PyPerlish Design History

Table of Contents

• 2000-02-22
• 2000-03-03
• 2001-09-29
• 2001-12-28

2000-02-22

The problem was to help perl programmers get on board python. And to ease the migration of perl cgi's to python.

I'd done a Modula-3 "perlish" module, which had the same goal, so I used those algorithms and test cases. Found the python was much less verbose, and the *args and **keywords were very handy.

Got the basic package working.

Reworked mkpythonproj some more, making the default a simple script and making full projects into GPL'd packages.

Since args are passed by reference, can't set values in, e.g.:

  p.open(MYFILE,"< myfile.dat") or p.die("cannot open file")

So I'll have to stay with the python semantics:

  MYFILE=p.open("< myfile.dat")
  if not MYFILE:
    p.die("cannot open file")

2000-03-03

Worked out expansion of '$' and '@' in text strings. This is done in "_ex", which is called internally by pyperl functions. Thus, it needs to look back two levels (to the pyperl ftn and then to the original caller). Made it available externally via the "__call__", since it is a major value of the perl syntax. That is, I'd use it even if no other perl-ism.

Reworked the treatment of file operators, to allow for invalid paths.

2001-09-29

2001-12-28

Convert to unittest and pydoc styles. Discovered that "split" was not properly responding to maxcnt=0, so fixed that. Moved documentation back to simple text files.