[EMAIL PROTECTED] a écrit : > Most unclear. My apologies. > > I'm trying to structure a tokenizer. The stupid concatenations are > just placeholders for the actual tokenizing work. By rebuilding the > input they demonstrate that the framework correctly processes all the > input. > > I'm currently using a C-style design (my own pointers into an array of > strings) but I suspect I've got the solution an old C guy would have. > As a Python newbie I don't yet think in Python.
Then the first move is to carefully eval existing solutions: http://wiki.python.org/moin/LanguageParsing -- http://mail.python.org/mailman/listinfo/python-list