> This is very cool. Can it (internally, at least) just be called "inline" or
> "Inline" though? There's no particular need to distinguish it from Perl's
> Inline module within the scope of Python programs, is there?
No, there really isn't any need to call the module PyInline. However, I
didn't want to call the project Inline because that would conflict with
Brian's work and to call the module something different than the project
would perhaps be confusing.. Thoughts??
> Were you able to leverage any of the techniqueus used in Brian Ingerson's
> work?
I looked at Brian's Inline to figure out PyInline's high level design and
also to figure out how one builds arbitrary chunks of "anonymous" code
(i.e., one idea is to build the code in a subdirectory which has in its
name the MD5 checksum of the code).
Brian's Inline also provides a few abstractions which I will be adding to
PyInline in future. In particular, PyInline currently only supports C.
In order to support the addition of more languages, I need to abstract out
the concept of building an arbitrary chunk of code in some language and
then "binding" to it.
TTUL
Ken
|
