Design overview

The command-language translators manage storage by creating a stack from a single area allocated at the start of the program.

Because the input is free-format, the translators move it into a buffer area that can hold data spanning two or more source records. The analysis of the source is mainly table driven.

The translators build the replacement source code for each EXEC command in a form appropriate to the language:

Errors in the source can be detected. Spelling corrections are made to the source, and any unrecognizable or duplicate keywords and options are ignored. For COBOL, PL/I, and C, the translator produces error diagnostics that are collected together on the output listing. The assembler language translator, however, produces error diagnostics in the translated output following the EXEC command in which the error occurred.

[[ Contents Previous Page | Next Page Index ]]