Subject: SEN1 meeting tomorrow, talk by Anya
From: Bas Basten
To: sen1-list 
Date: Wed, 2 Feb 2011 14:27:14 +0100 (CET)

Dear all,

Tomorrow it's Anya's turn to share her research ideas with us at the SEN1 meeting. This is her abstract:


Program Databases, Compiler Stages and Integrated Editors
---------------------------------------------------------

Classic compiler design uses a series of stages or phases (forming a 
pipeline) to process a program from source code to executable.
In an integrated editing environment such as Eclipse, the process 
becomes more complicated. The editor should be able to give feedback on 
semantic errors, provide context-sensitive help and quick 
jump-to-definition, while still being fast and responsive. The editor 
needs to perform semantic analysis and store relevant information � 
possibly a lot of information, for large projects with thousands of 
modules. Different features may require different analyses, giving a 
less clear 'pipeline' than in traditional compilers.

I've been looking at ways to solve this in a generic way (for Eclipse 
and Rascal), with a program database to store information from various 
stages, and on-demand scheduling of stages. I don't have a perfect 
solution yet (not sure that there is one), but I'd like to report on my 
ideas so far, and maybe see if you guys have some thoughts as well.

---------------------------------------------------------

We start at 10am in room L016. See you there!

Cheers,
Bas