herb 0.4.1-x86_64-linux-gnu → 0.4.3-x86_64-linux-gnu

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 913d5b1c9dcdb9655fad659e70cac6f0b9df1187020a5c86594d8f555f3665a7
-  data.tar.gz: 3ff81a4045b60e60944de4f23cdba79e116d007995b1a281adf816c08c7ad994
+  metadata.gz: 9c738e73ddaeddb08aa35b3ecd9127acdec79adae5a2545e0795e174e1a4c9af
+  data.tar.gz: 45d7300bbd0f84b0c121c9289de7a80587fa93580f0d2babf8a64246a68492a8
 SHA512:
-  metadata.gz: 3e318c828b332060bdf562bcd45e9f40b31f23a339b5884b872cfc49df51b1af71a47f81465267f4bc6940425d6d40f601939f53045ad1c59462703a0252a944
-  data.tar.gz: 49712caf44b90a614fb55474f4b4e02ef71aedf3f53e7ab1a958b1baaf45db5149e36ab5a43761d6d144e5f8c33ffdfc28432f7c747c73d225233b33655a5129
+  metadata.gz: de86bfc92f1c123967c89f62abbe837d4ccd3fdf8b88088f568c36a9945f870a9a984aa99f94dba6b3c56a74b773b3ce5c509d833d517cf0a5b1ca4290088e5e
+  data.tar.gz: 6bdeb2827459ec5f31246e24fc0bc3d7e30ab3286a56c058cac668c3890bea0d816865104d55ab5528a914751f87692a78b333ad43351d69d74e1566164356a7

data/README.md

@@ -43,11 +43,11 @@ You can use Herb programmatically in **Ruby**, as well as in **JavaScript** via
 
 For a complete overview of all available tools, libraries, and integrations, visit the [**Projects page**](https://herb-tools.dev/projects) on our documentation site.
 
-## Motiviation
+## Motivation
 
 HTML+ERB templates never really had good, accurate, and reliable tooling. While developer tooling for Ruby code improved significantly in the last few years (especially with the introduction of the new Prism parser), HTML+ERB files remained underserved, lacking fundamental support like syntax checking, auto-formatting, linting, and structural understanding.
 
-At the same time, with the rise of tools like [Hotwire](https://hotwired.dev), [Stimulus](https://stimulus.hotwired.dev), [Turbo](https://turbo.hotwired.dev), [HTMX](https://htmx.org), [Unploy](https://unpoly.com), and [Alpine.js](https://alpinejs.dev), advanced HTML templating became increasingly relevant (again). Developers expect modern, reliable, and precise tooling, especially given the robust ecosystem available to JavaScript frameworks and libraries.
+At the same time, with the rise of tools like [Hotwire](https://hotwired.dev), [Stimulus](https://stimulus.hotwired.dev), [Turbo](https://turbo.hotwired.dev), [HTMX](https://htmx.org), [Unpoly](https://unpoly.com), and [Alpine.js](https://alpinejs.dev), advanced HTML templating became increasingly relevant (again). Developers expect modern, reliable, and precise tooling, especially given the robust ecosystem available to JavaScript frameworks and libraries.
 
 Herb was built to close this tooling gap, providing proper tooling for HTML+ERB that matches what modern developers expect in the age of language servers, LLMs, and AI-driven workflows.
 

data/ext/herb/error_helpers.c

@@ -1,5 +1,5 @@
 // NOTE: This file is generated by the templates/template.rb script and should not
-// be modified manually. See /Users/marcoroth/Development/herb-release-4/templates/ext/herb/error_helpers.c.erb
+// be modified manually. See /Users/marcoroth/Development/herb-release-7/templates/ext/herb/error_helpers.c.erb
 
 #include <ruby.h>
 

data/ext/herb/error_helpers.h

@@ -1,5 +1,5 @@
 // NOTE: This file is generated by the templates/template.rb script and should not
-// be modified manually. See /Users/marcoroth/Development/herb-release-4/templates/ext/herb/error_helpers.h.erb
+// be modified manually. See /Users/marcoroth/Development/herb-release-7/templates/ext/herb/error_helpers.h.erb
 
 #ifndef HERB_EXTENSION_ERROR_HELPERS_H
 #define HERB_EXTENSION_ERROR_HELPERS_H

data/ext/herb/nodes.c

@@ -1,5 +1,5 @@
 // NOTE: This file is generated by the templates/template.rb script and should not
-// be modified manually. See /Users/marcoroth/Development/herb-release-4/templates/ext/herb/nodes.c.erb
+// be modified manually. See /Users/marcoroth/Development/herb-release-7/templates/ext/herb/nodes.c.erb
 
 #include <ruby.h>
 
@@ -415,7 +415,7 @@ static VALUE rb_erb_content_node_from_c_struct(AST_ERB_CONTENT_NODE_T* erb_conte
   VALUE erb_content_node_tag_opening = rb_token_from_c_struct(erb_content_node->tag_opening);
   VALUE erb_content_node_content = rb_token_from_c_struct(erb_content_node->content);
   VALUE erb_content_node_tag_closing = rb_token_from_c_struct(erb_content_node->tag_closing);
-  /* #<Herb::Template::AnalyzedRubyField:0x00007ffffed6eba0 @name="analyzed_ruby", @options={kind: nil}> */
+  /* #<Herb::Template::AnalyzedRubyField:0x00007fffe335db40 @name="analyzed_ruby", @options={kind: nil}> */
   VALUE erb_content_node_analyzed_ruby = Qnil;
   VALUE erb_content_node_parsed = (erb_content_node->parsed) ? Qtrue : Qfalse;
   VALUE erb_content_node_valid = (erb_content_node->valid) ? Qtrue : Qfalse;

data/ext/herb/nodes.h

@@ -1,5 +1,5 @@
 // NOTE: This file is generated by the templates/template.rb script and should not
-// be modified manually. See /Users/marcoroth/Development/herb-release-4/templates/ext/herb/nodes.h.erb
+// be modified manually. See /Users/marcoroth/Development/herb-release-7/templates/ext/herb/nodes.h.erb
 
 #ifndef HERB_EXTENSION_NODES_H
 #define HERB_EXTENSION_NODES_H

data/lib/herb/ast/nodes.rb

@@ -2,7 +2,7 @@
 # typed: true
 
 # NOTE: This file is generated by the templates/template.rb script and should not be
-# modified manually. See /Users/marcoroth/Development/herb-release-4/templates/lib/herb/ast/nodes.rb.erb
+# modified manually. See /Users/marcoroth/Development/herb-release-7/templates/lib/herb/ast/nodes.rb.erb
 
 module Herb
   module AST

data/lib/herb/cli.rb

@@ -110,8 +110,8 @@ class Herb::CLI
                   project.no_interactive = no_interactive
                   project.no_log_file = no_log_file
                   project.no_timing = no_timing
-                  project.parse!
-                  exit(0)
+                  has_issues = project.parse!
+                  exit(has_issues ? 1 : 0)
                 when "parse"
                   Herb.parse(file_content)
                 when "lex"

data/lib/herb/errors.rb

@@ -2,7 +2,7 @@
 # typed: true
 
 # NOTE: This file is generated by the templates/template.rb script and should not be
-# modified manually. See /Users/marcoroth/Development/herb-release-4/templates/lib/herb/errors.rb.erb
+# modified manually. See /Users/marcoroth/Development/herb-release-7/templates/lib/herb/errors.rb.erb
 
 module Herb
   module Errors

data/lib/herb/project.rb

@@ -366,6 +366,8 @@ module Herb
         end
 
         puts "\nResults saved to #{output_file}" unless no_log_file
+
+        problem_files.any?
       ensure
         log.close unless no_log_file
       end

data/lib/herb/version.rb

@@ -2,5 +2,5 @@
 # typed: true
 
 module Herb
-  VERSION = "0.4.1"
+  VERSION = "0.4.3"
 end

data/lib/herb/visitor.rb

@@ -2,7 +2,7 @@
 # typed: true
 
 # NOTE: This file is generated by the templates/template.rb script and should not be
-# modified manually. See /Users/marcoroth/Development/herb-release-4/templates/lib/herb/visitor.rb.erb
+# modified manually. See /Users/marcoroth/Development/herb-release-7/templates/lib/herb/visitor.rb.erb
 
 module Herb
   class Visitor

data/sig/serialized_ast_errors.rbs

@@ -2,7 +2,7 @@
 # typed: true
 
 # NOTE: This file is generated by the templates/template.rb script and should not be
-# modified manually. See /Users/marcoroth/Development/herb-release-4/templates/sig/serialized_ast_errors.rbs.erb
+# modified manually. See /Users/marcoroth/Development/herb-release-7/templates/sig/serialized_ast_errors.rbs.erb
 
 module Herb
   type serialized_unexpected_error = serialized_error & {

data/sig/serialized_ast_nodes.rbs

@@ -2,7 +2,7 @@
 # typed: true
 
 # NOTE: This file is generated by the templates/template.rb script and should not be
-# modified manually. See /Users/marcoroth/Development/herb-release-4/templates/sig/serialized_ast_nodes.rbs.erb
+# modified manually. See /Users/marcoroth/Development/herb-release-7/templates/sig/serialized_ast_nodes.rbs.erb
 
 module Herb
   type serialized_document_node = serialized_node & {

data/src/analyze.c

@@ -39,7 +39,7 @@ static analyzed_ruby_T* herb_analyze_ruby(char* source) {
   search_in_nodes(analyzed);
   search_rescue_nodes(analyzed);
   search_ensure_nodes(analyzed);
-  search_yield_nodes(analyzed);
+  search_yield_nodes(analyzed->root, analyzed);
   search_block_closing_nodes(analyzed);
 
   return analyzed;
@@ -95,8 +95,13 @@ static control_type_t detect_control_type(AST_ERB_CONTENT_NODE_T* erb_node) {
 
   if (!ruby) { return CONTROL_TYPE_UNKNOWN; }
 
-  if (ruby->valid) { return CONTROL_TYPE_UNKNOWN; }
+  if (ruby->valid) {
+    if (has_yield_node(ruby)) { return CONTROL_TYPE_YIELD; }
+    return CONTROL_TYPE_UNKNOWN;
+  }
 
+  if (has_yield_node(ruby)) { return CONTROL_TYPE_YIELD; }
+  if (has_block_node(ruby)) { return CONTROL_TYPE_BLOCK; }
   if (has_if_node(ruby)) { return CONTROL_TYPE_IF; }
   if (has_elsif_node(ruby)) { return CONTROL_TYPE_ELSIF; }
   if (has_else_node(ruby)) { return CONTROL_TYPE_ELSE; }
@@ -112,8 +117,6 @@ static control_type_t detect_control_type(AST_ERB_CONTENT_NODE_T* erb_node) {
   if (has_while_node(ruby)) { return CONTROL_TYPE_WHILE; }
   if (has_until_node(ruby)) { return CONTROL_TYPE_UNTIL; }
   if (has_for_node(ruby)) { return CONTROL_TYPE_FOR; }
-  if (has_block_node(ruby)) { return CONTROL_TYPE_BLOCK; }
-  if (has_yield_node(ruby)) { return CONTROL_TYPE_YIELD; }
   if (has_block_closing(ruby)) { return CONTROL_TYPE_BLOCK_CLOSE; }
 
   return CONTROL_TYPE_UNKNOWN;
@@ -1020,10 +1023,22 @@ static array_T* rewrite_node_array(AST_NODE_T* node, array_T* array, analyze_rub
       case CONTROL_TYPE_UNTIL:
       case CONTROL_TYPE_FOR:
       case CONTROL_TYPE_BLOCK:
-      case CONTROL_TYPE_YIELD:
         index = process_control_structure(node, array, index, new_array, context, type);
         continue;
 
+      case CONTROL_TYPE_YIELD: {
+        AST_NODE_T* yield_node = create_control_node(erb_node, array_init(8), NULL, NULL, type);
+
+        if (yield_node) {
+          array_append(new_array, yield_node);
+        } else {
+          array_append(new_array, item);
+        }
+
+        index++;
+        break;
+      }
+
       default:
         array_append(new_array, item);
         index++;

data/src/analyze_helpers.c

@@ -279,10 +279,14 @@ bool search_ensure_nodes(analyzed_ruby_T* analyzed) {
   return false;
 }
 
-bool search_yield_nodes(analyzed_ruby_T* analyzed) {
-  if (has_error_message(analyzed, "Invalid yield")) {
+bool search_yield_nodes(const pm_node_t* node, void* data) {
+  analyzed_ruby_T* analyzed = (analyzed_ruby_T*) data;
+
+  if (node->type == PM_YIELD_NODE) {
     analyzed->has_yield_node = true;
     return true;
+  } else {
+    pm_visit_child_nodes(node, search_yield_nodes, analyzed);
   }
 
   return false;

data/src/ast_nodes.c

@@ -1,5 +1,5 @@
 // NOTE: This file is generated by the templates/template.rb script and should not
-// be modified manually. See /Users/marcoroth/Development/herb-release-4/templates/src/ast_nodes.c.erb
+// be modified manually. See /Users/marcoroth/Development/herb-release-7/templates/src/ast_nodes.c.erb
 
 #include <stdio.h>
 #include <stdbool.h>

data/src/ast_pretty_print.c

@@ -1,5 +1,5 @@
 // NOTE: This file is generated by the templates/template.rb script and should not
-// be modified manually. See /Users/marcoroth/Development/herb-release-4/templates/src/ast_pretty_print.c.erb
+// be modified manually. See /Users/marcoroth/Development/herb-release-7/templates/src/ast_pretty_print.c.erb
 
 #include "include/ast_node.h"
 #include "include/ast_nodes.h"

data/src/errors.c

@@ -1,5 +1,5 @@
 // NOTE: This file is generated by the templates/template.rb script and should not
-// be modified manually. See /Users/marcoroth/Development/herb-release-4/templates/src/errors.c.erb
+// be modified manually. See /Users/marcoroth/Development/herb-release-7/templates/src/errors.c.erb
 
 #include "include/array.h"
 #include "include/errors.h"

data/src/include/analyze_helpers.h

@@ -44,6 +44,6 @@ bool search_when_nodes(analyzed_ruby_T* analyzed);
 bool search_in_nodes(analyzed_ruby_T* analyzed);
 bool search_rescue_nodes(analyzed_ruby_T* analyzed);
 bool search_ensure_nodes(analyzed_ruby_T* analyzed);
-bool search_yield_nodes(analyzed_ruby_T* analyzed);
+bool search_yield_nodes(const pm_node_t* node, void* data);
 
 #endif

data/src/include/ast_nodes.h

@@ -1,5 +1,5 @@
 // NOTE: This file is generated by the templates/template.rb script and should not
-// be modified manually. See /Users/marcoroth/Development/herb-release-4/templates/src/include/ast_nodes.h.erb
+// be modified manually. See /Users/marcoroth/Development/herb-release-7/templates/src/include/ast_nodes.h.erb
 
 #ifndef HERB_AST_NODES_H
 #define HERB_AST_NODES_H

data/src/include/ast_pretty_print.h

@@ -1,5 +1,5 @@
 // NOTE: This file is generated by the templates/template.rb script and should not
-// be modified manually. See /Users/marcoroth/Development/herb-release-4/templates/src/include/ast_pretty_print.h.erb
+// be modified manually. See /Users/marcoroth/Development/herb-release-7/templates/src/include/ast_pretty_print.h.erb
 
 #ifndef HERB_AST_PRETTY_PRINT_H
 #define HERB_AST_PRETTY_PRINT_H

data/src/include/errors.h

@@ -1,5 +1,5 @@
 // NOTE: This file is generated by the templates/template.rb script and should not
-// be modified manually. See /Users/marcoroth/Development/herb-release-4/templates/src/include/errors.h.erb
+// be modified manually. See /Users/marcoroth/Development/herb-release-7/templates/src/include/errors.h.erb
 
 #ifndef HERB_ERRORS_H
 #define HERB_ERRORS_H

data/src/include/utf8.h

@@ -0,0 +1,11 @@
+#ifndef HERB_UTF8_H
+#define HERB_UTF8_H
+
+#include <stdbool.h>
+#include <stdlib.h>
+
+int utf8_char_byte_length(unsigned char first_byte);
+int utf8_sequence_length(const char* str, size_t position, size_t max_length);
+bool utf8_is_valid_continuation_byte(unsigned char byte);
+
+#endif

data/src/include/version.h

@@ -1,6 +1,6 @@
 #ifndef HERB_VERSION_H
 #define HERB_VERSION_H
 
-#define HERB_VERSION "0.4.1"
+#define HERB_VERSION "0.4.3"
 
 #endif

data/src/lexer.c

@@ -1,6 +1,7 @@
 #include "include/buffer.h"
 #include "include/lexer_peek_helpers.h"
 #include "include/token.h"
+#include "include/utf8.h"
 #include "include/util.h"
 
 #include <ctype.h>
@@ -84,6 +85,23 @@ static void lexer_advance(lexer_T* lexer) {
   }
 }
 
+static void lexer_advance_utf8_bytes(lexer_T* lexer, int byte_count) {
+  if (byte_count <= 0) { return; }
+
+  if (lexer_has_more_characters(lexer) && !lexer_eof(lexer)) {
+    if (!is_newline(lexer->current_character)) { lexer->current_column++; }
+
+    lexer->current_position += byte_count;
+
+    if (lexer->current_position >= lexer->source_length) {
+      lexer->current_position = lexer->source_length;
+      lexer->current_character = '\0';
+    } else {
+      lexer->current_character = lexer->source[lexer->current_position];
+    }
+  }
+}
+
 static void lexer_advance_by(lexer_T* lexer, const size_t count) {
   for (size_t i = 0; i < count; i++) {
     lexer_advance(lexer);
@@ -116,6 +134,35 @@ static token_T* lexer_advance_current(lexer_T* lexer, const token_type_T type) {
   return lexer_advance_with(lexer, (char[]) { lexer->current_character, '\0' }, type);
 }
 
+static token_T* lexer_advance_utf8_character(lexer_T* lexer, const token_type_T type) {
+  int char_byte_length = utf8_sequence_length(lexer->source, lexer->current_position, lexer->source_length);
+
+  if (char_byte_length <= 1) { return lexer_advance_current(lexer, type); }
+
+  char* utf8_char = malloc(char_byte_length + 1);
+
+  if (!utf8_char) { return lexer_advance_current(lexer, type); }
+
+  for (int i = 0; i < char_byte_length; i++) {
+    if (lexer->current_position + i >= lexer->source_length) {
+      free(utf8_char);
+      return lexer_advance_current(lexer, type);
+    }
+
+    utf8_char[i] = lexer->source[lexer->current_position + i];
+  }
+
+  utf8_char[char_byte_length] = '\0';
+
+  lexer_advance_utf8_bytes(lexer, char_byte_length);
+
+  token_T* token = token_init(utf8_char, type, lexer);
+
+  free(utf8_char);
+
+  return token;
+}
+
 static token_T* lexer_match_and_advance(lexer_T* lexer, const char* value, const token_type_T type) {
   if (strncmp(lexer->source + lexer->current_position, value, strlen(value)) == 0) {
     return lexer_advance_with(lexer, value, type);
@@ -232,7 +279,7 @@ token_T* lexer_next_token(lexer_T* lexer) {
   if (isspace(lexer->current_character)) { return lexer_parse_whitespace(lexer); }
 
   if (lexer->current_character == '\xC2' && lexer_peek(lexer, 1) == '\xA0') {
-    return lexer_advance_with(lexer, "\xC2\xA0", TOKEN_NBSP);
+    return lexer_advance_utf8_character(lexer, TOKEN_NBSP);
   }
 
   switch (lexer->current_character) {
@@ -282,7 +329,7 @@ token_T* lexer_next_token(lexer_T* lexer) {
     default: {
       if (isalnum(lexer->current_character)) { return lexer_parse_identifier(lexer); }
 
-      return lexer_advance_current(lexer, TOKEN_CHARACTER);
+      return lexer_advance_utf8_character(lexer, TOKEN_CHARACTER);
     }
   }
 }

data/src/parser.c

@@ -9,6 +9,7 @@
 #include "include/parser_helpers.h"
 #include "include/token.h"
 #include "include/token_matchers.h"
+#include "include/util.h"
 
 #include <stdio.h>
 #include <stdlib.h>
@@ -184,14 +185,96 @@ static AST_HTML_TEXT_NODE_T* parser_parse_text_content(parser_T* parser, array_T
 
 static AST_HTML_ATTRIBUTE_NAME_NODE_T* parser_parse_html_attribute_name(parser_T* parser) {
   array_T* errors = array_init(8);
-  token_T* identifier = parser_consume_if_present(parser, TOKEN_IDENTIFIER);
 
-  if (identifier == NULL) { parser_append_unexpected_token_error(parser, TOKEN_IDENTIFIER, errors); }
+  token_T* at_token = parser_consume_if_present(parser, TOKEN_AT);
+  token_T* first_token = NULL;
+
+  if (at_token != NULL) {
+    first_token = parser_consume_if_present(parser, TOKEN_IDENTIFIER);
+
+    if (first_token == NULL) {
+      parser_append_unexpected_token_error(parser, TOKEN_IDENTIFIER, errors);
+
+      AST_HTML_ATTRIBUTE_NAME_NODE_T* attribute_name =
+        ast_html_attribute_name_node_init(at_token, at_token->location->start, at_token->location->end, errors);
+
+      token_free(at_token);
+
+      return attribute_name;
+    }
+  } else {
+    first_token = parser_consume_if_present(parser, TOKEN_IDENTIFIER);
+
+    if (first_token == NULL) {
+      parser_append_unexpected_token_error(parser, TOKEN_IDENTIFIER, errors);
+      AST_HTML_ATTRIBUTE_NAME_NODE_T* attribute_name = ast_html_attribute_name_node_init(NULL, NULL, NULL, errors);
+      return attribute_name;
+    }
+  }
+
+  buffer_T name_buffer = buffer_new();
+
+  position_T* start_position;
+
+  if (at_token != NULL) {
+    buffer_append(&name_buffer, at_token->value);
+    start_position = position_copy(at_token->location->start);
+  } else {
+    start_position = position_copy(first_token->location->start);
+  }
+
+  buffer_append(&name_buffer, first_token->value);
+
+  position_T* end_position = position_copy(first_token->location->end);
+  size_t range_end = first_token->range->to;
+
+  while (parser->current_token->type == TOKEN_CHARACTER && parser->current_token->value
+         && strcmp(parser->current_token->value, ".") == 0) {
+
+    token_T* dot_token = parser_advance(parser);
+
+    buffer_append(&name_buffer, dot_token->value);
+    position_free(end_position);
+
+    end_position = position_copy(dot_token->location->end);
+    range_end = dot_token->range->to;
+
+    token_free(dot_token);
+
+    if (parser->current_token->type == TOKEN_IDENTIFIER) {
+      token_T* next_identifier = parser_advance(parser);
+
+      buffer_append(&name_buffer, next_identifier->value);
+      position_free(end_position);
+
+      end_position = position_copy(next_identifier->location->end);
+      range_end = next_identifier->range->to;
+      token_free(next_identifier);
+    } else {
+      break;
+    }
+  }
+
+  token_T* combined_token = calloc(1, sizeof(token_T));
+  combined_token->value = herb_strdup(name_buffer.value);
+  combined_token->type = TOKEN_IDENTIFIER;
+  combined_token->location =
+    location_from(start_position->line, start_position->column, end_position->line, end_position->column);
+
+  size_t range_start = at_token != NULL ? at_token->range->from : first_token->range->from;
+  combined_token->range = range_init(range_start, range_end);
 
   AST_HTML_ATTRIBUTE_NAME_NODE_T* attribute_name =
-    ast_html_attribute_name_node_init(identifier, identifier->location->start, identifier->location->end, errors);
+    ast_html_attribute_name_node_init(combined_token, start_position, end_position, errors);
 
-  token_free(identifier);
+  buffer_free(&name_buffer);
+  position_free(start_position);
+  position_free(end_position);
+  token_free(first_token);
+
+  if (at_token != NULL) { token_free(at_token); }
+
+  token_free(combined_token);
 
   return attribute_name;
 }
@@ -390,10 +473,15 @@ static AST_HTML_OPEN_TAG_NODE_T* parser_parse_html_open_tag(parser_T* parser) {
       continue;
     }
 
+    if (parser->current_token->type == TOKEN_AT) {
+      array_append(children, parser_parse_html_attribute(parser));
+      continue;
+    }
+
     parser_append_unexpected_error(
       parser,
       "Unexpected Token",
-      "TOKEN_IDENTIFIER, TOKEN_ERB_START,TOKEN_WHITESPACE, or TOKEN_NEWLINE",
+      "TOKEN_IDENTIFIER, TOKEN_AT, TOKEN_ERB_START,TOKEN_WHITESPACE, or TOKEN_NEWLINE",
       errors
     );
   }
@@ -441,6 +529,12 @@ static AST_HTML_CLOSE_TAG_NODE_T* parser_parse_html_close_tag(parser_T* parser)
 
   token_T* tag_opening = parser_consume_expected(parser, TOKEN_HTML_TAG_START_CLOSE, errors);
   token_T* tag_name = parser_consume_expected(parser, TOKEN_IDENTIFIER, errors);
+
+  while (token_is_any_of(parser, TOKEN_WHITESPACE, TOKEN_NEWLINE)) {
+    token_T* whitespace = parser_advance(parser);
+    token_free(whitespace);
+  }
+
   token_T* tag_closing = parser_consume_expected(parser, TOKEN_HTML_TAG_END, errors);
 
   if (tag_name != NULL && is_void_element(tag_name->value) && parser_in_svg_context(parser) == false) {
@@ -618,12 +712,14 @@ static void parser_parse_in_data_state(parser_T* parser, array_T* children, arra
     if (token_is_any_of(
           parser,
           TOKEN_AMPERSAND,
+          TOKEN_AT,
           TOKEN_CHARACTER,
           TOKEN_COLON,
           TOKEN_DASH,
           TOKEN_EQUALS,
           TOKEN_EXCLAMATION,
           TOKEN_IDENTIFIER,
+          TOKEN_NBSP,
           TOKEN_NEWLINE,
           TOKEN_PERCENT,
           TOKEN_QUOTE,
@@ -639,8 +735,8 @@ static void parser_parse_in_data_state(parser_T* parser, array_T* children, arra
     parser_append_unexpected_error(
       parser,
       "Unexpected token",
-      "TOKEN_ERB_START, TOKEN_HTML_DOCTYPE, TOKEN_HTML_COMMENT_START, TOKEN_IDENTIFIER, TOKEN_WHITESPACE, or "
-      "TOKEN_NEWLINE",
+      "TOKEN_ERB_START, TOKEN_HTML_DOCTYPE, TOKEN_HTML_COMMENT_START, TOKEN_IDENTIFIER, TOKEN_WHITESPACE, "
+      "TOKEN_NBSP, TOKEN_AT, or TOKEN_NEWLINE",
       errors
     );
   }

data/src/utf8.c

@@ -0,0 +1,46 @@
+#include "include/utf8.h"
+
+// UTF-8 byte patterns:
+//   0xxxxxxx = 1 byte (ASCII)
+//   110xxxxx = 2 bytes
+//   1110xxxx = 3 bytes
+//   11110xxx = 4 bytes
+int utf8_char_byte_length(unsigned char first_byte) {
+  if ((first_byte & 0x80) == 0) {
+    return 1;
+  } else if ((first_byte & 0xE0) == 0xC0) {
+    return 2;
+  } else if ((first_byte & 0xF0) == 0xE0) {
+    return 3;
+  } else if ((first_byte & 0xF8) == 0xF0) {
+    return 4;
+  }
+
+  return 1;
+}
+
+// Continuation bytes have pattern 10xxxxxx
+bool utf8_is_valid_continuation_byte(unsigned char byte) {
+  return (byte & 0xC0) == 0x80;
+}
+
+int utf8_sequence_length(const char* str, size_t position, size_t max_length) {
+  if (position >= max_length) { return 0; }
+
+  unsigned char first_byte = (unsigned char) str[position];
+  int expected_length = utf8_char_byte_length(first_byte);
+
+  if (position + expected_length > max_length) {
+    return 1; // Not enough bytes, treat as single byte
+  }
+
+  if (expected_length > 1) {
+    for (int i = 1; i < expected_length; i++) {
+      if (!utf8_is_valid_continuation_byte((unsigned char) str[position + i])) {
+        return 1; // Invalid continuation byte, treat first byte as single byte
+      }
+    }
+  }
+
+  return expected_length;
+}

data/src/visitor.c

@@ -1,5 +1,5 @@
 // NOTE: This file is generated by the templates/template.rb script and should not
-// be modified manually. See /Users/marcoroth/Development/herb-release-4/templates/src/visitor.c.erb
+// be modified manually. See /Users/marcoroth/Development/herb-release-7/templates/src/visitor.c.erb
 
 #include <stdio.h>
 

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification
 name: herb
 version: !ruby/object:Gem::Version
-  version: 0.4.1
+  version: 0.4.3
 platform: x86_64-linux-gnu
 authors:
 - Marco Roth
 bindir: exe
 cert_chain: []
-date: 2025-07-22 00:00:00.000000000 Z
+date: 2025-08-03 00:00:00.000000000 Z
 dependencies: []
 description: Powerful and seamless HTML-aware ERB parsing and tooling.
 email:
@@ -125,6 +125,7 @@ files:
 - src/include/token.h
 - src/include/token_matchers.h
 - src/include/token_struct.h
+- src/include/utf8.h
 - src/include/util.h
 - src/include/version.h
 - src/include/visitor.h
@@ -144,6 +145,7 @@ files:
 - src/ruby_parser.c
 - src/token.c
 - src/token_matchers.c
+- src/utf8.c
 - src/util.c
 - src/visitor.c
 homepage: https://herb-tools.dev