Browse Prior Art Database

Automatically Building Hyper-Text Databases of Source Code

IP.com Disclosure Number: IPCOM000106386D
Original Publication Date: 1993-Nov-01
Included in the Prior Art Database: 2005-Mar-21
Document File: 4 page(s) / 129K

Publishing Venue

IBM

Related People

Cobbett, M: AUTHOR

Abstract

Disclosed is a hyper-text database to store program source code and documentation files in a form that allows convenient and efficient access. Described also is use of a tool to build the database. Solved are existing problems of maintenance and support of large programs requiring knowledge of program flow, links between the components, cross references to the location of variables and the data. Also the building of hyper-text links for program source code and documentation.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 40% of the total text.

Automatically Building Hyper-Text Databases of Source Code

      Disclosed is a hyper-text database to store program source code
and documentation files in a form that allows convenient and
efficient access.  Described also is use of a tool to build the
database.  Solved are existing problems of maintenance and support of
large programs requiring knowledge of program flow, links between the
components, cross references to the location of variables and the
data.  Also the building of hyper-text links for program source code
and documentation.

      Benefits are a user-friendly interface with 'instant' online
access to program information.  No special requirements are placed on
the programmer while developing code, the use of the tool is done on
program completion making use of existing documentation.  Hyper-text
links are simple to generate.  Users can navigate through a piece of
source code easier than if they had to edit/browse files on-line, or
pour over printed listings.  The hyper-text database is built
automatically, only requiring a simple AWK file to be written to do
the source-to-database format conversion.

Newcomers to unfamiliar software source may encounter the following
problems:

1.  Not knowing in which source file a function or constant is
    defined.

2.  Not having instant access to all source files needed, ie
    requiring to browse through a large number of source files at one
    time.

3.  Not having access to the source at all, to view it and learn
    about it on a different machine, or one which isn't connected to
    the source repository, i.e., the LAN in one case.

4.  Not having enough disk space to store all of the source on local
    hard disk.

5.  Not being able to "follow" the flow of execution of code through
    the program, mainly because it takes a long time finding where
    functions, types, structures, and constants are defined, and what
    they mean.

6.  When copying the source, sometimes it is difficult to gather it
    all together, finding later that files are missing.

      Also the problem that when in a remote location, without access
to the source, compelled when debugging to continually print off lots
of listings (which get out of date quickly).

      Desirable is a One-File database, which is compressed avoiding
excessive use of disk-space.  It becomes easy to copy around systems,
which contains all other source code, (header files and all), which
also contains imbedded hyper-text links between functions, and is
built automatically by a build-like process; possibly overnight, to
keep the database current throughout development stages.

      The disclosed solution is to convert normal source-code, such
as C and Assembler, but it could also apply to any programming
language, and create a hyper-text database which can be browsed
easily using a hyper-text browser.  Microsoft's QuickHelp utility was
used in the example at the end of...