Implement canonical interface names in wit-component

crates/wasmparser/src/validator/component.rs

@@ -4682,7 +4682,13 @@ impl ComponentNameContext {
             let implements = ComponentName::new_with_features(implements, offset, *features)
                 .with_context(|| format!("`{implements}` is not a valid name"))?;
             match implements.kind() {
-                ComponentNameKind::Interface(_) => {}
+                ComponentNameKind::Interface(iface_name) => {
+                    if let Some(suffix) = version_suffix {
+                        if let Err(e) = iface_name.version(Some(suffix)) {
+                            bail!(offset, "invalid interface version: {e}");
+                        }
+                    }
+                }
                 _ => bail!(offset, "name `{implements}` must be an interface"),
             }
         }

crates/wit-component/src/encoding.rs

@@ -99,7 +99,7 @@ const TLS_BASE_SET: &str = "$set-tls-base";
 pub(crate) mod fixup;
 
 mod wit;
-pub use wit::{encode, encode_world};
+pub use wit::{encode, encode_with_options, encode_world};
 
 mod types;
 use types::{InstanceTypeEncoder, RootTypeEncoder, TypeEncodingMaps, ValtypeEncoder};
@@ -609,12 +609,29 @@ impl<'a> EncodingState<'a> {
         let instance_type_idx = self
             .component
             .type_instance(Some(&format!("ty-{name}")), &ty);
+
+        let mut implements = info.implements.clone();
+        let (import_name, version_suffix) = if self.info.encoder.emit_canonical_names {
+            let suffix = resolve.version_suffix_of(interface_id);
+            if implements.is_some() {
+                implements = resolve.canon_id_of(interface_id);
+                (name.to_string(), suffix)
+            } else {
+                let canon_name = resolve
+                    .canon_id_of(interface_id)
+                    .unwrap_or_else(|| name.to_string());
+                (canon_name, suffix)
+            }
+        } else {
+            (name.to_string(), None)
+        };
+
         let instance_idx = self.component.import(
             wasm_encoder::ComponentExternName {
-                name: name.into(),
-                implements: info.implements.as_deref().map(|s| s.into()),
+                name: import_name.into(),
+                implements: implements.map(|s| s.into()),
                 external_id: info.external_id.as_deref().map(|s| s.into()),
-                version_suffix: None,
+                version_suffix: version_suffix.map(|s| s.into()),
             },
             ComponentTypeRef::Instance(instance_type_idx),
         );
@@ -762,7 +779,11 @@ impl<'a> EncodingState<'a> {
         let world = &resolve.worlds[self.info.encoder.metadata.world];
 
         for export_name in exports {
-            let export_string = resolve.name_world_key(export_name);
+            let export_string = if self.info.encoder.emit_canonical_names {
+                resolve.name_canon_world_key(export_name)
+            } else {
+                resolve.name_world_key(export_name)
+            };
             match &world.exports[export_name] {
                 WorldItem::Function(func) => {
                     let ty = self
@@ -993,12 +1014,25 @@ impl<'a> EncodingState<'a> {
             component_index,
             imports,
         );
+        let mut implements = resolve.implements_value(key, item);
+        let export_version_suffix = if self.info.encoder.emit_canonical_names {
+            match (&implements, key, item) {
+                (Some(_), _, WorldItem::Interface { id, .. }) => {
+                    implements = resolve.canon_id_of(*id);
+                    resolve.version_suffix_of(*id)
+                }
+                (None, WorldKey::Interface(id), _) => resolve.version_suffix_of(*id),
+                _ => None,
+            }
+        } else {
+            None
+        };
         let idx = self.component.export(
             wasm_encoder::ComponentExternName {
                 name: export_name.into(),
-                implements: resolve.implements_value(key, item).map(|s| s.into()),
+                implements: implements.map(|s| s.into()),
                 external_id: resolve.external_id_value(key, item).map(|s| s.into()),
-                version_suffix: None,
+                version_suffix: export_version_suffix.map(|s| s.into()),
             },
             ComponentExportKind::Instance,
             instance_index,
@@ -3290,6 +3324,7 @@ pub struct ComponentEncoder {
     pub(super) reject_legacy_names: bool,
     debug_names: bool,
     shim_return_call_ref: bool,
+    emit_canonical_names: bool,
 }
 
 impl ComponentEncoder {
@@ -3357,6 +3392,20 @@ impl ComponentEncoder {
         self
     }
 
+    /// Sets whether to emit canonical interface names in the component binary.
+    ///
+    /// When enabled, import/export names use canonical version prefixes (e.g.,
+    /// `wasi:cli/exit@0.2` instead of `wasi:cli/exit@0.2.1`) and the
+    /// `version_suffix` field is populated. This also forces merging of
+    /// imports that share the same canonical version prefix.
+    /// This flag subsumes the `merge_imports_based_on_semver` flag.
+    ///
+    /// This is disabled by default.
+    pub fn emit_canonical_names(&mut self, emit: bool) -> &mut Self {
+        self.emit_canonical_names = emit;
+        self
+    }
+
     /// Sets whether to reject the historical mangling/name scheme for core wasm
     /// imports/exports as they map to the component model.
     ///
@@ -3509,7 +3558,11 @@ impl ComponentEncoder {
             bail!("a module is required when encoding a component");
         }
 
-        if self.merge_imports_based_on_semver.unwrap_or(true) {
+        if self.emit_canonical_names {
+            self.metadata
+                .resolve
+                .merge_world_imports_based_on_canonical_version(self.metadata.world)?;
+        } else if self.merge_imports_based_on_semver.unwrap_or(true) {
             self.metadata
                 .resolve
                 .merge_world_imports_based_on_semver(self.metadata.world)?;

crates/wit-component/src/encoding/wit.rs

@@ -25,7 +25,16 @@ use wit_parser::*;
 /// The binary returned can be [`decode`d](crate::decode) to recover the WIT
 /// package provided.
 pub fn encode(resolve: &Resolve, package: PackageId) -> Result<Vec<u8>> {
-    let mut component = encode_component(resolve, package)?;
+    encode_with_options(resolve, package, false)
+}
+
+/// Same as [`encode`] but with an option to emit canonical interface names.
+pub fn encode_with_options(
+    resolve: &Resolve,
+    package: PackageId,
+    canonical_names: bool,
+) -> Result<Vec<u8>> {
+    let mut component = encode_component_with_options(resolve, package, canonical_names)?;
     component.raw_custom_section(&crate::base_producers().raw_custom_section());
     Ok(component.finish())
 }
@@ -48,11 +57,16 @@ pub fn encode(resolve: &Resolve, package: PackageId) -> Result<Vec<u8>> {
 ///
 /// The binary returned can be [`decode`d](crate::decode) to recover the WIT
 /// package provided.
-pub fn encode_component(resolve: &Resolve, package: PackageId) -> Result<ComponentBuilder> {
+pub fn encode_component_with_options(
+    resolve: &Resolve,
+    package: PackageId,
+    canonical_names: bool,
+) -> Result<ComponentBuilder> {
     let mut encoder = Encoder {
         component: ComponentBuilder::default(),
         resolve,
         package,
+        canonical_names,
     };
     encoder.run()?;
 
@@ -67,6 +81,15 @@ pub fn encode_component(resolve: &Resolve, package: PackageId) -> Result<Compone
 
 /// Encodes a `world` as a component type.
 pub fn encode_world(resolve: &Resolve, world_id: WorldId) -> Result<ComponentType> {
+    encode_world_with_options(resolve, world_id, false)
+}
+
+/// Same as [`encode_world`] but with an option to emit canonical names.
+pub fn encode_world_with_options(
+    resolve: &Resolve,
+    world_id: WorldId,
+    canonical_names: bool,
+) -> Result<ComponentType> {
     let mut component = InterfaceEncoder::new(resolve);
     let world = &resolve.worlds[world_id];
     log::trace!("encoding world {}", world.name);
@@ -93,9 +116,10 @@ pub fn encode_world(resolve: &Resolve, world_id: WorldId) -> Result<ComponentTyp
                 continue;
             }
         };
-        component
-            .outer
-            .import(component_extern_name(resolve, key, import), ty);
+        component.outer.import(
+            component_extern_name(resolve, key, import, canonical_names),
+            ty,
+        );
     }
     // Encode the exports
     for (key, export) in world.exports.iter() {
@@ -113,9 +137,10 @@ pub fn encode_world(resolve: &Resolve, world_id: WorldId) -> Result<ComponentTyp
             }
             WorldItem::Type { .. } => unreachable!(),
         };
-        component
-            .outer
-            .export(component_extern_name(resolve, key, export), ty);
+        component.outer.export(
+            component_extern_name(resolve, key, export, canonical_names),
+            ty,
+        );
     }
 
     Ok(component.outer)
@@ -125,19 +150,40 @@ fn component_extern_name(
     resolve: &Resolve,
     key: &WorldKey,
     item: &WorldItem,
+    canonical_names: bool,
 ) -> wasm_encoder::ComponentExternName<'static> {
-    ComponentExternName {
-        name: resolve.name_world_key(key).into(),
-        implements: resolve.implements_value(key, item).map(|s| s.into()),
-        external_id: resolve.external_id_value(key, item).map(|s| s.into()),
-        version_suffix: None,
+    if canonical_names {
+        let name = resolve.name_canon_world_key(key);
+        let mut implements = resolve.implements_value(key, item);
+        let version_suffix = match (&implements, key, item) {
+            (Some(_), _, WorldItem::Interface { id, .. }) => {
+                implements = resolve.canon_id_of(*id);
+                resolve.version_suffix_of(*id)
+            }
+            (None, WorldKey::Interface(id), _) => resolve.version_suffix_of(*id),
+            _ => None,
+        };
+        ComponentExternName {
+            name: name.into(),
+            implements: implements.map(|s| s.into()),
+            external_id: resolve.external_id_value(key, item).map(|s| s.into()),
+            version_suffix: version_suffix.map(|s| s.into()),
+        }
+    } else {
+        ComponentExternName {
+            name: resolve.name_world_key(key).into(),
+            implements: resolve.implements_value(key, item).map(|s| s.into()),
+            external_id: resolve.external_id_value(key, item).map(|s| s.into()),
+            version_suffix: None,
+        }
     }
 }
 
 struct Encoder<'a> {
     component: ComponentBuilder,
     resolve: &'a Resolve,
     package: PackageId,
+    canonical_names: bool,
 }
 
 impl Encoder<'_> {
@@ -153,7 +199,8 @@ impl Encoder<'_> {
         // For each `world` encode it directly as a component and then create a
         // wrapper component that exports that component.
         for (name, &world) in self.resolve.packages[self.package].worlds.iter() {
-            let component_ty = encode_world(self.resolve, world)?;
+            let component_ty =
+                encode_world_with_options(self.resolve, world, self.canonical_names)?;
 
             let world = &self.resolve.worlds[world];
             let mut wrapper = ComponentType::new();
@@ -197,11 +244,15 @@ impl Encoder<'_> {
         for interface in interfaces {
             encoder.interface = Some(interface);
             let iface = &self.resolve.interfaces[interface];
-            let name = self.resolve.id_of(interface).unwrap();
+            let name = if self.canonical_names {
+                self.resolve.canon_id_of(interface).unwrap()
+            } else {
+                self.resolve.id_of(interface).unwrap()
+            };
             if interface == id {
                 let idx = encoder.encode_instance(interface)?;
                 log::trace!("exporting self as {idx}");
-                encoder.outer.export(name, ComponentTypeRef::Instance(idx));
+                encoder.outer.export(&name, ComponentTypeRef::Instance(idx));
             } else {
                 encoder.push_instance();
                 for (_, id) in iface.types.iter() {
@@ -212,7 +263,7 @@ impl Encoder<'_> {
                 encoder.outer.ty().instance(&instance);
                 encoder.import_map.insert(interface, encoder.instances);
                 encoder.instances += 1;
-                encoder.outer.import(name, ComponentTypeRef::Instance(idx));
+                encoder.outer.import(&name, ComponentTypeRef::Instance(idx));
             }
         }
 

crates/wit-component/src/lib.rs

@@ -17,7 +17,7 @@ mod printing;
 mod targets;
 mod validation;
 
-pub use encoding::{ComponentEncoder, LibraryInfo, encode};
+pub use encoding::{ComponentEncoder, LibraryInfo, encode, encode_with_options};
 pub use linking::Linker;
 pub use printing::*;
 pub use targets::*;

crates/wit-component/tests/components.rs

@@ -116,6 +116,7 @@ fn run_test(path: &Path) -> Result<()> {
                 .debug_names(true)
                 .shim_return_call_ref(config.return_call_ref)
                 .realloc_via_memory_grow(config.realloc_via_memory_grow)
+                .emit_canonical_names(config.merge_imports_based_on_canonical_version)
                 .module(&module)?;
             for adapter in adapters {
                 let (name, wasm) = read_name_and_module("adapt-", &adapter?, &resolve, pkg_id)?;
@@ -247,6 +248,7 @@ struct Config {
     use_built_in_libdl: bool,
     return_call_ref: bool,
     realloc_via_memory_grow: bool,
+    merge_imports_based_on_canonical_version: bool,
 }
 
 /// Reads the configuration for the test located at `path`.