Fix example patch tests and stop tracking prev and next tokens

The example patch tests were broken because the parser was
initialized but never run. Additionally, none of the example patches
were complex enough to hit the recursion depth limit. I added a few
longer patches and run the parser, which reproduced the recursion
depth limit within tests.

To fix the recursion limit, which was caused by deep copying tokens,
and by extension all other tokens given that they are doubly linked,
I removed tracking of previous and next tokens. This was initially
implemented to handle multi-word instructions by traversing to
neighboring tokens, but that is now handled by TokenRegistry, and
there are no other apparent use cases. If I find that I need that
functionality in the future, it can be re-implemented and the _clone
method can be rewritten to avoid deeply copying neighboring tokens.

src/spinasm_lsp/parser.py

@@ -52,10 +52,6 @@ class Token:
         The symbol parsed by asfv1 representing the token.
     range : lsp.Range
         The location range of the token in the source file.
-    next_token : Token | None
-        The token that follows this token in the source file.
-    prev_token : Token | None
-        The token that precedes this token in the source file.
     """
 
     def __init__(
@@ -67,8 +63,6 @@ def __init__(
 
         self.symbol: Symbol = symbol
         self.range: lsp.Range = lsp.Range(start=start, end=end)
-        self.next_token: Token | None = None
-        self.prev_token: Token | None = None
 
     def __repr__(self) -> str:
         return self.symbol["stxt"]
@@ -135,11 +129,6 @@ def register_token(self, token: Token) -> None:
         if token.range.start.line not in self._tokens_by_line:
             self._tokens_by_line[token.range.start.line] = []
 
-        # Record the previous and next token for each token to allow traversing
-        if self._prev_token:
-            token.prev_token = self._prev_token
-            self._prev_token.next_token = token
-
         # Store the token on its line
         self._tokens_by_line[token.range.start.line].append(token)
         self._prev_token = token

tests/patches/LICENSE.md

@@ -0,0 +1,5 @@
+The `Basic` and `GA_DEMO` patches are reproduced from the Spin Semiconductor [Guitar Amp Applications](https://www.spinsemi.com/guitar_amp_application.html) written by Keith Barr, Copyright (C) 2018 Spin Semiconductor, Inc.
+
+The `rev_pl` patches are reproduced from the [Free DSP Programs](https://www.spinsemi.com/programs.php) licensed under the Spin Semiconductor Open Reverb License, and were submitted by Harmon Grold.
+
+The `freeverb` patch is reproduced from an [FV-1 port by basilrush](http://www.spinsemi.com/forum/viewtopic.php?t=309) of the public domain [Freeverb algorithm by Jezar](https://github.com/sinshu/freeverb).

tests/patches/freeverb.spn

@@ -0,0 +1,240 @@
+; freeverb from
+; http://www.spinsemi.com/forum/viewtopic.php?t=309
+
+; pot0: combfeedback
+; pot1: 
+; pot2: 
+
+equ comb1filt 32
+equ comb2filt 33
+equ comb3filt 34
+equ comb4filt 35
+equ comb5filt 36
+equ comb6filt 37
+equ comb7filt 38
+equ comb8filt 39
+
+equ combR1filt 32+12
+equ combR2filt 33+12
+equ combR3filt 34+12
+equ combR4filt 35+12
+equ combR5filt 36+12
+equ combR6filt 37+12
+equ combR7filt 38+12
+equ combR8filt 39+12
+
+equ combfeedback pot0
+equ stereospread 23
+equ filtco 0.5
+equ allpassamt 0.5
+mem comb1 1116
+mem comb2 1188
+mem comb3 1277
+mem comb4 1356
+mem comb5 1422
+mem comb6 1491
+mem comb7 1557
+mem comb8 1617
+
+mem combR1 1116+stereospread
+mem combR2 1188+stereospread
+mem combR3 1277+stereospread
+mem combR4 1356+stereospread
+mem combR5 1422+stereospread
+mem combR6 1491+stereospread
+mem combR7 1557+stereospread
+mem combR8 1617+stereospread
+
+mem allpass1 556
+mem allpass2 441
+mem allpass3 341
+mem allpass4 225
+
+mem allpassR1 556+stereospread
+mem allpassR2 441+stereospread
+mem allpassR3 341+stereospread
+mem allpassR4 225+stereospread
+
+; LEFT CHANNEL
+
+;* NOW do freeverb style comb with lowpass filters, sum outputs
+; output = delay output
+; delay-input = [filteredoutput * feedback + input]
+
+; get delay and one pole low pass
+rda comb1#, 1-filtco ; read delay output
+rdax comb1filt, filtco ; filter using one sample delay
+wrax comb1filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra comb1, 0 ; write to the buffer
+
+rda comb2#, 1-filtco ; read delay output
+rdax comb2filt, filtco ; filter using one sample delay
+wrax comb2filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra comb2, 0 ; write to the buffer
+
+rda comb3#, 1-filtco ; read delay output
+rdax comb3filt, filtco ; filter using one sample delay
+wrax comb3filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra comb3, 0 ; write to the buffer
+
+rda comb4#, 1-filtco ; read delay output
+rdax comb4filt, filtco ; filter using one sample delay
+wrax comb4filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra comb4, 0 ; write to the buffer
+
+rda comb5#, 1-filtco ; read delay output
+rdax comb5filt, filtco ; filter using one sample delay
+wrax comb5filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra comb5, 0 ; write to the buffer
+
+rda comb6#, 1-filtco ; read delay output
+rdax comb6filt, filtco ; filter using one sample delay
+wrax comb6filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra comb6, 0 ; write to the buffer
+
+rda comb7#, 1-filtco ; read delay output
+rdax comb7filt, filtco ; filter using one sample delay
+wrax comb7filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra comb7, 0 ; write to the buffer
+
+rda comb8#, 1-filtco ; read delay output
+rdax comb8filt, filtco ; filter using one sample delay
+wrax comb8filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra comb8, 0 ; write to the buffer
+
+; RIGHT CHANNEL
+
+;* NOW do freeverb style combR with lowpass filters, sum outputs
+; output = delay output
+; delay-input = [filteredoutput * feedback + input]
+
+; get delay and one pole low pass
+rda combR1#, 1-filtco ; read delay output
+rdax combR1filt, filtco ; filter using one sample delay
+wrax combR1filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra combR1, 0 ; write to the buffer
+
+rda combR2#, 1-filtco ; read delay output
+rdax combR2filt, filtco ; filter using one sample delay
+wrax combR2filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra combR2, 0 ; write to the buffer
+
+rda combR3#, 1-filtco ; read delay output
+rdax combR3filt, filtco ; filter using one sample delay
+wrax combR3filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra combR3, 0 ; write to the buffer
+
+rda combR4#, 1-filtco ; read delay output
+rdax combR4filt, filtco ; filter using one sample delay
+wrax combR4filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra combR4, 0 ; write to the buffer
+
+rda combR5#, 1-filtco ; read delay output
+rdax combR5filt, filtco ; filter using one sample delay
+wrax combR5filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra combR5, 0 ; write to the buffer
+
+rda combR6#, 1-filtco ; read delay output
+rdax combR6filt, filtco ; filter using one sample delay
+wrax combR6filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra combR6, 0 ; write to the buffer
+
+rda combR7#, 1-filtco ; read delay output
+rdax combR7filt, filtco ; filter using one sample delay
+wrax combR7filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra combR7, 0 ; write to the buffer
+
+rda combR8#, 1-filtco ; read delay output
+rdax combR8filt, filtco ; filter using one sample delay
+wrax combR8filt, 1 ; store one sample delay, scale by feedback amount
+mulx combfeedback
+rdax adcl, 0.5 ; add the input
+wra combR8, 0 ; write to the buffer
+
+
+;LEFT CHANNEL
+
+;* NOW sum outputs
+rda comb1#, 1
+rda comb2#, 1
+rda comb3#, 1
+rda comb4#, 1
+rda comb5#, 1
+rda comb6#, 1
+rda comb7#, 1
+rda comb8#, 1
+
+; feed through allpass filters in series
+
+; get delay -> add to -1*acc ->
+
+rda allpass1#, -allpassamt
+wrap allpass1, allpassamt
+rda allpass2#, -allpassamt
+wrap allpass2, allpassamt
+rda allpass3#, -allpassamt
+wrap allpass3, allpassamt
+rda allpass4#, -allpassamt
+wrap allpass4, allpassamt
+
+
+;** OUTPUT
+wrax dacl, 0
+
+; RIGHT CHANNEL
+
+;* NOW sum outputs
+rda combR1#, 1
+rda combR2#, 1
+rda combR3#, 1
+rda combR4#, 1
+rda combR5#, 1
+rda combR6#, 1
+rda combR7#, 1
+rda combR8#, 1
+
+; feed through allpass filters in series
+
+; get delay -> add to -1*acc ->
+
+rda allpassR1#, -allpassamt
+wrap allpassR1, allpassamt
+rda allpassR2#, -allpassamt
+wrap allpassR2, allpassamt
+rda allpassR3#, -allpassamt
+wrap allpassR3, allpassamt
+;rda allpassR4#, -allpassamt
+;wrap allpassR4, allpassamt
+
+;** OUTPUT
+wrax dacr, 0