Fix HCL parser for comments with CR-only line endings (#6532)

* Support CR-only line endings in HCL line comments

The HCL lexer's LINE_COMMENT rule only supported LF (\n) or CRLF (\r\n)
line endings, but not CR-only (\r) line endings (classic Mac-style).

This caused parsing failures for files with CR-only line endings after
comments, such as "# comment\rlocals { a = 1 }".

The fix changes the lexer rule from:
  LINE_COMMENT : ('//' | '#') ~[\r\n]* '\r'? ('\n' | EOF) -> channel(HIDDEN);
to:
  LINE_COMMENT : ('//' | '#') ~[\r\n]* ('\r\n' | '\r' | '\n' | EOF) -> channel(HIDDEN);

This now accepts all three line ending styles: CRLF, LF, and CR.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* Add comprehensive tests for HCL comments at file boundaries

Add extensive test coverage for HCL comments at the start and end of
Terraform/HCL files, including:

- Comments as the only content in a file
- Comments at the very first character of files
- Comments at the end without trailing newlines
- Multiple comments at start/end
- Block comments at file boundaries
- CRLF and CR-only line endings
- Unicode and emoji in comments
- Realistic Terraform patterns with header/footer comments
- Edge cases with heredocs and comments

The CR-only line ending tests pass after the lexer grammar fix.

One test is disabled pending further investigation:
- emojiInComment: Emoji (surrogate pair) in comments at the start of
  the file causes cursor tracking issues due to a mismatch between
  ANTLR character indices and the parser visitor's code point tracking.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* Restore original headers

* Copy the `antlrGeneration` approach from rewrite-toml

* Handle all newlines in NEWLINE

* Shorten LINE_COMMENT

---------

Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

Author: timtebeek

rewrite-hcl/build.gradle.kts

@@ -2,8 +2,9 @@ plugins {
     id("org.openrewrite.build.language-library")
 }
 
-// run manually with `./gradlew rewrite-hcl:generateAntlrSources` when you need to regenerate
-// be sure to use `implementation("org.antlr:antlr4:4.13.2")` below when generating
+val antlrGeneration by configurations.creating {
+    extendsFrom(configurations.implementation.get())
+}
 tasks.register<JavaExec>("generateAntlrSources") {
     mainClass.set("org.antlr.v4.Tool")
 
@@ -13,7 +14,7 @@ tasks.register<JavaExec>("generateAntlrSources") {
             "-visitor"
     ) + fileTree("src/main/antlr").matching { include("**/*.g4") }.map { it.path }
 
-    classpath = sourceSets["main"].runtimeClasspath
+    classpath = antlrGeneration
 
     finalizedBy("licenseFormat")
 }
@@ -28,6 +29,10 @@ dependencies {
     implementation("org.antlr:antlr4-runtime:4.13.2")
     implementation("io.micrometer:micrometer-core:1.9.+")
 
+    antlrGeneration("org.antlr:antlr4:4.13.2"){
+        exclude(group = "com.ibm.icu", module = "icu4j")
+    }
+
     testImplementation(project(":rewrite-test"))
     testImplementation("org.junit-pioneer:junit-pioneer:2.0.0")
 }

rewrite-hcl/src/main/antlr/HCLLexer.g4

@@ -56,10 +56,10 @@ Identifier
 
 // Lexical Elements - Comments and Whitespace
 // https://github.com/hashicorp/hcl2/blob/master/hcl/hclsyntax/spec.md#comments-and-whitespace
-WS              : [ \t\r\u000C]+                           -> channel(HIDDEN);
+WS              : [ \t\u000C]+                             -> channel(HIDDEN);
 COMMENT         : '/*' .*? '*/'                            -> channel(HIDDEN);
-LINE_COMMENT    : ('//' | '#') ~[\r\n]* '\r'? ('\n' | EOF) -> channel(HIDDEN);
-NEWLINE         : '\n'                                     -> channel(HIDDEN);
+LINE_COMMENT    : ('//' | '#') ~[\r\n]* -> channel(HIDDEN);
+NEWLINE         : ('\r\n' | '\r' | '\n')                   -> channel(HIDDEN);
 
 fragment LetterOrDigit
     : Letter

rewrite-hcl/src/main/java/org/openrewrite/hcl/internal/grammar/HCLLexer.interp

@@ -49,6 +49,7 @@ HP_COMMENT=48
 HP_LINE_COMMENT=49
 HTemplateLiteral=50
 HTemplateLiteralChar=51
+H_NEWLINE=52
 'if'=3
 'in'=4
 'null'=6
@@ -81,3 +82,4 @@ HTemplateLiteralChar=51
 '%'=41
 '...'=42
 '~'=43
+'\n'=52

rewrite-hcl/src/main/java/org/openrewrite/hcl/internal/grammar/HCLLexer.java

@@ -39,7 +39,8 @@ public class HCLParser extends Parser {
 		OR=27, NEQ=28, GT=29, QUESTION=30, RBRACK=31, RPAREN=32, MUL=33, NOT=34, 
 		LEQ=35, DOT=36, DIV=37, GEQ=38, ARROW=39, COMMA=40, MOD=41, ELLIPSIS=42, 
 		TILDE=43, TEMPLATE_INTERPOLATION_START=44, TemplateStringLiteral=45, TemplateStringLiteralChar=46, 
-		HP_WS=47, HP_COMMENT=48, HP_LINE_COMMENT=49, HTemplateLiteral=50, HTemplateLiteralChar=51;
+		HP_WS=47, HP_COMMENT=48, HP_LINE_COMMENT=49, HTemplateLiteral=50, HTemplateLiteralChar=51, 
+		H_NEWLINE=52;
 	public static final int
 		RULE_configFile = 0, RULE_body = 1, RULE_bodyContent = 2, RULE_attribute = 3, 
 		RULE_block = 4, RULE_blockLabel = 5, RULE_expression = 6, RULE_exprTerm = 7, 
@@ -74,7 +75,8 @@ private static String[] makeLiteralNames() {
 			null, null, null, null, null, null, null, null, "'+'", "'&&'", "'=='", 
 			"'<'", "'::'", "':'", "'['", "'('", "'-'", "'||'", "'!='", "'>'", "'?'", 
 			"']'", "')'", "'*'", "'!'", "'<='", "'.'", "'/'", "'>='", "'=>'", "','", 
-			"'%'", "'...'", "'~'"
+			"'%'", "'...'", "'~'", null, null, null, null, null, null, null, null, 
+			"'\\n'"
 		};
 	}
 	private static final String[] _LITERAL_NAMES = makeLiteralNames();
@@ -87,7 +89,7 @@ private static String[] makeSymbolicNames() {
 			"NEQ", "GT", "QUESTION", "RBRACK", "RPAREN", "MUL", "NOT", "LEQ", "DOT", 
 			"DIV", "GEQ", "ARROW", "COMMA", "MOD", "ELLIPSIS", "TILDE", "TEMPLATE_INTERPOLATION_START", 
 			"TemplateStringLiteral", "TemplateStringLiteralChar", "HP_WS", "HP_COMMENT", 
-			"HP_LINE_COMMENT", "HTemplateLiteral", "HTemplateLiteralChar"
+			"HP_LINE_COMMENT", "HTemplateLiteral", "HTemplateLiteralChar", "H_NEWLINE"
 		};
 	}
 	private static final String[] _SYMBOLIC_NAMES = makeSymbolicNames();
@@ -3511,7 +3513,7 @@ private boolean exprTerm_sempred(ExprTermContext _localctx, int predIndex) {
 	}
 
 	public static final String _serializedATN =
-		"\u0004\u00013\u0187\u0002\u0000\u0007\u0000\u0002\u0001\u0007\u0001\u0002"+
+		"\u0004\u00014\u0187\u0002\u0000\u0007\u0000\u0002\u0001\u0007\u0001\u0002"+
 		"\u0002\u0007\u0002\u0002\u0003\u0007\u0003\u0002\u0004\u0007\u0004\u0002"+
 		"\u0005\u0007\u0005\u0002\u0006\u0007\u0006\u0002\u0007\u0007\u0007\u0002"+
 		"\b\u0007\b\u0002\t\u0007\t\u0002\n\u0007\n\u0002\u000b\u0007\u000b\u0002"+

rewrite-hcl/src/main/java/org/openrewrite/hcl/internal/grammar/HCLLexer.tokens

@@ -49,6 +49,7 @@ HP_COMMENT=48
 HP_LINE_COMMENT=49
 HTemplateLiteral=50
 HTemplateLiteralChar=51
+H_NEWLINE=52
 'if'=3
 'in'=4
 'null'=6
@@ -81,3 +82,4 @@ HTemplateLiteralChar=51
 '%'=41
 '...'=42
 '~'=43
+'\n'=52