Thursday, December 27, 2018

How is the new tokenizer?

I have the code done, starting test.  The design is all driven by the current set of punctuation list and the client can =swap out those operator lists.  There are a couple of operator strings, the eol strings identifies the end of line markers. Then we have expr marks as well as file name marks and expression marks, all four can be set independently in one call using a mask and a list of stings pointers. I use loadable address space, so in fact, the client can actually change them independent of a console command.

Call this version 2 of the intelligent tokenizer, and wait until version 3 before concluding it works completely. I need another day of test for the version 2.  The double ops (--, ++, += ,-= ,/= ,*=.>>,<<, &&.||)  are not working yet. I have no figured out the double ops yet.

It is not glamorous work, not like Hardly working.

No comments: