The workings of the program can be broken up to 6 steps:

• initialization
• reading of input file
• first processing step
• second processing step
• generating and writing binary data
• quit

Init

• initializing state variables
• parsing commandline arguments
• loading instructions data

Reading

• reading input file
• tokenizing
• stripping whitespace and comments from tokens
• recognizing type (directive/instruction/label)
• result is a linked list with tokens
• can be repeated via .include directives

First step

• label: record position
• instruction: find and link instruction data
• instruction: identify addressmode
• instruction: try parsing operand (but forward-ref labels might be invalid)
• directive: do directive processing
• calculate binsizes
• remove all tokens with binsize=0
• conditional compilation logic
• can repeat reading step via include directive

Second step

• get token operands (now all should be valid)

Third step

• generate and concatenate binary data, write it to file