added string value assignment, leftstr, rightstr, substr functions

compiler/res/prog8lib/prog8lib.asm

@@ -2078,3 +2078,127 @@ ror2_array_uw	.proc
 		sta  (c64.SCRATCH_ZPWORD1),y
 +		rts
 		.pend
+
+		
+strcpy		.proc
+		; copy a string (0-terminated) from A/Y to (ZPWORD1)
+		; it is assumed the target string is large enough.
+		sta  c64.SCRATCH_ZPWORD2
+		sty  c64.SCRATCH_ZPWORD2+1
+		ldy  #$ff
+-		iny
+		lda  (c64.SCRATCH_ZPWORD2),y
+		sta  (c64.SCRATCH_ZPWORD1),y
+		bne  -
+		rts
+		.pend
+		
+		
+func_leftstr	.proc
+		; leftstr(source, target, length) with params on stack
+		inx
+		lda  c64.ESTACK_LO,x
+		tay			; length
+		inx
+		lda  c64.ESTACK_LO,x
+		sta  c64.SCRATCH_ZPWORD2
+		lda  c64.ESTACK_HI,x
+		sta  c64.SCRATCH_ZPWORD2+1
+		inx
+		lda  c64.ESTACK_LO,x
+		sta  c64.SCRATCH_ZPWORD1
+		lda  c64.ESTACK_HI,x
+		sta  c64.SCRATCH_ZPWORD1+1
+		lda  #0
+		sta  (c64.SCRATCH_ZPWORD2),y
+-		dey
+		cpy  #$ff
+		bne  +
+		rts
++		lda  (c64.SCRATCH_ZPWORD1),y
+		sta  (c64.SCRATCH_ZPWORD2),y
+		jmp  -
+		.pend
+
+func_rightstr	.proc
+		; rightstr(source, target, length) with params on stack
+		; make place for the 4 parameters for substr()
+		dex
+		dex
+		dex
+		dex
+		; X-> .
+		; x+1 -> length of segment
+		; x+2 -> start index
+		; X+3 -> target LO+HI
+		; X+4 -> source LO+HI
+		; original parameters:
+		;  x+5 -> original length LO
+		;  x+6 -> original targetLO + HI
+		;  x+7 -> original sourceLO + HI
+		; replicate paramters:
+		lda  c64.ESTACK_LO+5,x
+		sta  c64.ESTACK_LO+1,x
+		lda  c64.ESTACK_LO+6,x
+		sta  c64.ESTACK_LO+3,x
+		lda  c64.ESTACK_HI+6,x
+		sta  c64.ESTACK_HI+3,x
+		lda  c64.ESTACK_LO+7,x
+		sta  c64.ESTACK_LO+4,x
+		sta  c64.SCRATCH_ZPWORD1
+		lda  c64.ESTACK_HI+7,x
+		sta  c64.ESTACK_HI+4,x
+		sta  c64.SCRATCH_ZPWORD1+1
+		; determine string length
+		ldy  #0
+-		lda  (c64.SCRATCH_ZPWORD1),y
+		beq  +
+		iny
+		bne  -
++		tya
+		sec
+		sbc  c64.ESTACK_LO+1,x  ; start index = strlen - segment length
+		sta  c64.ESTACK_LO+2,x
+		jsr  func_substr
+		; unwind original params
+		inx
+		inx
+		inx
+		rts
+		.pend
+
+func_substr	.proc
+		; substr(source, target, start, length) with params on stack
+		inx
+		ldy  c64.ESTACK_LO,x	; length
+		inx
+		lda  c64.ESTACK_LO,x	; start
+		sta  c64.SCRATCH_ZPB1
+		inx
+		lda  c64.ESTACK_LO,x	
+		sta  c64.SCRATCH_ZPWORD2
+		lda  c64.ESTACK_HI,x
+		sta  c64.SCRATCH_ZPWORD2+1
+		inx
+		lda  c64.ESTACK_LO,x	
+		sta  c64.SCRATCH_ZPWORD1
+		lda  c64.ESTACK_HI,x
+		sta  c64.SCRATCH_ZPWORD1+1
+		; adjust src location
+		clc
+		lda  c64.SCRATCH_ZPWORD1
+		adc  c64.SCRATCH_ZPB1
+		sta  c64.SCRATCH_ZPWORD1
+		bcc  +
+		inc  c64.SCRATCH_ZPWORD1+1
++		lda  #0
+		sta  (c64.SCRATCH_ZPWORD2),y
+		jmp  _startloop
+-		lda  (c64.SCRATCH_ZPWORD1),y
+		sta  (c64.SCRATCH_ZPWORD2),y
+_startloop	dey
+		cpy  #$ff
+		bne  -
+		rts
+
+		.pend

compiler/res/version.txt

@@ -1 +1 @@
-2.1
+2.2-SNAPSHOT

compiler/src/prog8/compiler/target/c64/codegen/AssignmentAsmGen.kt

@@ -773,6 +773,21 @@ internal class AssignmentAsmGen(private val program: Program, private val errors
                     }
                 }
             }
+            DataType.STR -> {
+                val identifier = assign.value as? IdentifierReference
+                        ?: throw AssemblyError("string value assignment expects identifier value")
+                val sourceName = asmgen.asmIdentifierName(identifier)
+                asmgen.out("""
+                    lda  #<$targetName
+                    sta  ${C64Zeropage.SCRATCH_W1}
+                    lda  #>$targetName
+                    sta  ${C64Zeropage.SCRATCH_W1+1}
+                    lda  #<$sourceName
+                    ldy  #>$sourceName
+                    jsr  prog8_lib.strcpy
+                """)
+                return true
+            }
             else -> throw AssemblyError("assignment to identifier: invalid target datatype: $targetType")
         }
         return false

compiler/src/prog8/functions/BuiltinFunctions.kt

@@ -87,7 +87,20 @@ val BuiltinFunctions = mapOf(
                             FParam("address", IterableDatatypes + DataType.UWORD),
                             FParam("numwords", setOf(DataType.UWORD)),
                             FParam("wordvalue", setOf(DataType.UWORD, DataType.WORD))), null),
-    "strlen"      to FSignature(true, listOf(FParam("string", setOf(DataType.STR))), DataType.UBYTE, ::builtinStrlen)
+    "strlen"      to FSignature(true, listOf(FParam("string", setOf(DataType.STR))), DataType.UBYTE, ::builtinStrlen),
+    "substr"      to FSignature(false, listOf(
+            FParam("source", IterableDatatypes + DataType.UWORD),
+            FParam("target", IterableDatatypes + DataType.UWORD),
+            FParam("start", setOf(DataType.UBYTE)),
+            FParam("length", setOf(DataType.UBYTE))), null),
+    "leftstr"      to FSignature(false, listOf(
+            FParam("source", IterableDatatypes + DataType.UWORD),
+            FParam("target", IterableDatatypes + DataType.UWORD),
+            FParam("length", setOf(DataType.UBYTE))), null),
+    "rightstr"      to FSignature(false, listOf(
+            FParam("source", IterableDatatypes + DataType.UWORD),
+            FParam("target", IterableDatatypes + DataType.UWORD),
+            FParam("length", setOf(DataType.UBYTE))), null)
 )
 
 fun builtinMax(array: List<Number>): Number = array.maxBy { it.toDouble() }!!

docs/source/index.rst

@@ -156,7 +156,7 @@ Design principles and features
 - The compiler tries to optimize the program and generated code a bit, but hand-tuning of the
   performance or space-critical parts will likely still be required. This is supported by
   the ability to easily write embedded assembly code directly in the program source code.
-- There are many built-in functions, such as ``sin``, ``cos``, ``rnd``, ``abs``, ``min``, ``max``, ``sqrt``, ``msb``, ``rol``, ``ror``, ``swap``, ``memset``, ``memcopy``, ``sort`` and ``reverse``
+- There are many built-in functions, such as ``sin``, ``cos``, ``rnd``, ``abs``, ``min``, ``max``, ``sqrt``, ``msb``, ``rol``, ``ror``, ``swap``, ``memset``, ``memcopy``, ``substr``, ``sort`` and ``reverse`` (and others)
 - Assembling the generated code into a program wil be done by an external cross-assembler tool.
 
 

docs/source/programming.rst

@@ -279,16 +279,23 @@ This @-prefix can also be used for character byte values.
 
 
 You can concatenate two string literals using '+' (not very useful though) or repeat
-a string literal a given number of times using '*'::
+a string literal a given number of times using '*'. You can also assign a new string
+value to another string. No bounds check is done so be sure the destination string is
+large enough to contain the new value::
 
     str string1 = "first part" + "second part"
     str string2 = "hello!" * 10
 
+    string1 = string2
+    string1 = "new value"
+
 
 .. caution::
-    Avoid changing strings after they've been created.
+    It's probably best to avoid changing strings after they've been created. This
+    includes changing certain letters by index, or by assigning a new value, or by
+    modifying the string via other means for example ``substr`` function and its cousins.
     This is because if your program exits and is restarted (without loading it again),
-    it will then start working with the changed strings instead of the original ones.
+    it will then start working with the changed strings instead of the original ones!
     The same is true for arrays.
 
 
@@ -802,6 +809,22 @@ memsetw(address, numwords, wordvalue)
     Efficiently set a part of memory to the given (u)word value.
     But the most efficient will always be to write a specialized fill routine in assembly yourself!
 
+leftstr(source, target, length)
+    Copies the left side of the source string of the given length to target string.
+    It is assumed the target string buffer is large enough to contain the result.
+    Modifies in-place, doesn't return a value (so can't be used in an expression).
+
+rightstr(source, target, length)
+    Copies the right side of the source string of the given length to target string.
+    It is assumed the target string buffer is large enough to contain the result.
+    Modifies in-place, doesn't return a value (so can't be used in an expression).
+
+substr(source, target, start, length)
+    Copies a segment from the source string, starting at the given index,
+    and of the given length to target string.
+    It is assumed the target string buffer is large enough to contain the result.
+    Modifies in-place, doesn't return a value (so can't be used in an expression).
+
 swap(x, y)
     Swap the values of numerical variables (or memory locations) x and y in a fast way.
 

docs/source/todo.rst

@@ -3,7 +3,6 @@ TODO
 ====
 
 - finalize (most) of the still missing "new" assignment asm code generation
-
 - aliases for imported symbols for example perhaps '%alias print = c64scr.print'
 - option to load library files from a directory instead of the embedded ones (easier library development/debugging)
 - investigate support for 8bitguy's Commander X16 platform https://murray2.com/forums/commander-x16.9/  and https://github.com/commanderx16/x16-docs

examples/strings.p8

@@ -1,36 +0,0 @@
-%import c64lib
-%zeropage basicsafe
-
-main {
-
-sub start() {
-
-    str s1 = "apple"
-    str s2 = "banana"
-    byte[] a1 = [66,77,88,0]
-    ubyte i1 = 101
-    uword w1 = 000
-
-    c64.STROUT(s1)
-    c64.CHROUT('\n')
-    c64.STROUT(a1)
-    c64.CHROUT('\n')
-
-    c64scr.print("bla\n")
-
-;    c64scr.print_uwhex(s1, true)
-;    w1 = &s1
-;    c64scr.print_uwhex(w1, true)
-;
-;    c64scr.print_uwhex(a1, true)
-;    w1 = &a1
-;    c64scr.print_uwhex(w1, true)
-;
-;    s1 = s1
-;    s1 = s2
-;    s2 = "zzz"
-
-}
-}
-
-