XTRAN User's Manual

Chapter 1 ("mini" for demo)

(This chapter revised June 30, 2004)

XTRAN is a specialized expert system developed by XTRAN, LLC for the symbolic manipulation of computer languages: Second generation or 2GL (assembly) languages, third generation or 3GL languages, fourth generation or 4GL languages, scripting languages, markup languages, and data base languages — in fact, any well-defined computer language. XTRAN provides the ability to analyze, translate, re-engineer, and generate these languages.

The following is a brief overview of XTRAN's capabilities. For more information, please contact XTRAN, LLC.

Code analysis

XTRAN can analyze all of the computer languages it accommodates. When analyzing, XTRAN is configured with one or more input languages, and operates in three phases:

The input phase reads XTRAN commands and rules, source code to be analyzed, and, optionally, data declarations in the source or a target language. From these it builds XTRAN's initial internal data bases: Source language, commands, and rules. XTRAN converts all languages to the same proprietary internal representation.
The analysis phase performs the requested analysis, by perusing the internal representation of the source language and accumulating analysis results for eventual output.
The output phase generates output text from the internal representation of the analysis results.

The fact that XTRAN's code analysis is driven by user-specified rules means that complex analyses can be set up and executed quickly. And, of course, analysis rules that prove to be generally useful can be saved and reused as desired.

Code translation

XTRAN can translate a variety of source languages, including assembly languages, into a variety of target languages. When translating, XTRAN is configured with one or more input languages and one or more output languages, and operates in four phases:

The input phase reads XTRAN commands and rules, source code to be operated on, and, optionally, data declarations in the source or target language. From these it builds XTRAN's initial internal data bases: Source language, commands, and rules. XTRAN converts all languages to the same proprietary internal representation.
The translation phase performs the requested translation, by translating the internal representation of the source language into an internal representation of the target language.
The built-in re-engineering phase raises the level and quality of the target code through built-in structuring and optimization rules. It also provides a generalized mechanism for user-specified optimizations and re-engineering.
The output phase generates, from the internal representation of the target language, XTRAN's output: text target language programs.

XTRAN's translation phase is guided by user-editable rules, which specify the target code to be generated for each source language operator or statement. XTRAN is not a mechanical, line by line translator; it is fully symbolic, and its sophisticated rules capability works on an entire body of code at a time.

Code improvement and re-engineering

XTRAN can also be configured to provide the ability to improve existing code through structuring, standardization, and any other rule-driven transformation. This includes the ability to enforce programming standards and conventions. XTRAN can provide this code improvement capability for all of the languages it accommodates. It includes the ability to apply XTRAN's sophisticated rule-based behavior to code improvement and standardization. An example is structuring poorly structured code.

XTRAN can re-engineer a variety of source languages, including assembly languages. When doing so, XTRAN is configured with one or more input languages and the same output languages, and operates in four phases:

The input phase reads XTRAN commands and rules, source code to be operated on, and, optionally, data declarations in any of the source languages. From these it builds XTRAN's initial internal data bases: Source language, commands, and rules. XTRAN converts all languages to the same proprietary internal representation.
The re-engineering phase performs the requested re-engineering, by operating on the internal representation of the source language as specified by the user's rules.
The built-in re-engineering phase raises the level and quality of the code through built-in structuring and optimization rules.
The output phase generates, from the internal representation of the source language, XTRAN's output: text language programs.

XTRAN's re-engineering phase is guided by user-editable rules, which specify the operations to be performed on the input language before it is output.

Code generation

XTRAN provides the capability of generating target code from any higher-level language. It is even possible to custom design computer languages as needed and have XTRAN compile them into 3GL code. XTRAN's user-specified rule capabilities allow sophisticated code generation, not just mechanical expansion.

XTRAN can generate essentially any computer language. When doing so, XTRAN is configured with one or more input languages and one or more output languages, and operates in four phases:

The input phase reads XTRAN commands and rules, and source code to be operated on. From these it builds XTRAN's initial internal data bases: Source language, commands, and rules. XTRAN converts all computer languages to the same proprietary internal representation.
The code generation phase performs the requested code generation, by operating on the internal representation of the source language, as specified by the user's rules, to generate target code.
The built-in re-engineering phase may raise the level and quality of the target code through built-in structuring and optimization rules.
The output phase generates, from the internal representation of the target language, XTRAN's output: Text target language programs.

XTRAN's code generation phase is guided by user-editable rules, which specify the operations to be performed.