tanjoodo / sicasm Goto Github PK

The tokenization is a group of functions that very badly organized. Make it a separate unit and separate the stages nicely.

This should preferably be done before issue #3.

Currently, the tokenizer assumes there are no spaces in any of the three fields. While this makes sense for the label and opcode field, it does not make any sense for the operands field. As the delimiter in that field is the ,.

This means the whole operands field should be parsed by a different mechanism altogether.

This currently does a lot in terms of parsing the assembly and making sure it is valid. There are still some problems when it comes to generating the object file. Mostly related to the fact the book optimizes for outputting a text file as an object file (which is idiotic imo). Once we fix this we should be able to correctly generate code.

Assembler should choose the best available addressing mode unless prompted otherwise by the BASE directive