prism 0.27.0 → 0.28.0

data/ext/prism/extension.c

@@ -32,6 +32,7 @@ ID rb_option_id_frozen_string_literal;
 ID rb_option_id_line;
 ID rb_option_id_scopes;
 ID rb_option_id_version;
+ID rb_prism_source_id_for;
 
 /******************************************************************************/
 /* IO of Ruby code                                                            */
@@ -599,8 +600,7 @@ parse_lex_input(pm_string_t *input, const pm_options_t *options, bool return_nod
 
     VALUE source_string = rb_str_new((const char *) pm_string_source(input), pm_string_length(input));
     VALUE offsets = rb_ary_new();
-    VALUE source_argv[] = { source_string, LONG2NUM(parser.start_line), offsets };
-    VALUE source = rb_class_new_instance(3, source_argv, rb_cPrismSource);
+    VALUE source = rb_funcall(rb_cPrismSource, rb_prism_source_id_for, 3, source_string, LONG2NUM(parser.start_line), offsets);
 
     parse_lex_data_t parse_lex_data = {
         .source = source,
@@ -1243,7 +1243,7 @@ static_inspect(int argc, VALUE *argv, VALUE self) {
     pm_node_t *node = ((pm_program_node_t *) program)->statements->body.nodes[0];
 
     pm_buffer_t buffer = { 0 };
-    pm_static_literal_inspect(&buffer, &parser, node);
+    pm_static_literal_inspect(&buffer, &parser.newline_list, parser.start_line, parser.encoding->name, node);
 
     rb_encoding *encoding = rb_enc_find(parser.encoding->name);
     VALUE result = rb_enc_str_new(pm_buffer_value(&buffer), pm_buffer_length(&buffer), encoding);
@@ -1379,6 +1379,8 @@ Init_prism(void) {
     rb_option_id_scopes = rb_intern_const("scopes");
     rb_option_id_version = rb_intern_const("version");
 
+    rb_prism_source_id_for = rb_intern("for");
+
     /**
      * The version of the prism library.
      */

data/ext/prism/extension.h

@@ -1,7 +1,7 @@
 #ifndef PRISM_EXT_NODE_H
 #define PRISM_EXT_NODE_H
 
-#define EXPECTED_PRISM_VERSION "0.27.0"
+#define EXPECTED_PRISM_VERSION "0.28.0"
 
 #include <ruby.h>
 #include <ruby/encoding.h>

data/include/prism/ast.h

@@ -1235,6 +1235,7 @@ typedef struct pm_and_node {
  *
  * Type: PM_ARGUMENTS_NODE
  * Flags:
+ *    PM_ARGUMENTS_NODE_FLAGS_CONTAINS_KEYWORDS
  *    PM_ARGUMENTS_NODE_FLAGS_CONTAINS_KEYWORD_SPLAT
  *
  * @extends pm_node_t
@@ -2427,21 +2428,11 @@ typedef struct pm_constant_path_node {
     struct pm_node *parent;
 
     /**
-     * ConstantPathNode#child
+     * ConstantPathNode#name
      *
-     * The right-hand node of the path. Always a `ConstantReadNode` in a
-     * valid Ruby syntax tree.
-     *
-     *     ::Foo
-     *       ^^^
-     *
-     *     self::Test
-     *           ^^^^
-     *
-     *     a.b::C
-     *          ^
+     * The name of the constant being accessed. This could be `nil` in the event of a syntax error.
      */
-    struct pm_node *child;
+    pm_constant_id_t name;
 
     /**
      * ConstantPathNode#delimiter_loc
@@ -2455,6 +2446,19 @@ typedef struct pm_constant_path_node {
      *        ^^
      */
     pm_location_t delimiter_loc;
+
+    /**
+     * ConstantPathNode#name_loc
+     *
+     * The location of the name of the constant.
+     *
+     *     ::Foo
+     *       ^^^
+     *
+     *     One::Two
+     *          ^^^
+     */
+    pm_location_t name_loc;
 } pm_constant_path_node_t;
 
 /**
@@ -2533,14 +2537,19 @@ typedef struct pm_constant_path_target_node {
     struct pm_node *parent;
 
     /**
-     * ConstantPathTargetNode#child
+     * ConstantPathTargetNode#name
      */
-    struct pm_node *child;
+    pm_constant_id_t name;
 
     /**
      * ConstantPathTargetNode#delimiter_loc
      */
     pm_location_t delimiter_loc;
+
+    /**
+     * ConstantPathTargetNode#name_loc
+     */
+    pm_location_t name_loc;
 } pm_constant_path_target_node_t;
 
 /**
@@ -5455,6 +5464,8 @@ typedef struct pm_retry_node {
  * ReturnNode
  *
  * Type: PM_RETURN_NODE
+ * Flags:
+ *    PM_RETURN_NODE_FLAGS_REDUNDANT
  *
  * @extends pm_node_t
  */
@@ -6029,8 +6040,11 @@ typedef struct pm_yield_node {
  * Flags for arguments nodes.
  */
 typedef enum pm_arguments_node_flags {
+    /** if arguments contain keywords */
+    PM_ARGUMENTS_NODE_FLAGS_CONTAINS_KEYWORDS = 1,
+
     /** if arguments contain keyword splat */
-    PM_ARGUMENTS_NODE_FLAGS_CONTAINS_KEYWORD_SPLAT = 1,
+    PM_ARGUMENTS_NODE_FLAGS_CONTAINS_KEYWORD_SPLAT = 2,
 } pm_arguments_node_flags_t;
 
 /**
@@ -6167,6 +6181,14 @@ typedef enum pm_regular_expression_flags {
     PM_REGULAR_EXPRESSION_FLAGS_FORCED_US_ASCII_ENCODING = 1024,
 } pm_regular_expression_flags_t;
 
+/**
+ * Flags for return nodes.
+ */
+typedef enum pm_return_node_flags {
+    /** a return statement that is redundant because it is the last statement in a method */
+    PM_RETURN_NODE_FLAGS_REDUNDANT = 1,
+} pm_return_node_flags_t;
+
 /**
  * Flags for shareable constant nodes.
  */

data/include/prism/diagnostic.h

@@ -152,12 +152,13 @@ typedef enum {
     PM_ERR_HASH_ROCKET,
     PM_ERR_HASH_TERM,
     PM_ERR_HASH_VALUE,
+    PM_ERR_HEREDOC_IDENTIFIER,
     PM_ERR_HEREDOC_TERM,
     PM_ERR_INCOMPLETE_QUESTION_MARK,
     PM_ERR_INCOMPLETE_VARIABLE_CLASS,
-    PM_ERR_INCOMPLETE_VARIABLE_CLASS_3_3_0,
+    PM_ERR_INCOMPLETE_VARIABLE_CLASS_3_3,
     PM_ERR_INCOMPLETE_VARIABLE_INSTANCE,
-    PM_ERR_INCOMPLETE_VARIABLE_INSTANCE_3_3_0,
+    PM_ERR_INCOMPLETE_VARIABLE_INSTANCE_3_3,
     PM_ERR_INSTANCE_VARIABLE_BARE,
     PM_ERR_INVALID_BLOCK_EXIT,
     PM_ERR_INVALID_CHARACTER,
@@ -172,14 +173,16 @@ typedef enum {
     PM_ERR_INVALID_NUMBER_DECIMAL,
     PM_ERR_INVALID_NUMBER_HEXADECIMAL,
     PM_ERR_INVALID_NUMBER_OCTAL,
-    PM_ERR_INVALID_NUMBER_UNDERSCORE,
+    PM_ERR_INVALID_NUMBER_UNDERSCORE_INNER,
+    PM_ERR_INVALID_NUMBER_UNDERSCORE_TRAILING,
     PM_ERR_INVALID_PERCENT,
     PM_ERR_INVALID_PRINTABLE_CHARACTER,
     PM_ERR_INVALID_RETRY_AFTER_ELSE,
     PM_ERR_INVALID_RETRY_AFTER_ENSURE,
     PM_ERR_INVALID_RETRY_WITHOUT_RESCUE,
+    PM_ERR_INVALID_SYMBOL,
     PM_ERR_INVALID_VARIABLE_GLOBAL,
-    PM_ERR_INVALID_VARIABLE_GLOBAL_3_3_0,
+    PM_ERR_INVALID_VARIABLE_GLOBAL_3_3,
     PM_ERR_INVALID_YIELD,
     PM_ERR_IT_NOT_ALLOWED_NUMBERED,
     PM_ERR_IT_NOT_ALLOWED_ORDINARY,
@@ -223,6 +226,7 @@ typedef enum {
     PM_ERR_PARAMETER_STAR,
     PM_ERR_PARAMETER_UNEXPECTED_FWD,
     PM_ERR_PARAMETER_WILD_LOOSE_COMMA,
+    PM_ERR_PARAMETER_UNEXPECTED_NO_KW,
     PM_ERR_PATTERN_CAPTURE_DUPLICATE,
     PM_ERR_PATTERN_EXPRESSION_AFTER_BRACKET,
     PM_ERR_PATTERN_EXPRESSION_AFTER_COMMA,
@@ -276,6 +280,9 @@ typedef enum {
     PM_ERR_UNARY_RECEIVER,
     PM_ERR_UNDEF_ARGUMENT,
     PM_ERR_UNEXPECTED_BLOCK_ARGUMENT,
+    PM_ERR_UNEXPECTED_INDEX_BLOCK,
+    PM_ERR_UNEXPECTED_INDEX_KEYWORDS,
+    PM_ERR_UNEXPECTED_SAFE_NAVIGATION,
     PM_ERR_UNEXPECTED_TOKEN_CLOSE_CONTEXT,
     PM_ERR_UNEXPECTED_TOKEN_IGNORE,
     PM_ERR_UNTIL_TERM,
@@ -296,7 +303,7 @@ typedef enum {
     PM_WARN_COMPARISON_AFTER_COMPARISON,
     PM_WARN_DOT_DOT_DOT_EOL,
     PM_WARN_EQUAL_IN_CONDITIONAL,
-    PM_WARN_EQUAL_IN_CONDITIONAL_3_3_0,
+    PM_WARN_EQUAL_IN_CONDITIONAL_3_3,
     PM_WARN_END_IN_METHOD,
     PM_WARN_DUPLICATED_HASH_KEY,
     PM_WARN_DUPLICATED_WHEN_CLAUSE,

data/include/prism/options.h

@@ -49,8 +49,8 @@ typedef enum {
     /** The current version of prism. */
     PM_OPTIONS_VERSION_LATEST = 0,
 
-    /** The vendored version of prism in CRuby 3.3.0. */
-    PM_OPTIONS_VERSION_CRUBY_3_3_0 = 1
+    /** The vendored version of prism in CRuby 3.3.x. */
+    PM_OPTIONS_VERSION_CRUBY_3_3 = 1
 } pm_options_version_t;
 
 /**

data/include/prism/parser.h

@@ -10,6 +10,7 @@
 #include "prism/ast.h"
 #include "prism/encoding.h"
 #include "prism/options.h"
+#include "prism/static_literals.h"
 #include "prism/util/pm_constant_pool.h"
 #include "prism/util/pm_list.h"
 #include "prism/util/pm_newline_list.h"
@@ -717,6 +718,15 @@ struct pm_parser {
     /** The current parsing context. */
     pm_context_node_t *current_context;
 
+    /**
+     * The hash keys for the hash that is currently being parsed. This is not
+     * usually necessary because it can pass it down the various call chains,
+     * but in the event that you're parsing a hash that is being directly
+     * pushed into another hash with **, we need to share the hash keys so that
+     * we can warn for the nested hash as well.
+     */
+    pm_static_literals_t *current_hash_keys;
+
     /**
      * The encoding functions for the current file is attached to the parser as
      * it's parsing so that it can change with a magic comment.

data/include/prism/static_literals.h

@@ -8,8 +8,7 @@
 
 #include "prism/defines.h"
 #include "prism/ast.h"
-#include "prism/node.h"
-#include "prism/parser.h"
+#include "prism/util/pm_newline_list.h"
 
 #include <assert.h>
 #include <stdbool.h>
@@ -92,12 +91,13 @@ typedef struct {
 /**
  * Add a node to the set of static literals.
  *
- * @param parser The parser that created the node.
+ * @param newline_list The list of newline offsets to use to calculate lines.
+ * @param start_line The line number that the parser starts on.
  * @param literals The set of static literals to add the node to.
  * @param node The node to add to the set.
  * @return A pointer to the node that is being overwritten, if there is one.
  */
-pm_node_t * pm_static_literals_add(const pm_parser_t *parser, pm_static_literals_t *literals, pm_node_t *node);
+pm_node_t * pm_static_literals_add(const pm_newline_list_t *newline_list, int32_t start_line, pm_static_literals_t *literals, pm_node_t *node);
 
 /**
  * Free the internal memory associated with the given static literals set.
@@ -110,9 +110,11 @@ void pm_static_literals_free(pm_static_literals_t *literals);
  * Create a string-based representation of the given static literal.
  *
  * @param buffer The buffer to write the string to.
- * @param parser The parser that created the node.
+ * @param newline_list The list of newline offsets to use to calculate lines.
+ * @param start_line The line number that the parser starts on.
+ * @param encoding_name The name of the encoding of the source being parsed.
  * @param node The node to create a string representation of.
  */
-PRISM_EXPORTED_FUNCTION void pm_static_literal_inspect(pm_buffer_t *buffer, const pm_parser_t *parser, const pm_node_t *node);
+PRISM_EXPORTED_FUNCTION void pm_static_literal_inspect(pm_buffer_t *buffer, const pm_newline_list_t *newline_list, int32_t start_line, const char *encoding_name, const pm_node_t *node);
 
 #endif

data/include/prism/version.h

@@ -14,7 +14,7 @@
 /**
  * The minor version of the Prism library as an int.
  */
-#define PRISM_VERSION_MINOR 27
+#define PRISM_VERSION_MINOR 28
 
 /**
  * The patch version of the Prism library as an int.
@@ -24,6 +24,6 @@
 /**
  * The version of the Prism library as a constant string.
  */
-#define PRISM_VERSION "0.27.0"
+#define PRISM_VERSION "0.28.0"
 
 #endif

data/lib/prism/dot_visitor.rb

@@ -1287,13 +1287,15 @@ module Prism
         digraph.edge("#{id}:parent -> #{node_id(parent)};")
       end
 
-      # child
-      table.field("child", port: true)
-      digraph.edge("#{id}:child -> #{node_id(node.child)};")
+      # name
+      table.field("name", node.name.inspect)
 
       # delimiter_loc
       table.field("delimiter_loc", location_inspect(node.delimiter_loc))
 
+      # name_loc
+      table.field("name_loc", location_inspect(node.name_loc))
+
       digraph.nodes << <<~DOT
         #{id} [
           label=<#{table.to_dot.gsub(/\n/, "\n  ")}>
@@ -1367,13 +1369,15 @@ module Prism
         digraph.edge("#{id}:parent -> #{node_id(parent)};")
       end
 
-      # child
-      table.field("child", port: true)
-      digraph.edge("#{id}:child -> #{node_id(node.child)};")
+      # name
+      table.field("name", node.name.inspect)
 
       # delimiter_loc
       table.field("delimiter_loc", location_inspect(node.delimiter_loc))
 
+      # name_loc
+      table.field("name_loc", location_inspect(node.name_loc))
+
       digraph.nodes << <<~DOT
         #{id} [
           label=<#{table.to_dot.gsub(/\n/, "\n  ")}>
@@ -3999,6 +4003,9 @@ module Prism
       table = Table.new("ReturnNode")
       id = node_id(node)
 
+      # flags
+      table.field("flags", return_node_flags_inspect(node))
+
       # keyword_loc
       table.field("keyword_loc", location_inspect(node.keyword_loc))
 
@@ -4569,6 +4576,7 @@ module Prism
     # comma-separated list.
     def arguments_node_flags_inspect(node)
       flags = [] #: Array[String]
+      flags << "contains_keywords" if node.contains_keywords?
       flags << "contains_keyword_splat" if node.contains_keyword_splat?
       flags.join(", ")
     end
@@ -4671,6 +4679,14 @@ module Prism
       flags.join(", ")
     end
 
+    # Inspect a node that has return_node_flags flags to display the flags as a
+    # comma-separated list.
+    def return_node_flags_inspect(node)
+      flags = [] #: Array[String]
+      flags << "redundant" if node.redundant?
+      flags.join(", ")
+    end
+
     # Inspect a node that has shareable_constant_node_flags flags to display the flags as a
     # comma-separated list.
     def shareable_constant_node_flags_inspect(node)

data/lib/prism/dsl.rb

@@ -10,7 +10,7 @@ module Prism
   # The DSL module provides a set of methods that can be used to create prism
   # nodes in a more concise manner. For example, instead of writing:
   #
-  #     source = Prism::Source.new("[1]")
+  #     source = Prism::Source.for("[1]")
   #
   #     Prism::ArrayNode.new(
   #       [
@@ -28,7 +28,7 @@ module Prism
   #
   # you could instead write:
   #
-  #     source = Prism::Source.new("[1]")
+  #     source = Prism::Source.for("[1]")
   #
   #     ArrayNode(
   #       IntegerNode(Prism::IntegerBaseFlags::DECIMAL, 1, Location(source, 1, 1)), source),
@@ -228,8 +228,8 @@ module Prism
     end
 
     # Create a new ConstantPathNode node
-    def ConstantPathNode(parent, child, delimiter_loc, source = nil, location = Location())
-      ConstantPathNode.new(source, parent, child, delimiter_loc, location)
+    def ConstantPathNode(parent, name, delimiter_loc, name_loc, source = nil, location = Location())
+      ConstantPathNode.new(source, parent, name, delimiter_loc, name_loc, location)
     end
 
     # Create a new ConstantPathOperatorWriteNode node
@@ -243,8 +243,8 @@ module Prism
     end
 
     # Create a new ConstantPathTargetNode node
-    def ConstantPathTargetNode(parent, child, delimiter_loc, source = nil, location = Location())
-      ConstantPathTargetNode.new(source, parent, child, delimiter_loc, location)
+    def ConstantPathTargetNode(parent, name, delimiter_loc, name_loc, source = nil, location = Location())
+      ConstantPathTargetNode.new(source, parent, name, delimiter_loc, name_loc, location)
     end
 
     # Create a new ConstantPathWriteNode node
@@ -698,8 +698,8 @@ module Prism
     end
 
     # Create a new ReturnNode node
-    def ReturnNode(keyword_loc, arguments, source = nil, location = Location())
-      ReturnNode.new(source, keyword_loc, arguments, location)
+    def ReturnNode(flags, keyword_loc, arguments, source = nil, location = Location())
+      ReturnNode.new(source, flags, keyword_loc, arguments, location)
     end
 
     # Create a new SelfNode node

data/lib/prism/ffi.rb

@@ -317,7 +317,7 @@ module Prism
         buffer.read
       end
 
-      Serialize.load_tokens(Source.new(code), serialized)
+      Serialize.load_tokens(Source.for(code), serialized)
     end
 
     def parse_common(string, code, options) # :nodoc:
@@ -329,7 +329,7 @@ module Prism
       LibRubyParser::PrismBuffer.with do |buffer|
         LibRubyParser.pm_serialize_parse_comments(buffer.pointer, string.pointer, string.length, dump_options(options))
 
-        source = Source.new(code)
+        source = Source.for(code)
         loader = Serialize::Loader.new(source, buffer.read)
 
         loader.load_header
@@ -343,7 +343,7 @@ module Prism
       LibRubyParser::PrismBuffer.with do |buffer|
         LibRubyParser.pm_serialize_parse_lex(buffer.pointer, string.pointer, string.length, dump_options(options))
 
-        source = Source.new(code)
+        source = Source.for(code)
         loader = Serialize::Loader.new(source, buffer.read)
 
         tokens = loader.load_tokens