Type annotations

Signed-off-by: Ben Sherman <[email protected]>

Author: bentsherman

docs/migrations/25-10.md

@@ -8,6 +8,51 @@ This page summarizes the upcoming changes in Nextflow 25.10, which will be relea
 This page is a work in progress and will be updated as features are finalized. It should not be considered complete until the 25.10 release.
 :::
 
+## New features
+
+<h3>Type annotations</h3>
+
+Type annotations are a way to denote the *type* of a variable. They are useful both for documenting and validating your pipeline code.
+
+```nextflow
+workflow RNASEQ {
+  take:
+  reads: Channel<Path>
+  index: Channel<Path>
+
+  main:
+  samples_ch = QUANT( reads.combine(index) )
+
+  emit:
+  samples: Channel<Path> = samples_ch
+}
+
+def isSraId(id: String) -> Boolean {
+  return id.startsWith('SRA')
+}
+```
+
+The following declarations can be annotated with types:
+
+- Pipeline parameters (the `params` block)
+- Workflow takes and emits
+- Function parameters and returns
+- Local variables
+- Closure parameters
+- Workflow outputs (the `output` block)
+
+Type annotations can refer to any of the {ref}`standard types <stdlib-types>`.
+
+Type annotations can be appended with `?` to denote that the value can be `null`:
+
+```nextflow
+def x_opt: String? = null
+```
+
+:::{note}
+While Nextflow inherited type annotations of the form `<type> <name>` from Groovy, this syntax was deprecated in the {ref}`strict syntax <strict-syntax-page>`. Groovy-style type annotations are still allowed for functions and local variables, but will be automatically converted to Nextflow-stype type annotations when formatting code with the language server or `nextflow lint`.
+:::
+
 ## Enhancements
 
 <h3>New syntax for workflow handlers</h3>

docs/strict-syntax.md

@@ -299,10 +299,22 @@ def str = 'hello'
 def meta = [:]
 ```
 
-:::{note}
-Because type annotations are useful for providing type checking at runtime, the language server will not report errors for Groovy-style type annotations at this time. Type annotations will be addressed in a future version of the Nextflow language specification.
+:::{versionadded} 25.10.0
 :::
 
+Local variables can be declared with a type annotation:
+
+```nextflow
+def a: Integer = 1
+def b: Integer = 2
+def (c: Integer, d: Integer) = [3, 4]
+def (e: Integer, f: Integer) = [5, 6]
+def str: String = 'hello'
+def meta: Map = [:]
+```
+
+While Groovy-style type annotations are still supported, the linter and language server will automatically convert them to Nextflow-style type annotations when formatting code. Groovy-style type annotations will not be supported in a future version.
+
 ### Strings
 
 Groovy supports a wide variety of strings, including multi-line strings, dynamic strings, slashy strings, multi-line dynamic slashy strings, and more.

modules/nf-lang/src/main/antlr/ScriptParser.g4

@@ -246,15 +246,25 @@ workflowBody
     ;
 
 workflowTakes
-    :   identifier (sep identifier)*
+    :   workflowTake (sep workflowTake)*
+    ;
+
+workflowTake
+    :   identifier (COLON type)?
+    |   statement
     ;
 
 workflowEmits
-    :   statement (sep statement)*
+    :   workflowEmit (sep workflowEmit)*
+    ;
+
+workflowEmit
+    :   nameTypePair (ASSIGN expression)?
+    |   statement
     ;
 
 workflowPublishers
-    :   statement (sep statement)*
+    :   workflowEmit (sep workflowEmit)*
     ;
 
 // -- output definition
@@ -269,14 +279,19 @@ outputBody
     ;
 
 outputDeclaration
-    :   identifier LBRACE nls blockStatements? RBRACE
+    :   identifier (COLON type)? LBRACE nls blockStatements? RBRACE
     |   statement
     ;
 
 // -- function definition
 functionDef
-    :   (DEF | legacyType | DEF legacyType) identifier LPAREN nls (formalParameterList nls)? rparen nls LBRACE
-        nls blockStatements? RBRACE
+    :   DEF
+        identifier LPAREN nls (formalParameterList nls)? rparen (ARROW type)?
+        nls LBRACE nls blockStatements? RBRACE
+
+    |   (legacyType | DEF legacyType)
+        identifier LPAREN nls (formalParameterList nls)? rparen
+        nls LBRACE nls blockStatements? RBRACE
     ;
 
 // -- incomplete script declaration
@@ -321,7 +336,12 @@ tryCatchStatement
     ;
 
 catchClause
-    :   CATCH LPAREN catchTypes? identifier rparen nls statementOrBlock
+    :   CATCH LPAREN catchVariable rparen nls statementOrBlock
+    ;
+
+catchVariable
+    :   identifier (COLON catchTypes)?
+    |   catchTypes identifier
     ;
 
 catchTypes
@@ -335,19 +355,28 @@ assertStatement
 
 // -- variable declaration
 variableDeclaration
-    :   (DEF | legacyType | DEF legacyType) identifier (nls ASSIGN nls initializer=expression)?
-    |   DEF variableNames nls ASSIGN nls initializer=expression
+    :   DEF nameTypePair (nls ASSIGN nls initializer=expression)?
+    |   DEF nameTypePairs nls ASSIGN nls initializer=expression
+    |   (legacyType | DEF legacyType) identifier (nls ASSIGN nls initializer=expression)?
     ;
 
-variableNames
-    :   LPAREN identifier (COMMA identifier)+ rparen
+nameTypePairs
+    :   LPAREN nameTypePair (COMMA nameTypePair)+ rparen
+    ;
+
+nameTypePair
+    :   identifier (COLON type)?
     ;
 
 // -- assignment statement
 multipleAssignmentStatement
     :   variableNames nls ASSIGN nls expression
     ;
 
+variableNames
+    :   LPAREN identifier (COMMA identifier)+ rparen
+    ;
+
 assignmentStatement
     :   target=expression nls
         op=(ASSIGN
@@ -578,7 +607,8 @@ formalParameterList
     ;
 
 formalParameter
-    :   DEF? legacyType? identifier (nls ASSIGN nls expression)?
+    :   identifier (COLON type)? (nls ASSIGN nls expression)?
+    |   DEF? legacyType? identifier (nls ASSIGN nls expression)?
     ;
 
 closureWithLabels
@@ -650,7 +680,7 @@ namedArg
 //
 type
     :   primitiveType
-    |   qualifiedClassName typeArguments?
+    |   qualifiedClassName typeArguments? QUESTION?
     ;
 
 primitiveType

modules/nf-lang/src/main/java/nextflow/script/ast/ASTNodeMarker.java

@@ -45,6 +45,9 @@ public enum ASTNodeMarker {
     // the MethodNode targeted by a variable expression (PropertyNode)
     METHOD_VARIABLE_TARGET,
 
+    // denotes a nullable type annotation (ClassNode)
+    NULLABLE,
+
     // the starting quote sequence of a string literal or gstring expression
     QUOTE_CHAR,
 

modules/nf-lang/src/main/java/nextflow/script/ast/OutputNode.java

@@ -16,6 +16,7 @@
 package nextflow.script.ast;
 
 import org.codehaus.groovy.ast.ASTNode;
+import org.codehaus.groovy.ast.ClassNode;
 import org.codehaus.groovy.ast.stmt.Statement;
 
 /**
@@ -25,10 +26,12 @@
  */
 public class OutputNode extends ASTNode {
     public final String name;
+    public final ClassNode type;
     public final Statement body;
 
-    public OutputNode(String name, Statement body) {
+    public OutputNode(String name, ClassNode type, Statement body) {
         this.name = name;
+        this.type = type;
         this.body = body;
     }
 }

modules/nf-lang/src/main/java/nextflow/script/ast/ScriptVisitorSupport.java

@@ -63,7 +63,6 @@ public void visitParam(ParamNode node) {
 
     @Override
     public void visitWorkflow(WorkflowNode node) {
-        visit(node.takes);
         visit(node.main);
         visit(node.emits);
         visit(node.publishers);

modules/nf-lang/src/main/java/nextflow/script/ast/WorkflowNode.java

@@ -38,16 +38,14 @@
  * @author Ben Sherman <[email protected]>
  */
 public class WorkflowNode extends MethodNode {
-    public final Statement takes;
     public final Statement main;
     public final Statement emits;
     public final Statement publishers;
     public final Statement onComplete;
     public final Statement onError;
 
-    public WorkflowNode(String name, Statement takes, Statement main, Statement emits, Statement publishers, Statement onComplete, Statement onError) {
-        super(name, 0, dummyReturnType(emits), dummyParams(takes), ClassNode.EMPTY_ARRAY, EmptyStatement.INSTANCE);
-        this.takes = takes;
+    public WorkflowNode(String name, Parameter[] takes, Statement main, Statement emits, Statement publishers, Statement onComplete, Statement onError) {
+        super(name, 0, dummyReturnType(emits), takes, ClassNode.EMPTY_ARRAY, EmptyStatement.INSTANCE);
         this.main = main;
         this.emits = emits;
         this.publishers = publishers;
@@ -56,7 +54,7 @@ public WorkflowNode(String name, Statement takes, Statement main, Statement emit
     }
 
     public WorkflowNode(String name, Statement main) {
-        this(name, EmptyStatement.INSTANCE, main, EmptyStatement.INSTANCE, EmptyStatement.INSTANCE, EmptyStatement.INSTANCE, EmptyStatement.INSTANCE);
+        this(name, Parameter.EMPTY_ARRAY, main, EmptyStatement.INSTANCE, EmptyStatement.INSTANCE, EmptyStatement.INSTANCE, EmptyStatement.INSTANCE);
     }
 
     public boolean isEntry() {
@@ -67,13 +65,6 @@ public boolean isCodeSnippet() {
         return getLineNumber() == -1;
     }
 
-    private static Parameter[] dummyParams(Statement takes) {
-        return asBlockStatements(takes)
-            .stream()
-            .map((stmt) -> new Parameter(ClassHelper.dynamicType(), ""))
-            .toArray(Parameter[]::new);
-    }
-
     private static ClassNode dummyReturnType(Statement emits) {
         var cn = new ClassNode(Record.class);
         asBlockStatements(emits).stream()

modules/nf-lang/src/main/java/nextflow/script/control/ResolveVisitor.java

@@ -22,6 +22,7 @@
 
 import groovy.lang.Tuple2;
 import nextflow.script.ast.ASTNodeMarker;
+import nextflow.script.types.Bag;
 import org.codehaus.groovy.GroovyBugError;
 import org.codehaus.groovy.ast.ASTNode;
 import org.codehaus.groovy.ast.ClassCodeExpressionTransformer;
@@ -60,9 +61,16 @@
  */
 public class ResolveVisitor extends ClassCodeExpressionTransformer {
 
-    public static final String[] DEFAULT_PACKAGE_PREFIXES = { "java.lang.", "java.util.", "java.io.", "java.net.", "groovy.lang.", "groovy.util." };
-
-    public static final String[] EMPTY_STRING_ARRAY = new String[0];
+    public static final ClassNode[] STANDARD_TYPES = {
+        ClassHelper.makeCached(Bag.class),
+        ClassHelper.Boolean_TYPE,
+        ClassHelper.Integer_TYPE,
+        ClassHelper.Number_TYPE,
+        ClassHelper.STRING_TYPE,
+        ClassHelper.LIST_TYPE,
+        ClassHelper.MAP_TYPE,
+        ClassHelper.SET_TYPE
+    };
 
     private SourceUnit sourceUnit;
 
@@ -113,12 +121,16 @@ public boolean resolve(ClassNode type) {
             return true;
         if( type.isResolved() )
             return true;
-        if( resolveFromModule(type) )
+        if( !type.hasPackageName() && resolveFromModule(type) )
+            return true;
+        if( !type.hasPackageName() && resolveFromStandardTypes(type) )
             return true;
         if( resolveFromLibImports(type) )
             return true;
         if( !type.hasPackageName() && resolveFromDefaultImports(type) )
             return true;
+        if( !type.hasPackageName() && resolveFromGroovyImports(type) )
+            return true;
         return resolveFromClassResolver(type.getName()) != null;
     }
 
@@ -156,10 +168,19 @@ protected boolean resolveFromModule(ClassNode type) {
         return false;
     }
 
+    protected boolean resolveFromStandardTypes(ClassNode type) {
+        for( var cn : STANDARD_TYPES ) {
+            if( cn.getNameWithoutPackage().equals(type.getName()) ) {
+                type.setRedirect(cn);
+                return true;
+            }
+        }
+        return false;
+    }
+
     protected boolean resolveFromLibImports(ClassNode type) {
-        var name = type.getName();
         for( var cn : libImports ) {
-            if( name.equals(cn.getName()) ) {
+            if( cn.getName().equals(type.getName()) ) {
                 type.setRedirect(cn);
                 return true;
             }
@@ -168,22 +189,29 @@ protected boolean resolveFromLibImports(ClassNode type) {
     }
 
     protected boolean resolveFromDefaultImports(ClassNode type) {
-        // resolve from script imports
-        var typeName = type.getName();
         for( var cn : defaultImports ) {
-            if( typeName.equals(cn.getNameWithoutPackage()) ) {
+            if( cn.getNameWithoutPackage().equals(type.getName()) ) {
                 type.setRedirect(cn);
                 return true;
             }
         }
-        // resolve from default imports cache
+        return false;
+    }
+
+    private static final String[] DEFAULT_PACKAGE_PREFIXES = { "java.lang.", "java.util.", "java.io.", "java.net.", "groovy.lang.", "groovy.util." };
+
+    private static final String[] EMPTY_STRING_ARRAY = new String[0];
+
+    protected boolean resolveFromGroovyImports(ClassNode type) {
+        var typeName = type.getName();
+        // resolve from Groovy imports cache
         var packagePrefixSet = DEFAULT_IMPORT_CLASS_AND_PACKAGES_CACHE.get(typeName);
         if( packagePrefixSet != null ) {
-            if( resolveFromDefaultImports(type, packagePrefixSet.toArray(EMPTY_STRING_ARRAY)) )
+            if( resolveFromGroovyImports(type, packagePrefixSet.toArray(EMPTY_STRING_ARRAY)) )
                 return true;
         }
-        // resolve from default imports
-        if( resolveFromDefaultImports(type, DEFAULT_PACKAGE_PREFIXES) ) {
+        // resolve from Groovy imports
+        if( resolveFromGroovyImports(type, DEFAULT_PACKAGE_PREFIXES) ) {
             return true;
         }
         if( "BigInteger".equals(typeName) ) {
@@ -202,7 +230,7 @@ protected boolean resolveFromDefaultImports(ClassNode type) {
         DEFAULT_IMPORT_CLASS_AND_PACKAGES_CACHE.putAll(VMPluginFactory.getPlugin().getDefaultImportClasses(DEFAULT_PACKAGE_PREFIXES));
     }
 
-    protected boolean resolveFromDefaultImports(ClassNode type, String[] packagePrefixes) {
+    protected boolean resolveFromGroovyImports(ClassNode type, String[] packagePrefixes) {
         var typeName = type.getName();
         for( var packagePrefix : packagePrefixes ) {
             var redirect = resolveFromClassResolver(packagePrefix + typeName);

modules/nf-lang/src/main/java/nextflow/script/control/ScriptResolveVisitor.java

@@ -85,6 +85,8 @@ public void visitParam(ParamNode node) {
 
     @Override
     public void visitWorkflow(WorkflowNode node) {
+        for( var take : node.getParameters() )
+            resolver.resolveOrFail(take.getType(), take);
         resolver.visit(node.main);
         resolver.visit(node.emits);
         resolver.visit(node.publishers);