Ophis/doc/tutor6.sgm

<chapter>
  <title>Expressions</title>

  <para>
    Ophis permits a reasonably rich set of arithmetic operations to be
    done at assemble time.  So far, all of our arguments and values
    have either been constants or label names.  In this chapter, we
    will modify the <literal>print</literal> macro so that it calls a
    subroutine to do the actual printing.  This will shrink the final
    code size a fair bit.
  </para>

  <para>
    Here's our printing routine.  It's fairly straightforward.
  </para>

<programlisting>
; PRINTSTR routine.  Accumulator stores the low byte of the address,
; X register stores the high byte.  Destroys the values of $10 and
; $11.

.scope
printstr:
        sta $10
        stx $11
        ldy #$00
_lp:    lda ($10), y
        beq _done
        jsr chrout
        iny
        bne _lp
_done:  rts
.scend
</programlisting>

  <para>
    However, now we are faced with the problem of what to do with
    the <literal>print</literal> macro.  We need to take a 16-bit
    value and store it in two 8-bit registers.  We can use
    the <literal>&lt;</literal> and <literal>&gt;</literal> operators
    to take the low or high byte of a word, respectively.
    The <literal>print</literal> macro becomes:
  </para>

  <programlisting>
.macro print
        lda #&lt;_1
        ldx #&gt;_1
        jsr printstr
.macend
</programlisting>

  <para>
    Also, since BASIC uses the locations $10 and $11, we should really
    cache them at the start of the program and restore them at the
    end:
  </para>

  <programlisting>
.data
.org $C000
.space cache 2
.text

        ; Save the zero page locations that printstr uses.
        lda $10
        sta cache
        lda $11
        sta cache+1

        ; ... main program goes here ...

        ; Restore the zero page values printstr uses.
        lda cache
        sta $10
        lda cache+1
        sta $11
</programlisting>

  <para>
    Note that we only have to name <literal>cache</literal> once, but
    can use addition to refer to any offset from it.
  </para>

  <para>
    Ophis supports following operations, with the following precedence
    levels (higher entries bind more tightly):
  </para>

  <table frame="all">
    <title>Ophis Operators</title>
    <tgroup cols='2'>
      <thead>
        <row>
          <entry align="center">Operators</entry>
          <entry align="center">Description</entry>
        </row>
      </thead>
      <tbody>
        <row><entry><literal>[ ]</literal></entry><entry>Parenthesized expressions</entry></row>
        <row><entry><literal>&lt; &gt;</literal></entry><entry>Byte selection (low, high)</entry></row>
        <row><entry><literal>* /</literal></entry><entry>Multiply, divide</entry></row>
        <row><entry><literal>+ -</literal></entry><entry>Add, subtract</entry></row>
        <row><entry><literal>| &amp; ^</literal></entry><entry>Bitwise OR, AND, XOR</entry></row>
      </tbody>
    </tgroup>
  </table>
  <para>
    Note that brackets, not parentheses, are used to group arithmetic
    operations.  This is because parentheses are used for the indirect
    addressing modes, and it makes parsing much easier.
  </para>

  <para>
    The code for this version of the code is
    in <xref linkend="tutor6-src" endterm="tutor6-fname">.
  </para>
</chapter>