CSY/reowolf Changeset - 291ee7a49c03 · Centrum Wiskunde & Informatica (CWI)

Changeset - 291ee7a49c03

Parent rev.

Child rev.

[Not reviewed]

0 8 0

mh - 3 years ago 2022-02-15 14:58:17
contact@maxhenger.nl

Add concept of TypeId for builtin types as well

8 files changed with 137 insertions and 69 deletions:

src/collections/string_pool.rs

src/protocol/ast.rs

src/protocol/ast_printer.rs

src/protocol/input_source.rs

src/protocol/mod.rs

src/protocol/parser/mod.rs

src/protocol/parser/pass_typing.rs

src/protocol/parser/type_table.rs

0 comments (0 inline, 0 general)

src/collections/string_pool.rs

➞

Show inline comments

@@ @@ -10,49 +10,49 @@ pub struct StringRef<'a> { @@
     data: *const u8,
     length: usize,
     _phantom: PhantomData<&'a [u8]>,
+}
 // As the StringRef is an immutable thing:
 unsafe impl Sync for StringRef<'_> {}
 unsafe impl Send for StringRef<'_> {}
 impl<'a> StringRef<'a> {
     /// `new` constructs a new StringRef whose data is not owned by the
     /// `StringPool`, hence cannot have a `'static` lifetime.
     pub(crate) fn new(data: &'a [u8]) -> StringRef<'a> {
         // This is an internal (compiler) function: so debug_assert that the
         // string is valid ascii. Most commonly the input will come from the
         // code's source file, which is checked for ASCII-ness anyway.
         debug_assert!(data.is_ascii());
         let length = data.len();
         let data = data.as_ptr();
         StringRef{ data, length, _phantom: PhantomData }
+    }
     /// `new_empty` creates a empty StringRef. It is a null pointer with a
     /// length of zero.
     pub(crate) fn new_empty() -> StringRef<'static> {
+    pub(crate) const fn new_empty() -> StringRef<'static> {
         StringRef{ data: null(), length: 0, _phantom: PhantomData }
+    }
     pub fn as_str(&self) -> &'a str {
         unsafe {
             let slice = std::slice::from_raw_parts::<'a, u8>(self.data, self.length);
             std::str::from_utf8_unchecked(slice)
+        }
+    }
     pub fn as_bytes(&self) -> &'a [u8] {
         unsafe {
             std::slice::from_raw_parts::<'a, u8>(self.data, self.length)
+        }
+    }
+}
 impl<'a> Debug for StringRef<'a> {
     fn fmt(&self, f: &mut Formatter<'_>) -> FmtResult {
         f.write_str("StringRef{ value: ")?;
         f.write_str(self.as_str())?;
         f.write_str(" }")
+    }
+}

src/protocol/ast.rs

➞

Show inline comments

@@ @@ -324,49 +324,49 @@ pub struct ImportModule { @@
 #[derive(Debug, Clone)]
 pub struct AliasedSymbol {
     pub name: Identifier,
     pub alias: Option<Identifier>,
     pub definition_id: DefinitionId,
+}
 #[derive(Debug, Clone)]
 pub struct ImportSymbols {
     pub this: ImportId,
     // Phase 1: parser
     pub span: InputSpan,
     pub module: Identifier,
     pub module_id: RootId,
     pub symbols: Vec<AliasedSymbol>,
+}
 #[derive(Debug, Clone)]
 pub struct Identifier {
     pub span: InputSpan,
     pub value: StringRef<'static>,
+}
 impl Identifier {
     pub(crate) fn new_empty(span: InputSpan) -> Identifier {
+    pub(crate) const fn new_empty(span: InputSpan) -> Identifier {
         return Identifier{
             span,
             value: StringRef::new_empty(),
         };
+    }
+}
 impl PartialEq for Identifier {
     fn eq(&self, other: &Self) -> bool {
         return self.value == other.value
+    }
+}
 impl Display for Identifier {
     fn fmt(&self, f: &mut Formatter<'_>) -> fmt::Result {
         write!(f, "{}", self.value.as_str())
+    }
+}
 #[derive(Debug, Clone, PartialEq, Eq)]
 pub enum ParserTypeVariant {
     // Special builtin, only usable by the compiler and not constructable by the
     // programmer
     Void,
@@ @@ -489,67 +489,68 @@ impl<'a> Iterator for ParserTypeIter<'a> { @@
+        }
         debug_assert!(depth == 0, "illegally constructed ParserType: {:?}", self.elements);
         return Some(&self.elements[start_element..self.cur_embedded_idx]);
+    }
+}
 /// ConcreteType is the representation of a type after the type inference and
 /// checker is finished. These are fully typed.
 #[derive(Debug, Clone, Copy, Eq, PartialEq, Hash)]
 pub enum ConcreteTypePart {
     // Special types (cannot be explicitly constructed by the programmer)
     Void,
     // Builtin types without nested types
     Message,
     Bool,
     UInt8, UInt16, UInt32, UInt64,
     SInt8, SInt16, SInt32, SInt64,
     Character, String,
     // Builtin types with one nested type
     Array,
     Slice,
     Input,
     Output,
     Pointer,
     // Tuple: variable number of nested types, will never be 1
     Tuple(u32),
     // User defined type with any number of nested types
     Instance(DefinitionId, u32),    // instance of data type
     Function(DefinitionId, u32),    // instance of function
     Component(DefinitionId, u32),   // instance of a connector
+}
 impl ConcreteTypePart {
     pub(crate) fn num_embedded(&self) -> u32 {
         use ConcreteTypePart::*;
         match self {
             Void | Message | Bool |
             UInt8 | UInt16 | UInt32 | UInt64 |
             SInt8 | SInt16 | SInt32 | SInt64 |
             Character | String =>
 ,
             Array | Slice | Input | Output =>
+            Array | Slice | Input | Output | Pointer =>
 ,
             Tuple(num_embedded) => *num_embedded,
             Instance(_, num_embedded) => *num_embedded,
             Function(_, num_embedded) => *num_embedded,
             Component(_, num_embedded) => *num_embedded,
+        }
+    }
+}
 #[derive(Debug, Clone, Eq, PartialEq)]
 pub struct ConcreteType {
     pub(crate) parts: Vec<ConcreteTypePart>
+}
 impl Default for ConcreteType {
     fn default() -> Self {
         Self{ parts: Vec::new() }
+    }
+}
 impl ConcreteType {
     /// Returns an iterator over the subtrees that are type arguments (e.g. an
     /// array element's type, or a polymorphic type's arguments) to the
     /// provided parent type (specified by its index in the `parts` array).
@@ @@ -614,48 +615,52 @@ impl ConcreteType { @@
             CTP::UInt32 => { target.push_str(KW_TYPE_UINT32_STR); },
             CTP::UInt64 => { target.push_str(KW_TYPE_UINT64_STR); },
             CTP::SInt8 => { target.push_str(KW_TYPE_SINT8_STR); },
             CTP::SInt16 => { target.push_str(KW_TYPE_SINT16_STR); },
             CTP::SInt32 => { target.push_str(KW_TYPE_SINT32_STR); },
             CTP::SInt64 => { target.push_str(KW_TYPE_SINT64_STR); },
             CTP::Character => { target.push_str(KW_TYPE_CHAR_STR); },
             CTP::String => { target.push_str(KW_TYPE_STRING_STR); },
             CTP::Array | CTP::Slice => {
                 idx = Self::render_type_part_at(parts, heap, idx, target);
                 target.push_str("[]");
             },
             CTP::Input => {
                 target.push_str(KW_TYPE_IN_PORT_STR);
                 target.push('<');
                 idx = Self::render_type_part_at(parts, heap, idx, target);
                 target.push('>');
             },
             CTP::Output => {
                 target.push_str(KW_TYPE_OUT_PORT_STR);
                 target.push('<');
                 idx = Self::render_type_part_at(parts, heap, idx, target);
                 target.push('>');
             },
             CTP::Pointer => {
                 target.push('*');
                 idx = Self::render_type_part_at(parts, heap, idx, target);
+            }
             CTP::Tuple(num_parts) => {
                 target.push('(');
                 if num_parts != 0 {
                     idx = Self::render_type_part_at(parts, heap, idx, target);
                     for _ in 1..num_parts {
                         target.push(',');
                         idx = Self::render_type_part_at(parts, heap, idx, target);
+                    }
+                }
                 target.push(')');
             },
             CTP::Instance(definition_id, num_poly_args) |
             CTP::Function(definition_id, num_poly_args) |
             CTP::Component(definition_id, num_poly_args) => {
                 let definition = &heap[definition_id];
                 target.push_str(definition.identifier().value.as_str());
                 if num_poly_args != 0 {
                     target.push('<');
                     for poly_arg_idx in 0..num_poly_args {
                         if poly_arg_idx != 0 {
                             target.push(',');
+                        }
                         idx = Self::render_type_part_at(parts, heap, idx, target);

src/protocol/ast_printer.rs

➞

Show inline comments

                         element_idx = write_element(target, heap, t, element_idx + 1);
+                    }
                     target.push('>');
+                }
+            }
+        }
         element_idx
+    }
     write_element(target, heap, t, 0);
+}
 fn write_concrete_type(target: &mut String, heap: &Heap, def_id: DefinitionId, t: &ConcreteType) {
     use ConcreteTypePart as CTP;
     fn write_concrete_part(target: &mut String, heap: &Heap, def_id: DefinitionId, t: &ConcreteType, mut idx: usize) -> usize {
         if idx >= t.parts.len() {
             return idx;
+        }
         match &t.parts[idx] {
             CTP::Void => target.push_str("void"),
             CTP::Message => target.push_str("msg"),
-            CTP::Bool => target.push_str("bool"),
+            CTP::Bool => target.push_str(KW_TYPE_BOOL_STR),
             CTP::UInt8 => target.push_str(KW_TYPE_UINT8_STR),
             CTP::UInt16 => target.push_str(KW_TYPE_UINT16_STR),
             CTP::UInt32 => target.push_str(KW_TYPE_UINT32_STR),
             CTP::UInt64 => target.push_str(KW_TYPE_UINT64_STR),
             CTP::SInt8 => target.push_str(KW_TYPE_SINT8_STR),
             CTP::SInt16 => target.push_str(KW_TYPE_SINT16_STR),
             CTP::SInt32 => target.push_str(KW_TYPE_SINT32_STR),
             CTP::SInt64 => target.push_str(KW_TYPE_SINT64_STR),
             CTP::Character => target.push_str(KW_TYPE_CHAR_STR),
             CTP::String => target.push_str(KW_TYPE_STRING_STR),
             CTP::Pointer => target.push('*'),
             CTP::Array => {
                 idx = write_concrete_part(target, heap, def_id, t, idx + 1);
                 target.push_str("[]");
             },
             CTP::Slice => {
                 idx = write_concrete_part(target, heap, def_id, t, idx + 1);
                 target.push_str("[..]");
+            }
             CTP::Input => {
                 target.push_str("in<");
                 idx = write_concrete_part(target, heap, def_id, t, idx + 1);
                 target.push('>');
             },
             CTP::Output => {
                 target.push_str("out<");
                 idx = write_concrete_part(target, heap, def_id, t, idx + 1);
                 target.push('>')
             },
             CTP::Tuple(num_embedded) => {
                 target.push('(');
                 for idx_embedded in 0..*num_embedded {
                     if idx_embedded != 0 {
                         target.push_str(", ");
+                    }

src/protocol/input_source.rs

➞

Show inline comments

@@ @@ -3,49 +3,49 @@ use std::sync::{RwLock, RwLockReadGuard}; @@
 use std::fmt::Write;
 #[derive(Debug, Clone, Copy)]
 pub struct InputPosition {
     pub line: u32,
     pub offset: u32,
+}
 impl InputPosition {
     pub(crate) fn with_offset(&self, offset: u32) -> Self {
         InputPosition { line: self.line, offset: self.offset + offset }
+    }
+}
 #[derive(Debug, Clone, Copy)]
 pub struct InputSpan {
     pub begin: InputPosition,
     pub end: InputPosition,
+}
 impl InputSpan {
     // This must only be used if you're sure that the span will not be involved
     // in creating an error message.
     #[inline]
     pub fn new() -> InputSpan {
+    pub const fn new() -> InputSpan {
         InputSpan{ begin: InputPosition{ line: 0, offset: 0 }, end: InputPosition{ line: 0, offset: 0 }}
+    }
     #[inline]
     pub fn from_positions(begin: InputPosition, end: InputPosition) -> Self {
         Self { begin, end }
+    }
+}
 /// Wrapper around source file with optional filename. Ensures that the file is
 /// only scanned once.
 pub struct InputSource {
     pub(crate) filename: String,
     pub(crate) input: Vec<u8>,
     // Iteration
     line: u32,
     offset: usize,
     // State tracking
     pub(crate) had_error: Option<ParseError>,
     // The offset_lookup is built on-demand upon attempting to report an error.
     // Only one procedure will actually create the lookup, afterwards only read
     // locks will be held.
     offset_lookup: RwLock<Vec<u32>>,
+}

src/protocol/mod.rs

➞

Show inline comments

@@ @@ -126,49 +126,49 @@ impl ProtocolDescription { @@
         // By now we're sure that all of the arguments are correct. So create
         // the connector.
         return Ok(Prompt::new(&self.types, &self.heap, definition_id, mono_index, arguments));
+    }
     fn lookup_module_root(&self, module_name: &[u8]) -> Option<RootId> {
         for module in self.modules.iter() {
             match &module.name {
                 Some(name) => if name.as_bytes() == module_name {
                     return Some(module.root_id);
                 },
                 None => if module_name.is_empty() {
                     return Some(module.root_id);
+                }
+            }
+        }
         return None;
+    }
     fn verify_same_type(&self, expected: &ConcreteType, expected_idx: usize, arguments: &ValueGroup, argument: &Value) -> bool {
         use ConcreteTypePart as CTP;
         match &expected.parts[expected_idx] {
             CTP::Void | CTP::Message | CTP::Slice | CTP::Function(_, _) | CTP::Component(_, _) => unreachable!(),
+            CTP::Void | CTP::Message | CTP::Slice | CTP::Pointer | CTP::Function(_, _) | CTP::Component(_, _) => unreachable!(),
             CTP::Bool => if let Value::Bool(_) = argument { true } else { false },
             CTP::UInt8 => if let Value::UInt8(_) = argument { true } else { false },
             CTP::UInt16 => if let Value::UInt16(_) = argument { true } else { false },
             CTP::UInt32 => if let Value::UInt32(_) = argument { true } else { false },
             CTP::UInt64 => if let Value::UInt64(_) = argument { true } else { false },
             CTP::SInt8 => if let Value::SInt8(_) = argument { true } else { false },
             CTP::SInt16 => if let Value::SInt16(_) = argument { true } else { false },
             CTP::SInt32 => if let Value::SInt32(_) = argument { true } else { false },
             CTP::SInt64 => if let Value::SInt64(_) = argument { true } else { false },
             CTP::Character => if let Value::Char(_) = argument { true } else { false },
             CTP::String => {
                 // Match outer string type and embedded character types
                 if let Value::String(heap_pos) = argument {
                     for element in &arguments.regions[*heap_pos as usize] {
                         if let Value::Char(_) = element {} else {
                             return false;
+                        }
+                    }
                 } else {
                     return false;
+                }
                 return true;
             },

src/protocol/parser/mod.rs

➞

Show inline comments

@@ @@ -4,149 +4,200 @@ pub(crate) mod tokens; @@
 pub(crate) mod token_parsing;
 pub(crate) mod pass_tokenizer;
 pub(crate) mod pass_symbols;
 pub(crate) mod pass_imports;
 pub(crate) mod pass_definitions;
 pub(crate) mod pass_definitions_types;
 pub(crate) mod pass_validation_linking;
 pub(crate) mod pass_rewriting;
 pub(crate) mod pass_typing;
 pub(crate) mod pass_stack_size;
 mod visitor;
 use tokens::*;
 use crate::collections::*;
 use visitor::Visitor;
 use pass_tokenizer::PassTokenizer;
 use pass_symbols::PassSymbols;
 use pass_imports::PassImport;
 use pass_definitions::PassDefinitions;
 use pass_validation_linking::PassValidationLinking;
 use pass_typing::{PassTyping, ResolveQueue};
 use pass_rewriting::PassRewriting;
 use pass_stack_size::PassStackSize;
 use symbol_table::*;
-use type_table::TypeTable;
+use type_table::*;
 use crate::protocol::ast::*;
 use crate::protocol::input_source::*;
 use crate::protocol::ast_printer::ASTWriter;
 use crate::protocol::parser::type_table::PolymorphicVariable;
 #[derive(Debug, PartialEq, Eq, PartialOrd, Ord)]
 pub enum ModuleCompilationPhase {
     Tokenized,              // source is tokenized
     SymbolsScanned,         // all definitions are linked to their type class
     ImportsResolved,        // all imports are added to the symbol table
     DefinitionsParsed,      // produced the AST for the entire module
     TypesAddedToTable,      // added all definitions to the type table
     ValidatedAndLinked,     // AST is traversed and has linked the required AST nodes
     Typed,                  // Type inference and checking has been performed
     Rewritten,              // Special AST nodes are rewritten into regular AST nodes
     // When we continue with the compiler:
     // StackSize
+}
 pub struct Module {
     // Buffers
     pub source: InputSource,
     pub tokens: TokenBuffer,
     // Identifiers
     pub root_id: RootId,
     pub name: Option<(PragmaId, StringRef<'static>)>,
     pub version: Option<(PragmaId, i64)>,
     pub phase: ModuleCompilationPhase,
+}
 // TODO: This is kind of wrong. Because when we're producing bytecode we would
 //       like the bytecode itself to not have the notion of the size of a pointer
 //       type. But until I figure out what we do want I'll just set everything
 //       to a 64-bit architecture.
 pub struct TargetArch {
     pub array_size_alignment: (usize, usize),
     pub slice_size_alignment: (usize, usize),
     pub string_size_alignment: (usize, usize),
     pub port_size_alignment: (usize, usize),
     pub pointer_size_alignment: (usize, usize),
     pub void_type_id: TypeId,
     pub message_type_id: TypeId,
     pub bool_type_id: TypeId,
     pub uint8_type_id: TypeId,
     pub uint16_type_id: TypeId,
     pub uint32_type_id: TypeId,
     pub uint64_type_id: TypeId,
     pub sint8_type_id: TypeId,
     pub sint16_type_id: TypeId,
     pub sint32_type_id: TypeId,
     pub sint64_type_id: TypeId,
     pub char_type_id: TypeId,
     pub string_type_id: TypeId,
     pub array_type_id: TypeId,
     pub slice_type_id: TypeId,
     pub input_type_id: TypeId,
     pub output_type_id: TypeId,
     pub pointer_type_id: TypeId,
+}
 impl TargetArch {
     fn new() -> Self {
         return Self{
             void_type_id: TypeId::new_invalid(),
             bool_type_id: TypeId::new_invalid(),
             message_type_id: TypeId::new_invalid(),
             uint8_type_id: TypeId::new_invalid(),
             uint16_type_id: TypeId::new_invalid(),
             uint32_type_id: TypeId::new_invalid(),
             uint64_type_id: TypeId::new_invalid(),
             sint8_type_id: TypeId::new_invalid(),
             sint16_type_id: TypeId::new_invalid(),
             sint32_type_id: TypeId::new_invalid(),
             sint64_type_id: TypeId::new_invalid(),
             char_type_id: TypeId::new_invalid(),
             string_type_id: TypeId::new_invalid(),
             array_type_id: TypeId::new_invalid(),
             slice_type_id: TypeId::new_invalid(),
             input_type_id: TypeId::new_invalid(),
             output_type_id: TypeId::new_invalid(),
             pointer_type_id: TypeId::new_invalid(),
+        }
+    }
+}
 pub struct PassCtx<'a> {
     heap: &'a mut Heap,
     symbols: &'a mut SymbolTable,
     pool: &'a mut StringPool,
     arch: &'a TargetArch,
+}
 pub struct Parser {
     // Storage of all information created/gathered during compilation.
     pub(crate) heap: Heap,
     pub(crate) string_pool: StringPool, // Do not deallocate, holds all strings
     pub(crate) modules: Vec<Module>,
     pub(crate) symbol_table: SymbolTable,
     pub(crate) type_table: TypeTable,
     // Compiler passes, used as little state machine that keep their memory
     // around.
     pass_tokenizer: PassTokenizer,
     pass_symbols: PassSymbols,
     pass_import: PassImport,
     pass_definitions: PassDefinitions,
     pass_validation: PassValidationLinking,
     pass_typing: PassTyping,
     pass_rewriting: PassRewriting,
     pass_stack_size: PassStackSize,
     // Compiler options
     pub write_ast_to: Option<String>,
     pub(crate) arch: TargetArch,
+}
 impl Parser {
     pub fn new() -> Self {
         let mut parser = Parser{
             heap: Heap::new(),
             string_pool: StringPool::new(),
             modules: Vec::new(),
             symbol_table: SymbolTable::new(),
             type_table: TypeTable::new(),
             pass_tokenizer: PassTokenizer::new(),
             pass_symbols: PassSymbols::new(),
             pass_import: PassImport::new(),
             pass_definitions: PassDefinitions::new(),
             pass_validation: PassValidationLinking::new(),
             pass_typing: PassTyping::new(),
             pass_rewriting: PassRewriting::new(),
             pass_stack_size: PassStackSize::new(),
             write_ast_to: None,
             arch: TargetArch {
                 array_size_alignment: (3*8, 8), // pointer, length, capacity
                 slice_size_alignment: (2*8, 8), // pointer, length
                 string_size_alignment: (3*8, 8), // pointer, length, capacity
                 port_size_alignment: (3*4, 4), // two u32s: connector + port ID
                 pointer_size_alignment: (8, 8),
+            }
             arch: TargetArch::new(),
         };
         parser.symbol_table.insert_scope(None, SymbolScope::Global);
         // Insert builtin types
         // TODO: At some point use correct values for size/alignment
         parser.arch.void_type_id    = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::Void], false, 0, 1);
         parser.arch.message_type_id = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::Message], false, 24, 8);
         parser.arch.bool_type_id    = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::Bool], false, 1, 1);
         parser.arch.uint8_type_id   = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::UInt8], false, 1, 1);
         parser.arch.uint16_type_id  = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::UInt16], false, 2, 2);
         parser.arch.uint32_type_id  = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::UInt32], false, 4, 4);
         parser.arch.uint64_type_id  = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::UInt64], false, 8, 8);
         parser.arch.sint8_type_id   = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::SInt8], false, 1, 1);
         parser.arch.sint16_type_id  = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::SInt16], false, 2, 2);
         parser.arch.sint32_type_id  = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::SInt32], false, 4, 4);
         parser.arch.sint64_type_id  = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::SInt64], false, 8, 8);
         parser.arch.char_type_id    = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::Character], false, 4, 4);
         parser.arch.string_type_id  = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::String], false, 24, 8);
         parser.arch.array_type_id   = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::Array, ConcreteTypePart::Void], true, 24, 8);
         parser.arch.slice_type_id   = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::Slice, ConcreteTypePart::Void], true, 16, 4);
         parser.arch.input_type_id   = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::Input, ConcreteTypePart::Void], true, 8, 8);
         parser.arch.output_type_id  = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::Output, ConcreteTypePart::Void], true, 8, 8);
         parser.arch.pointer_type_id = insert_builtin_type(&mut parser.type_table, vec![ConcreteTypePart::Pointer, ConcreteTypePart::Void], true, 8, 8);
         // Insert builtin functions
         fn quick_type(variants: &[ParserTypeVariant]) -> ParserType {
             let mut t = ParserType{ elements: Vec::with_capacity(variants.len()), full_span: InputSpan::new() };
             for variant in variants {
                 t.elements.push(ParserTypeElement{ element_span: InputSpan::new(), variant: variant.clone() });
+            }
+            t
+        }
         use ParserTypeVariant as PTV;
         insert_builtin_function(&mut parser, "get", &["T"], |id| (
             vec![
                 ("input", quick_type(&[PTV::Input, PTV::PolymorphicArgument(id.upcast(), 0)]))
             ],
             quick_type(&[PTV::PolymorphicArgument(id.upcast(), 0)])
         ));
         insert_builtin_function(&mut parser, "put", &["T"], |id| (
             vec![
                 ("output", quick_type(&[PTV::Output, PTV::PolymorphicArgument(id.upcast(), 0)])),
                 ("value", quick_type(&[PTV::PolymorphicArgument(id.upcast(), 0)])),
             ],
             quick_type(&[PTV::Void])
         ));
         insert_builtin_function(&mut parser, "fires", &["T"], |id| (
             vec![
@@ @@ -267,52 +318,68 @@ impl Parser { @@
             self.modules[module_idx].phase = ModuleCompilationPhase::Typed;
             let mut ctx = visitor::Ctx{
                 heap: &mut self.heap,
                 modules: &mut self.modules,
                 module_idx,
                 symbols: &mut self.symbol_table,
                 types: &mut self.type_table,
                 arch: &self.arch,
             };
             self.pass_rewriting.visit_module(&mut ctx);
             self.pass_stack_size.visit_module(&mut ctx);
+        }
         // Write out desired information
         if let Some(filename) = &self.write_ast_to {
             let mut writer = ASTWriter::new();
             let mut file = std::fs::File::create(std::path::Path::new(filename)).unwrap();
             writer.write_ast(&mut file, &self.heap);
+        }
         Ok(())
+    }
+}
 fn insert_builtin_type(type_table: &mut TypeTable, parts: Vec<ConcreteTypePart>, has_poly_var: bool, size: usize, alignment: usize) -> TypeId {
     const POLY_VARS: [PolymorphicVariable; 1] = [PolymorphicVariable{
         identifier: Identifier::new_empty(InputSpan::new()),
         is_in_use: false,
     }];
     let concrete_type = ConcreteType{ parts };
     let poly_var = if has_poly_var {
         POLY_VARS.as_slice()
     } else {
         &[]
     };
     return type_table.add_builtin_type(concrete_type, poly_var, size, alignment);
+}
 // Note: args and return type need to be a function because we need to know the function ID.
 fn insert_builtin_function<T: Fn(FunctionDefinitionId) -> (Vec<(&'static str, ParserType)>, ParserType)> (
     p: &mut Parser, func_name: &str, polymorphic: &[&str], arg_and_return_fn: T) {
     p: &mut Parser, func_name: &str, polymorphic: &[&str], arg_and_return_fn: T
 ) {
     let mut poly_vars = Vec::with_capacity(polymorphic.len());
     for poly_var in polymorphic {
         poly_vars.push(Identifier{ span: InputSpan::new(), value: p.string_pool.intern(poly_var.as_bytes()) });
+    }
     let func_ident_ref = p.string_pool.intern(func_name.as_bytes());
     let func_id = p.heap.alloc_function_definition(|this| FunctionDefinition{
         this,
         defined_in: RootId::new_invalid(),
         builtin: true,
         span: InputSpan::new(),
         identifier: Identifier{ span: InputSpan::new(), value: func_ident_ref.clone() },
         poly_vars,
         return_type: ParserType{ elements: Vec::new(), full_span: InputSpan::new() },
         parameters: Vec::new(),
         scope: ScopeId::new_invalid(),
         body: BlockStatementId::new_invalid(),
         num_expressions_in_body: -1,
     });
     let (arguments, return_type) = arg_and_return_fn(func_id);
     let mut parameters = Vec::with_capacity(arguments.len());
     for (arg_name, arg_type) in arguments {

src/protocol/parser/pass_typing.rs

➞

Show inline comments

@@ @@ -3762,48 +3762,49 @@ impl PassTyping { @@
             match concrete_part {
                 CTP::Void => parser_type.push(ITP::Void),
                 CTP::Message => {
                     parser_type.push(ITP::Message);
                     parser_type.push(ITP::UInt8)
                 },
                 CTP::Bool => parser_type.push(ITP::Bool),
                 CTP::UInt8 => parser_type.push(ITP::UInt8),
                 CTP::UInt16 => parser_type.push(ITP::UInt16),
                 CTP::UInt32 => parser_type.push(ITP::UInt32),
                 CTP::UInt64 => parser_type.push(ITP::UInt64),
                 CTP::SInt8 => parser_type.push(ITP::SInt8),
                 CTP::SInt16 => parser_type.push(ITP::SInt16),
                 CTP::SInt32 => parser_type.push(ITP::SInt32),
                 CTP::SInt64 => parser_type.push(ITP::SInt64),
                 CTP::Character => parser_type.push(ITP::Character),
                 CTP::String => {
                     parser_type.push(ITP::String);
                     parser_type.push(ITP::Character)
                 },
                 CTP::Array => parser_type.push(ITP::Array),
                 CTP::Slice => parser_type.push(ITP::Slice),
                 CTP::Input => parser_type.push(ITP::Input),
                 CTP::Output => parser_type.push(ITP::Output),
                 CTP::Pointer => unreachable!("pointer type during concrete to inference type conversion"),
                 CTP::Tuple(num) => parser_type.push(ITP::Tuple(*num)),
                 CTP::Instance(id, num) => parser_type.push(ITP::Instance(*id, *num)),
                 CTP::Function(_, _) => unreachable!("function type during concrete to inference type conversion"),
                 CTP::Component(_, _) => unreachable!("component type during concrete to inference type conversion"),
+            }
+        }
+    }
     /// Construct an error when an expression's type does not match. This
     /// happens if we infer the expression type from its arguments (e.g. the
     /// expression type of an addition operator is the type of the arguments)
     /// But the expression type was already set due to our parent (e.g. an
     /// "if statement" or a "logical not" always expecting a boolean)
     fn construct_expr_type_error(
         &self, ctx: &Ctx, expr_id: ExpressionId, arg_id: ExpressionId
     ) -> ParseError {
         // TODO: Expand and provide more meaningful information for humans
         let expr = &ctx.heap[expr_id];
         let arg_expr = &ctx.heap[arg_id];
         let expr_idx = expr.get_unique_id_in_definition();
         let arg_expr_idx = arg_expr.get_unique_id_in_definition();
         let expr_type = &self.expr_types[expr_idx as usize].expr_type;
         let arg_type = &self.expr_types[arg_expr_idx as usize].expr_type;

src/protocol/parser/type_table.rs

➞

Show inline comments

@@ @@ -119,50 +119,50 @@ impl DefinedTypeVariant { @@
     pub(crate) fn as_struct(&self) -> &StructType {
         match self {
             DefinedTypeVariant::Struct(v) => v,
             _ => unreachable!("Cannot convert {} to struct variant", self.type_class())
+        }
+    }
     pub(crate) fn as_enum(&self) -> &EnumType {
         match self {
             DefinedTypeVariant::Enum(v) => v,
             _ => unreachable!("Cannot convert {} to enum variant", self.type_class())
+        }
+    }
     pub(crate) fn as_union(&self) -> &UnionType {
         match self {
             DefinedTypeVariant::Union(v) => v,
             _ => unreachable!("Cannot convert {} to union variant", self.type_class())
+        }
+    }
+}
 pub struct PolymorphicVariable {
     identifier: Identifier,
     is_in_use: bool, // a polymorphic argument may be defined, but not used by the type definition
     pub(crate) identifier: Identifier,
     pub(crate) is_in_use: bool, // a polymorphic argument may be defined, but not used by the type definition
+}
 /// `EnumType` is the classical C/C++ enum type. It has various variants with
 /// an assigned integer value. The integer values may be user-defined,
 /// compiler-defined, or a mix of the two. If a user assigns the same enum
 /// value multiple times, we assume the user is an expert and we consider both
 /// variants to be equal to one another.
 pub struct EnumType {
     pub variants: Vec<EnumVariant>,
     pub minimum_tag_value: i64,
     pub maximum_tag_value: i64,
     pub tag_type: ConcreteType,
     pub size: usize,
     pub alignment: usize,
+}
 // TODO: Also support maximum u64 value
 pub struct EnumVariant {
     pub identifier: Identifier,
     pub value: i64,
+}
 /// `UnionType` is the algebraic datatype (or sum type, or discriminated union).
 /// A value is an element of the union, identified by its tag, and may contain
@@ @@ -762,48 +762,49 @@ impl TypeTable { @@
     /// going to be performing typechecking on it, and we don't want to
     /// check the same monomorph twice)
     pub(crate) fn reserve_procedure_monomorph_type_id(&mut self, definition_id: &DefinitionId, concrete_type: ConcreteType) -> TypeId {
         debug_assert_eq!(get_concrete_type_definition(&concrete_type.parts).unwrap(), *definition_id);
         let type_id = TypeId(self.mono_types.len() as i64);
         let base_type = self.definition_lookup.get_mut(definition_id).unwrap();
         self.mono_search_key.set(&concrete_type.parts, &base_type.poly_vars);
         debug_assert!(!self.mono_type_lookup.contains_key(&self.mono_search_key));
         self.mono_type_lookup.insert(self.mono_search_key.clone(), type_id);
         self.mono_types.push(MonoType::new_empty(type_id, concrete_type, MonoTypeVariant::Procedure(ProcedureMonomorph{
             arg_types: Vec::new(),
             expr_data: Vec::new(),
         })));
         return type_id;
+    }
     /// Adds a builtin type to the type table. As this is only called by the
     /// compiler during setup we assume it cannot fail.
     // TODO: Finish this train of thought, requires a little bit of design work
     pub(crate) fn add_builtin_type(&mut self, concrete_type: ConcreteType, poly_vars: &[PolymorphicVariable], size: usize, alignment: usize) -> TypeId {
         self.mono_search_key.set(&concrete_type.parts, poly_vars);
         debug_assert!(!self.mono_type_lookup.contains_key(&self.mono_search_key));
         debug_assert_ne!(alignment, 0);
         let type_id = TypeId(self.mono_types.len() as i64);
         self.mono_type_lookup.insert(self.mono_search_key.clone(), type_id);
         self.mono_types.push(MonoType{
             type_id,
             concrete_type,
             size,
             alignment,
             variant: MonoTypeVariant::Builtin,
         });
         return type_id;
+    }
     /// Adds a monomorphed type to the type table. If it already exists then the
     /// previous entry will be used.
     pub(crate) fn add_monomorphed_type(
         &mut self, modules: &[Module], heap: &Heap, arch: &TargetArch,
         definition_id: DefinitionId, concrete_type: ConcreteType
     ) -> Result<TypeId, ParseError> {
         debug_assert_eq!(definition_id, get_concrete_type_definition(&concrete_type.parts).unwrap());
         // Check if the concrete type was already added
         let definition = self.definition_lookup.get(&definition_id).unwrap();
         let poly_var_in_use = &definition.poly_vars;
@@ @@ -1773,48 +1774,50 @@ impl TypeTable { @@
     //--------------------------------------------------------------------------
     // Determining memory layout for types
     //--------------------------------------------------------------------------
     /// Should be called after type loops are detected (and resolved
     /// successfully). As a result of this call we expect the
     /// `encountered_types` array to be filled. We'll calculate size/alignment/
     /// offset values for those types in this routine.
     fn lay_out_memory_for_encountered_types(&mut self, arch: &TargetArch) {
         // Programmers note: this works like a little stack machine. We have
         // memory layout breadcrumbs which, like the type loop breadcrumbs, keep
         // track of the currently considered member type. This breadcrumb also
         // stores an index into the `size_alignment_stack`, which will be used
         // to store intermediate size/alignment pairs until all members are
         // resolved. Note that this `size_alignment_stack` is NOT an
         // optimization, we're working around borrowing rules here.
         // Just finished type loop detection, so we're left with the encountered
         // types only
         debug_assert!(self.type_loops.is_empty());
         debug_assert!(!self.encountered_types.is_empty());
         debug_assert!(self.memory_layout_breadcrumbs.is_empty());
         debug_assert!(self.size_alignment_stack.is_empty());
         let (ptr_size, ptr_align) = self.mono_types[arch.pointer_type_id.0 as usize].get_size_alignment().unwrap();
         // Push the first entry (the type we originally started with when we
         // were detecting type loops)
         let first_entry = &self.encountered_types[0];
         self.memory_layout_breadcrumbs.push(MemoryBreadcrumb{
             type_id: first_entry.type_id,
             next_member: 0,
             next_embedded: 0,
             first_size_alignment_idx: 0,
         });
         // Enter the main resolving loop
         'breadcrumb_loop: while !self.memory_layout_breadcrumbs.is_empty() {
             let cur_breadcrumb_idx = self.memory_layout_breadcrumbs.len() - 1;
             let mut breadcrumb = self.memory_layout_breadcrumbs[cur_breadcrumb_idx].clone();
             let mono_type = &self.mono_types[breadcrumb.type_id.0 as usize];
             match &mono_type.variant {
                 MonoTypeVariant::Builtin | MonoTypeVariant::Enum => {
                     // Size should already be computed
                     dbg_code!({
                         let mono_type = &self.mono_types[breadcrumb.type_id.0 as usize];
                         debug_assert!(mono_type.size != 0 && mono_type.alignment != 0);
                     });
                 },
@@ @@ -1855,49 +1858,48 @@ impl TypeTable { @@
                         breadcrumb.next_member += 1;
                         breadcrumb.next_embedded = 0;
+                    }
                     // If here then we can at least compute the stack size of
                     // the type, we'll have to come back at the very end to
                     // fill in the heap size/alignment/offset of each heap-
                     // allocated variant.
                     let mut max_size = mono_type.tag_size;
                     let mut max_alignment = mono_type.tag_size;
                     let mono_type = &mut self.mono_types[breadcrumb.type_id.0 as usize];
                     let union_type = mono_type.variant.as_union_mut();
                     let mut size_alignment_idx = breadcrumb.first_size_alignment_idx as usize;
                     for variant in &mut union_type.variants {
                         // We're doing stack computations, so always start with
                         // the tag size/alignment.
                         let mut variant_offset = union_type.tag_size;
                         let mut variant_alignment = union_type.tag_size;
                         if variant.lives_on_heap {
                             // Variant lives on heap, so just a pointer
                             let (ptr_size, ptr_align) = arch.pointer_size_alignment;
                             align_offset_to(&mut variant_offset, ptr_align);
                             variant_offset += ptr_size;
                             variant_alignment = variant_alignment.max(ptr_align);
                         } else {
                             // Variant lives on stack, so walk all embedded
                             // types.
                             for embedded in &mut variant.embedded {
                                 let (size, alignment) = self.size_alignment_stack[size_alignment_idx];
                                 embedded.size = size;
                                 embedded.alignment = alignment;
                                 size_alignment_idx += 1;
                                 align_offset_to(&mut variant_offset, alignment);
                                 embedded.offset = variant_offset;
                                 variant_offset += size;
                                 variant_alignment = variant_alignment.max(alignment);
+                            }
                         };
                         max_size = max_size.max(variant_offset);
                         max_alignment = max_alignment.max(variant_alignment);
+                    }
@@ @@ -2086,106 +2088,98 @@ impl TypeTable { @@
                 // At least one entry lives on the heap
                 mono_type.heap_size = max_size;
                 mono_type.heap_alignment = max_alignment;
+            }
+        }
         // And now, we're actually, properly, done
         self.encountered_types.clear();
+    }
     /// Attempts to compute size/alignment for the provided type. Note that this
     /// is called *after* type loops have been succesfully resolved. Hence we
     /// may assume that all monomorph entries exist, but we may not assume that
     /// those entries already have their size/alignment computed.
     // Passed parameters are messy. But need to strike balance between borrowing
     // and allocations in hot loops. So it is what it is.
     fn get_memory_layout_or_breadcrumb(
         definition_map: &DefinitionMap, mono_type_map: &MonoTypeMap, mono_types: &MonoTypeArray,
         search_key: &mut MonoSearchKey, arch: &TargetArch, parts: &[ConcreteTypePart],
         size_alignment_stack_len: usize,
     ) -> MemoryLayoutResult {
         use ConcreteTypePart as CTP;
         debug_assert!(!parts.is_empty());
         let (builtin_size, builtin_alignment) = match parts[0] {
             CTP::Void   => (0, 1),
             CTP::Message => arch.array_size_alignment,
             CTP::Bool   => (1, 1),
             CTP::UInt8  => (1, 1),
             CTP::UInt16 => (2, 2),
             CTP::UInt32 => (4, 4),
             CTP::UInt64 => (8, 8),
             CTP::SInt8  => (1, 1),
             CTP::SInt16 => (2, 2),
             CTP::SInt32 => (4, 4),
             CTP::SInt64 => (8, 8),
             CTP::Character => (4, 4),
             CTP::String => arch.string_size_alignment,
             CTP::Array => arch.array_size_alignment,
             CTP::Slice => arch.array_size_alignment,
             CTP::Input => arch.port_size_alignment,
             CTP::Output => arch.port_size_alignment,
         let type_id = match parts[0] {
             CTP::Void      => arch.void_type_id,
             CTP::Message   => arch.message_type_id,
             CTP::Bool      => arch.bool_type_id,
             CTP::UInt8     => arch.uint8_type_id,
             CTP::UInt16    => arch.uint16_type_id,
             CTP::UInt32    => arch.uint32_type_id,
             CTP::UInt64    => arch.uint64_type_id,
             CTP::SInt8     => arch.sint8_type_id,
             CTP::SInt16    => arch.sint16_type_id,
             CTP::SInt32    => arch.sint32_type_id,
             CTP::SInt64    => arch.sint64_type_id,
             CTP::Character => arch.char_type_id,
             CTP::String    => arch.string_type_id,
             CTP::Array     => arch.array_type_id,
             CTP::Slice     => arch.slice_type_id,
             CTP::Input     => arch.input_type_id,
             CTP::Output    => arch.output_type_id,
             CTP::Pointer   => arch.pointer_type_id,
             CTP::Tuple(_) => {
                 Self::set_search_key_to_tuple(search_key, definition_map, parts);
                 let type_id = mono_type_map.get(&search_key).copied().unwrap();
                 let mono_type = &mono_types[type_id.0 as usize];
                 if let Some((size, alignment)) = mono_type.get_size_alignment() {
                     return MemoryLayoutResult::TypeExists(size, alignment);
                 } else {
                     return MemoryLayoutResult::PushBreadcrumb(MemoryBreadcrumb{
                         type_id,
                         next_member: 0,
                         next_embedded: 0,
                         first_size_alignment_idx: size_alignment_stack_len as u32,
                     })
+                }
                 type_id
             },
             CTP::Instance(definition_id, _) => {
                 // Retrieve entry and the specific monomorph index by applying
                 // the full concrete type.
                 let definition_type = definition_map.get(&definition_id).unwrap();
                 search_key.set(parts, &definition_type.poly_vars);
                 let type_id = mono_type_map.get(&search_key).copied().unwrap();
                 let mono_type = &mono_types[type_id.0 as usize];
                 if let Some((size, alignment)) = mono_type.get_size_alignment() {
                     return MemoryLayoutResult::TypeExists(size, alignment);
                 } else {
                     return MemoryLayoutResult::PushBreadcrumb(MemoryBreadcrumb{
                         type_id,
                         next_member: 0,
                         next_embedded: 0,
                         first_size_alignment_idx: size_alignment_stack_len as u32,
                     });
+                }
                 type_id
             },
             CTP::Function(_, _) | CTP::Component(_, _) => {
                 todo!("storage for 'function pointers'");
+            }
         };
         return MemoryLayoutResult::TypeExists(builtin_size, builtin_alignment);
         let mono_type = &mono_types[type_id.0 as usize];
         if let Some((size, alignment)) = mono_type.get_size_alignment() {
             return MemoryLayoutResult::TypeExists(size, alignment);
         } else {
             return MemoryLayoutResult::PushBreadcrumb(MemoryBreadcrumb{
                 type_id,
                 next_member: 0,
                 next_embedded: 0,
                 first_size_alignment_idx: size_alignment_stack_len as u32,
             });
+        }
+    }
     /// Returns tag concrete type (always a builtin integer type), the size of
     /// that type in bytes (and implicitly, its alignment)
     fn variant_tag_type_from_values(min_val: i64, max_val: i64) -> (ConcreteType, usize) {
         debug_assert!(min_val <= max_val);
         let (part, size) = if min_val >= 0 {
             // Can be an unsigned integer
             if max_val <= (u8::MAX as i64) {
                 (ConcreteTypePart::UInt8, 1)
             } else if max_val <= (u16::MAX as i64) {
                 (ConcreteTypePart::UInt16, 2)
             } else if max_val <= (u32::MAX as i64) {
                 (ConcreteTypePart::UInt32, 4)
             } else {
                 (ConcreteTypePart::UInt64, 8)
+            }
         } else {
             // Must be a signed integer
             if min_val >= (i8::MIN as i64) && max_val <= (i8::MAX as i64) {
                 (ConcreteTypePart::SInt8, 1)
             } else if min_val >= (i16::MIN as i64) && max_val <= (i16::MAX as i64) {
                 (ConcreteTypePart::SInt16, 2)

0 comments (0 inline, 0 general)