B
Baron Samedi
When writing design documents, I don't really consider my design
complete until I have identified each data struct or class and when it
is initialized/destroyed and written(/read).
Source browsers can be useful to help me do this in legacy/third party
code, but I stumbled across something that got me thinking at
http://www.sgvsarc.com/overview_dataflow.htm
It diagrams nicely, showing the source code as a sort of zoomed out
flowchart, and showing usage of a specified variable.
I am just about to download the demo and give it a spin, but it got me
wondering how useful this would be with a global variable, or
something passed as a parameter, across multiple source files/classes.
As an added complication, what about passing by message?
Does anyone know of any slick tools for this sort of thing? I have
been looking lately at reverse engineering inherited legacy code,
running it through DoxyGen, producing function call trees and class
inheritance trees, but haven't yet given too much thought to data,
which is probably a big mistake.
complete until I have identified each data struct or class and when it
is initialized/destroyed and written(/read).
Source browsers can be useful to help me do this in legacy/third party
code, but I stumbled across something that got me thinking at
http://www.sgvsarc.com/overview_dataflow.htm
It diagrams nicely, showing the source code as a sort of zoomed out
flowchart, and showing usage of a specified variable.
I am just about to download the demo and give it a spin, but it got me
wondering how useful this would be with a global variable, or
something passed as a parameter, across multiple source files/classes.
As an added complication, what about passing by message?
Does anyone know of any slick tools for this sort of thing? I have
been looking lately at reverse engineering inherited legacy code,
running it through DoxyGen, producing function call trees and class
inheritance trees, but haven't yet given too much thought to data,
which is probably a big mistake.