[antlr-interest] Any info/advice on creating "Program Database files" ?

micheal_jor open.zone at virgin.net
Tue Aug 27 18:00:57 PDT 2002


Hi All,

Are there any useful articles/projects that deal with the issue of 
creating program database files - a persistent store of the info 
collected by the parser/semantic analyzer about all the files and 
source code of an application?.

My particular interest is developing metrics tools and other code 
analyzers for a couple of custom languages. We are experiencing 
increases in the number and sizes of the files involved in typical 
projects.

I imagined a scheme whereby the files are parsed and AST built for 
each file and stored (perhaps with additional info) into this program 
database. When a file is changed (or the AST is changed), it should 
be possible to simply merge the changes into the program database (or 
alternatively delete all entries related to the file and replace with 
new entries). The program database will contain the symbol table and 
suchlike too and allow similar changes.

Also, any thoughts on managing the collection of AST (one-per-file) 
collectively given that they ofen won't all fit into available memory 
together?

Micheal



 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ 



More information about the antlr-interest mailing list