Rewrite assembler

This is an almost complete rewrite of the entire assembler from the
ground up, with a different parsing strategy and a whole new symbol
resolution mechanism for automatically including library files.

The assembly syntax has also been slightly modified, with padding
tokens now being prefixed with '#' instead of '$', and a block-style
anonymous-label syntax which uses the '{' and '}' characters.

1 files changed, 90 insertions, 0 deletions

diff --git a/src/tokens/semantic.rs b/src/tokens/semantic.rs
new file mode 100644
index 0000000..ac5179c
--- /dev/null
+++ b/src/tokens/semantic.rs
@@ -0,0 +1,90 @@
+use crate::*;
+
+use SemanticTokenVariant as SemVar;
+
+
+pub struct SemanticToken {
+    pub source: SourceSpan,
+    pub bytecode: BytecodeSpan,
+    pub variant: SemanticTokenVariant,
+}
+
+
+pub enum SemanticTokenVariant {
+    LabelDefinition(LabelDefinition),
+    MacroDefinition(MacroDefinition),
+
+    /// Pointer to the matching label definition.
+    LabelReference(usize),
+    /// Pointer to the matching macro definition.
+    MacroInvocation(usize),
+
+    Literal(Value),
+    Padding(Value),
+    Instruction(Instruction),
+
+    Comment(String),
+    String(Vec<u8>),
+
+    /// Pointer to the matching block close.
+    BlockOpen(usize),
+    /// Pointer to the matching block open.
+    BlockClose(usize),
+    MarkOpen,
+    MarkClose,
+
+    Error(SemanticParseError),
+}
+
+impl std::fmt::Debug for SemanticToken {
+    fn fmt(&self, f: &mut std::fmt::Formatter) -> Result<(), std::fmt::Error> {
+        match &self.variant {
+            SemVar::LabelDefinition(def) => write!(f, "LabelDefinition({})", def.name),
+            SemVar::MacroDefinition(def) => write!(f, "MacroDefinition({})", def.name),
+            SemVar::LabelReference(pointer) => write!(f, "LabelReference(*{pointer})"),
+            SemVar::MacroInvocation(pointer) => write!(f, "MacroInvocation(*{pointer})"),
+            SemVar::Literal(value) => write!(f, "Literal({value})"),
+            SemVar::Padding(value) => write!(f, "Padding({value})"),
+            SemVar::Instruction(instr) => write!(f, "Instruction(0x{:02x})", instr.value),
+            SemVar::Comment(comment) => write!(f, "Comment({comment})"),
+            SemVar::String(string) => write!(f, "String({})", String::from_utf8_lossy(&string)),
+            SemVar::BlockOpen(_) => write!(f, "BlockOpen"),
+            SemVar::BlockClose(_) => write!(f, "BlockClose"),
+            SemVar::MarkOpen => write!(f, "MarkOpen"),
+            SemVar::MarkClose => write!(f, "MarkClose"),
+            SemVar::Error(_) => write!(f, "Error"),
+        }
+    }
+}
+
+
+pub struct LabelDefinition {
+    /// The absolute name of the label or sublabel.
+    pub name: String,
+    /// List of pointers to label reference tokens.
+    pub references: Vec<usize>,
+}
+
+
+pub struct MacroDefinition {
+    pub name: String,
+    pub references: Vec<usize>,
+    pub body_tokens: Vec<SemanticToken>,
+}
+
+
+pub enum SemanticParseError {
+    LabelDefinitionInMacroDefinition,
+    MacroDefinitionInMacroDefinition,
+
+    StrayMacroTerminator,
+    StrayBlockClose,
+    UnclosedBlock,
+
+    UndefinedSymbol(String),
+    RedefinedSymbol((String, SourceSpan)),
+
+    MacroInvocationBeforeDefinition((String, SourceSpan)),
+
+    SyntaxError(SyntacticParseError)
+}