Code review on org.jquantlib.util.stdlibc.Std

[Richard Gomes]

Code review on org.jquantlib.util.stdlibc.Std

=============
Relationships
=============
related to http://bugs.launchpad.net/bugs/jquantlib-311
parent of http://bugs.launchpad.net/bugs/jquantlib-129
has duplicate http://bugs.launchpad.net/bugs/jquantlib-235

[Ueli Hofstetter (hofi85)]

hi richard,

same thing here. plz. assign the stuff back to you before you start working on. sry. for saying this.

let me know when you're finished and when i can start implementing the remaining testcases for the std class.

cheers

ueli

[Richard Gomes (frgomes)]

I've done an initial code review aiming:

1. Better comments, headers and documentation referring to
http://gcc.gnu.org/onlinedocs/libstdc++/latest-doxygen/

2. Performance improvements, basically replacing List<Double> by double[]

3. Removal of old code which was intended to interface to fastutil (which is now removed).

4. Standardization of names of parameters. Example: "begin", "from" and "start" have been previously employed. Now I changed everything to "from".

5. "final private" everywhere

Please have a look at the source code and find FIXME tags like these:

1. The use of E_IUnaryFunction and E_IBinaryFunction are pretty confusing.
It's not clear what these interfaces/classes do (where are the comments and references, etc?) and, in particular, E_IBinaryFunction should have 3 generic arguments: 2 parameter types and 1 result type.

2. There's no such "std::multiplies"
This class returns an object which is extended from an abstract class from math package, which is something weird because 2 very related concepts which could not be apart are located in two very different places.

[Ueli Hofstetter (hofi85)]

std.java been made redundant/basically everything has been refactored

[Richard Gomes (frgomes)]

class Std removed.

Methods migrated to Array and/or Date

[Richard Gomes (frgomes)]

Issue reopened.

As translation goes ahead, we found other situations where C-style "std" functions are called.

Zahid created another Std.Java at org.jquantlib.util.

We should keep this file there and should keep this issue open until we reach near 100% of QuantLib translated.