UCHILE

To run the dataVisualizer you must have the following requirements met:

• Install in your LinuxOS or MacOS:
• graphviz
• python3-tk
• python3
• racket
• Create a virtual environement with

$ python3 -m venv venv

and then activate it with

$ source venv/bin/activate

• Install python modules with

$ pip install -r requirements

• Finally run the code

$ python dataVisualizer.py

Use instructions

To run it you must first connect to Racket (actually just call the script) by pressing Start racket. Then Connect to Racket. Write your code and PreProcess it! Set the options for your analysis and write initial variables if you need to. By pressing Do Analysis racket will receive the instructions and await for you to press on Next Step. You can move forward and backwards until its finished. When it is done an alert will pop up and you must restart the racket service to do another analysis

TroubleShooting

If an error occurs or the app never changes to state: connected maybe the socket has not been closed. To force it do

$ fuser -k 9876/tcp

with the port which is 9876 for this app.

Requirements

graphviz python3-tk

Dataflow Analysis

A Racket implementation of traditional dataflow analyses for an imperative language TIP.

Target: the TIP Language

These analyses are targeted for S-Expression based TIP language. The syntax is largely extracted from the great lecture notes Static Program Analysis[1].

An Example

{while {> 5 x}
  {{if {== x 3}
       {:= x 4}
       {:= x 5}}
   {:= x {- x 1}}}}

File Descriptions

• parser.rkt functions that parse s-exp based TIP to abstract syntax tree (AST).
• ast.rkt the abstract syntax tree structure definitions.
• cfg.rkt contrlo flow graph (CFG) structure definitions; CFG is transformed from AST.
• dfa.rkt chaotic iteration framework and algorithm, which operates on CFG.
• reaching-def.rkt reaching definition analysis.
• very-busy.rkt very busy expressions analysis.
• available-expr.rkt available expressions analysis.
• live-var.rkt live variables analysis.

See test cases in each files.

TODO

• SSA-based analysis
• Pointer analysis

References

[1] Static Program Analysis