tree-sitter-objectscript 1.9.22 → 1.10.1

package/common/scanner.h

@@ -314,6 +314,93 @@ static bool is_statement_or_class_keyword(const int32_t *text, uint32_t len) {
   return false;
 }
 
+// Consumes an ObjectScript interpolation (#(...)# or ##(...)##) embedded in
+// &html/&xml text. Assumes lookahead is '#'. Inside the parens, '>' and '\''
+// are ObjectScript operators, not XML syntax, so they must not be treated as
+// tag ends or attribute quotes.
+static void skip_cos_interpolation(TSLexer *lexer) {
+  advance(lexer);
+  if (lexer->lookahead == '#') {
+    advance(lexer);
+  }
+  if (lexer->lookahead != '(') {
+    return;
+  }
+  int depth = 0;
+  while (!lexer->eof(lexer) && lexer->lookahead != '\n') {
+    int32_t c = lexer->lookahead;
+    if (c == '"') {
+      advance(lexer);
+      while (!lexer->eof(lexer) && lexer->lookahead != '\n') {
+        if (lexer->lookahead == '"') {
+          advance(lexer);
+          if (lexer->lookahead != '"') {
+            break;
+          }
+        }
+        advance(lexer);
+      }
+      continue;
+    }
+    advance(lexer);
+    if (c == '(') {
+      depth++;
+    } else if (c == ')') {
+      depth--;
+      if (depth == 0) {
+        break;
+      }
+    }
+  }
+  // consume the closing # (or ##)
+  if (lexer->lookahead == '#') {
+    advance(lexer);
+    if (lexer->lookahead == '#') {
+      advance(lexer);
+    }
+  }
+}
+
+// Consumes the rest of an XML tag up to (not including) the closing '>'.
+// A quote only opens an attribute value directly after '=', so stray quotes in
+// hand-written HTML (e.g. height=24") don't swallow the rest of the file.
+static void skip_xml_tag_body(TSLexer *lexer) {
+  int32_t prev = 0;
+  while (!lexer->eof(lexer) && lexer->lookahead != '>') {
+    int32_t c = lexer->lookahead;
+    if (c == '#') {
+      skip_cos_interpolation(lexer);
+      prev = '#';
+      continue;
+    }
+    if ((c == '"' || c == '\'') && prev == '=') {
+      advance(lexer);
+      while (!lexer->eof(lexer)) {
+        if (lexer->lookahead == '#') {
+          skip_cos_interpolation(lexer);
+          continue;
+        }
+        if (lexer->lookahead == c) {
+          advance(lexer);
+          if (lexer->lookahead == c) {
+            advance(lexer);
+          } else {
+            break;
+          }
+        } else {
+          advance(lexer);
+        }
+      }
+      prev = c;
+      continue;
+    }
+    if (!iswspace(c)) {
+      prev = c;
+    }
+    advance(lexer);
+  }
+}
+
 static bool lex_fenced_text(TSLexer *lexer,
                             enum ObjectScript_Core_Scanner_TokenType desired_symbol, int32_t l_delim,
                             int32_t r_delim, bool is_xml, bool has_python_comment, bool is_text) {
@@ -338,6 +425,10 @@ static bool lex_fenced_text(TSLexer *lexer,
         advance(lexer);
     }
     int32_t c = lexer->lookahead;
+    if (c == '#' && is_xml) {
+        skip_cos_interpolation(lexer);
+        continue;
+    }
     if (c == '<' && is_xml) {
         bool comment = true;
         bool is_code = false;
@@ -417,27 +508,7 @@ static bool lex_fenced_text(TSLexer *lexer,
                 }
                 advance(lexer);
             }
-            while (!lexer->eof(lexer) && lexer->lookahead != '>') {
-                if (lexer->lookahead == '"' || lexer->lookahead == '\'') {
-                  int32_t quote = lexer->lookahead;
-                  advance(lexer);
-                  while (!lexer->eof(lexer)) {
-                    if (lexer->lookahead == quote) {
-                      advance(lexer);
-                      if (lexer->lookahead == quote) {
-                        advance(lexer);
-                      } else {
-                        break;
-                      }
-                    } else {
-                      advance(lexer);
-                    }
-                  }
-                  continue;
-                }
-                advance(lexer);
-                continue;
-            }
+            skip_xml_tag_body(lexer);
             if (lexer->lookahead == '>') {
                 advance(lexer);
             }
@@ -470,27 +541,7 @@ static bool lex_fenced_text(TSLexer *lexer,
         }
 
         else {
-            while (!lexer->eof(lexer) && lexer->lookahead != '>') {
-                if (lexer->lookahead == '"' || lexer->lookahead == '\'') {
-                  int32_t quote = lexer->lookahead;
-                  advance(lexer);
-                  while (!lexer->eof(lexer)) {
-                     if (lexer->lookahead == quote) {
-                      advance(lexer);
-                      if (lexer->lookahead == quote) {
-                        advance(lexer);
-                      } else {
-                        break;
-                      }
-                    } else {
-                      advance(lexer);
-                    }
-                  }
-                  continue;
-                }
-                advance(lexer);
-                continue;
-            }
+            skip_xml_tag_body(lexer);
             // consume the >
             advance(lexer);
             continue;

package/core/grammar.js

@@ -895,8 +895,14 @@ module.exports = grammar(objectscript_expr, {
           $.expression,
         ),
       ),
+    // #dim is a compile-time declaration, so IRIS allows it between the
+    // try block and its catch.
     command_trycatch: ($) =>
-      seq(build_block_no_params($, $.keyword_try), $.catch_block),
+      seq(
+        build_block_no_params($, $.keyword_try),
+        repeat($.pound_dim),
+        $.catch_block,
+      ),
 
     catch_block: ($) =>
       build_block($, $.keyword_catch, build_arguments($._target_var_arg)),

package/core/src/node-types.json

@@ -1161,6 +1161,10 @@
           "type": "post_conditional",
           "named": true
         },
+        {
+          "type": "pound_dim",
+          "named": true
+        },
         {
           "type": "statement",
           "named": true

package/core/src/parser.c

@@ -1,4 +1,4 @@
 [diffend] Oversized file quarantined before diffing.
 name: package/core/src/parser.c
-size: 44871078 bytes
-sha256: 8423e58ee56778fbc8bc68c35bcf9655568436dacea06337fca933c829fc18b5
+size: 45411547 bytes
+sha256: db4d938f4fa0731c1a1a4e9b3e4e6c15467fc1f895e5dcc2071fbe0351465e6b

package/expr/grammar.js

@@ -304,7 +304,7 @@ module.exports = grammar({
       ),
     keyword_super: (_) => /##SUPER/i,
     superclass_method_call: ($) => seq($.keyword_super, $.method_args),
-    expression_call: ($) => seq(alias(/##expression/i, $.keyword_pound_expression), $.method_args),
+    expression_call: ($) => seq(alias(choice(/##expression/i, /##safeexpression/i), $.keyword_pound_expression), $.method_args),
     extrinsic_function: ($) =>
       // $$tag^rtn or $$@var
       prec.left(seq('$$', $._extrinsic_reference, optional($.method_args))),