Flume Parsing Improved

Post date: Sep 10, 2010 8:3:1 PM

I found a bug in the parsing of Flume that led to problems handling scientific notation numbers (1.0e-2). The solution helped me realize I could make the handling of x or X more intelligent. I allow X to represent the multiplication operator (along with the C-standard '*'). Before, you had to have spaces around the X, otherwise Flume thought it might be part of a hexadecimal number (0x123ABC). But now, Flume only assumes x or X is part of a hexadecimal number if immediately preceded by zero. So now "6x7" expands nicely into "6 x 7" which is correctly multiplied, while "0x7" still represents the hex number 7. The fix for scientific numbers ensures they are recognized as such and correctly accept negative exponents and not assume the negative is a subtraction operator. Changes are in version the new ALPHA release 0.2.6.

Discuss Flume at the Tiny Core Linux Forum.