8.2 The COBOL copybook generation

In an XML file, the elements can be classified in 3 categories :

• Root : the first XML element. It's ignored in the generation process and is not part of the generated COBOL copybook.
• First child : the second XML element. Each time it appears, the meta-data analysis of the XML elements is (re-)started. It corresponds to the first COBOL item in the copybook. As this first child element includes all the other XML elements, its corresponding COBOL item is a grouping item, with the lowest level and a unique occurrence.
• Other child elements: the other XML elements. They are analyzed to extract the meta-data needed to build the rest of the COBOL copybook.

The COBOL copybook generation process take place in the following steps :

• The XML file is read record per record until the end of the file.
• The first child is then searched.
• Each XML element or attributes generates a COBOL item applying the following rules :
  • The name of the XML element or attribute is capitalized and dashes (''-'') are translated to underscore (''_''). For example : <cobol_item_01> becomes COBOL-ITEM-01. COBOL item names are checked against a COBOL reserved word list^8.1. A warning message is issued if the generated item name appears in this list,
  • Meta-data of the content^8.2 of the XML element and attributes is extracted : length, type and sign.
  • Length of item is rounded if requested (see sub-section ). A COBOL item is then created.
• Each time the XML first child is found, the analysis is restarted and the meta-data is updated if necessary. For example, if the content length of an XML element is greater than the current length of its corresponding COBOL item, the length is updated. The contrary is, of course, not true.
• When the end of the XML file is reached :
  • The item list is then checked against elementary or grouping items occurring more than one time. If this is the case, only the first occurrence of the item is kept, the others are deleted and an occurrence counter is created
  • The starting position of each COBOL item is computed. Although, it's not necessary for the COBOL copybook, it is used for final reporting.
  • The COBOL copybook is written, indenting the items according to their COBOL level.