Comment on revision r5d88604edd51c847a5c743d8fccf6046f37eae52 in pysph

[py...@googlecode.com]

Comment by prabhu.r...@gmail.com:

Line-by-line comments:

File: /source/pysph/base/particle_array.pyx
(r5d88604edd51c847a5c743d8fccf6046f37eae52)
===============================================================================

Line 1433: try:
-------------------------------------------------------------------------------
This code isn't clear at all and the documentation mentions nothing about
constants. Clearly you are sending a bunch of constants along with the
numpy array data. Why do you need the type(props[prop]) == str check? I
really think this part of the code could be better.

For more information:
http://code.google.com/p/pysph/source/detail?r=5d88604edd51c847a5c743d8fccf6046f37eae52

[py...@googlecode.com]

Comment by kunal.r....@gmail.com:

On Thu, Jul 14, 2011 at 1:08 AM, <py...@googlecode.com> wrote:

> Comment by prabhu.ramachandran:

> Line-by-line comments:

> File: /source/pysph/base/particle_**array.pyx (**
> r5d88604edd51c847a5c743d8fccf6**046f37eae52)
> ==============================**==============================**
> ===================

> Line 1433: try:
> ------------------------------**------------------------------**

> -------------------
> This code isn't clear at all and the documentation mentions nothing about
> constants. Clearly you are sending a bunch of constants along with the
> numpy array data. Why do you need the type(props[prop]) == str check? I
> really think this part of the code could be better.

> I needed to do this to properly load the cl_precision attribute for the
particle array. Given that we are not saving it currently, but when we do,
the string variable will have a length but is not an array.