBOL copybooks at run-time, but only once, during the first call, to achieve maximum performance
it is distributed as a source code module allowing you to change program name, maximum I/O data structure sizes and other parameters (XML definition, SOAP tags, namespaces, etc.), and to have as many copies as you want (say with different access/execution rights or transaction names in CICS)
converts data to variable or fixed size XML documents (reads only fixed size)
avaliable in standalone, batch and stored procedure version
