1. Tracer
- Posted by mistertrik at hotmail.com Oct 24, 2002
- 436 views
I like the trace utility, and I especially like the ability to check up on variables by using the '?' key. However, my problem at the moment is that there is an error deep within a large sequence. It's not a run-time error, but it produces an undesirable result. I can't see that particular entry, because all that you can see of the sequence is the first few members. I'd like to see extended support in the trace facility, to specify an entry or a slice of a variable, or even do multi-dimension databases. A further change, if possible... would be to allow lines that reference other lines - ie mysequence[row][col]. Ultimately, it would be nice if it could support any function that doesn't change variables outside the scope of the function. That might seem a bit strange... but consider this. Inside tracer window: if a variable is entered via the ? button, eg template[line] an entry would appear below as "template[line]=<it's value>" if a non-variable is entered via the ? button, it checks to see if it will change any variables outside the function, and if not will allow it. It puts the result in the window as "Result=<it's value>". Result would be a simple placeholder, and could not be called by a successive ? entry. Examples... template[line - 1] * template[line] would recognise that there aren't any function calls, and return that amount, under the moniker Result= template[length(template)] would recognise the presence of a function call, and check the function 'length' for any commands that would change a variable. If there aren't any, then it returns that amount, under the moniker Result= If it fails it can just return an 'disallowed' message. My $0.02 ===================================================== .______<-------------------\__ / _____<--------------------__|=== ||_ <-------------------/ \__| Mr Trick
2. Re: Tracer
- Posted by Robert Craig <rds at RapidEuphoria.com> Oct 24, 2002
- 460 views
Mr. Trick writes: > However, my problem at the moment is that there is an > error deep within a large sequence. It's not a run-time error, > but it produces an undesirable result. I can't see that > particular entry, because all that you can see of the sequence > is the first few members. I'd like to see extended support in > the trace facility, to specify an entry or a slice of a variable, > or even do multi-dimension databases. I'm going to make some improvements in this area, but for now you should try the pretty_print() routine that I've added to misc.e (see Recent User Contributions "Faster EDS"). You could do something like: f = open("debug", "w") pretty_print(f, x, {}) close(f) where x is your variable. It will print your variable, no matter how big or complicated, in a nice readable way. Regards, Rob Craig Rapid Deployment Software http://www.RapidEuphoria.com