Difference: FRCVLabMLSA_execution (3 vs. 4)

Revision 42017-07-12 - AnneMarieBogar

Line: 1 to 1

META TOPICPARENT	name="FRCVLabMLSA"

Line: 6 to 6

To run the software:

Changed:

<
<

1. Download software from https://git.io/MLSA

2. Run mlsapath.bash in the mlsa folder (adds mlcg.py to PATH) with command: $source mlsapath.bash

3. On the terminal, run the mlcg command with the desired folder or programs as arguments: $mlcg.py [list of args]

A good way to test if your MLSA installation is operating correctly is to test the code on the test programs in the test folder

1. Run the command: $python testcode.py 0

2. A document titled 'stats' will show if there are any differences between the code you generated and the ideal code

3. Delete the generated code by running the command: $python testcode.py -1

2. Repeat with numbers 1-5

>
>

On the terminal, run the mlcg command with the desired folder or programs as arguments:

$mlcg.py [list of args]

A good place to start is by testing out the code in MLSA's test folder by running the command:

$mlcg.py test/test5

The multilingual pipeline is called for all the programs in test5 and a single call graph is generated. Procedure calling between files in the same and different languages will be identified (for the limited set of interoperability calls that have been implemented) and the call graph will reflect this, but programs with no procedure calls in common are fine too. The resulting call graph is a forest of trees. Recursion is flagged after one full cycle and several other kind of interlanguage calls are also flagged, such as circularity in a system and the inability to determine a program called through one of the three APIs used for interoperability.

Permissions

Persons/group who can view/change the page:

View topic | History: r5 < r4 < r3 < r2 | More topic actions...