From fbb243a5002cab03d2c848d2d5f94520e3e52763 Mon Sep 17 00:00:00 2001
From: Simon Brown <simon@structurizr.com>
Date: Tue, 19 Sep 2023 11:40:01 +0100
Subject: [PATCH] Adds a way to set the character encoding used by the DSL
 parser (see https://github.com/structurizr/dsl/issues/338).

---
 docs/changelog.md                               |  1 +
 .../structurizr/dsl/StructurizrDslParser.java   | 17 ++++++++++++++++-
 src/test/dsl/iso-8859.dsl                       |  5 +++++
 src/test/java/com/structurizr/dsl/DslTests.java |  9 +++++++++
 4 files changed, 31 insertions(+), 1 deletion(-)
 create mode 100644 src/test/dsl/iso-8859.dsl

diff --git a/docs/changelog.md b/docs/changelog.md
index 445cb7a..534ae89 100644
--- a/docs/changelog.md
+++ b/docs/changelog.md
@@ -5,6 +5,7 @@
 - DSL identifiers (if present) will now be loaded when extending a JSON workspace (see https://github.com/structurizr/dsl/discussions/328).
 - Adds a `context` variable to inline/external scripts (see https://github.com/structurizr/dsl/issues/332).
 - Fixes https://github.com/structurizr/dsl/issues/324 (Groups with no curly braces breaks diagrams).
+- Adds a way to set the character encoding used by the DSL parser (see https://github.com/structurizr/dsl/issues/338).
 
 ## 1.32.0 (28th July 2023)
 
diff --git a/src/main/java/com/structurizr/dsl/StructurizrDslParser.java b/src/main/java/com/structurizr/dsl/StructurizrDslParser.java
index 98ebe7b..ea678c9 100644
--- a/src/main/java/com/structurizr/dsl/StructurizrDslParser.java
+++ b/src/main/java/com/structurizr/dsl/StructurizrDslParser.java
@@ -7,6 +7,7 @@
 
 import java.io.File;
 import java.io.IOException;
+import java.nio.charset.Charset;
 import java.nio.charset.StandardCharsets;
 import java.nio.file.Files;
 import java.util.*;
@@ -32,6 +33,7 @@ public final class StructurizrDslParser extends StructurizrDslTokens {
 
     private static final String STRUCTURIZR_DSL_IDENTIFIER_PROPERTY_NAME = "structurizr.dsl.identifier";
 
+    private Charset characterEncoding = StandardCharsets.UTF_8;
     private IdentifierScope identifierScope = IdentifierScope.Flat;
     private Stack<DslContext> contextStack;
     private Set<String> parsedTokens = new HashSet<>();
@@ -53,6 +55,19 @@ public StructurizrDslParser() {
         constants = new HashMap<>();
     }
 
+    /**
+     * Provides a way to change the character encoding used by the DSL parser.
+     *
+     * @param characterEncoding     a Charset instance
+     */
+    public void setCharacterEncoding(Charset characterEncoding) {
+        if (characterEncoding == null) {
+            throw new IllegalArgumentException("A character encoding must be specified");
+        }
+
+        this.characterEncoding = characterEncoding;
+    }
+
     IdentifierScope getIdentifierScope() {
         return identifierScope;
     }
@@ -125,7 +140,7 @@ public void parse(File path) throws StructurizrDslParserException {
         List<File> files = FileUtils.findFiles(path);
         try {
             for (File file : files) {
-                parse(Files.readAllLines(file.toPath(), StandardCharsets.UTF_8), file);
+                parse(Files.readAllLines(file.toPath(), characterEncoding), file);
             }
         } catch (IOException e) {
             throw new StructurizrDslParserException(e.getMessage());
diff --git a/src/test/dsl/iso-8859.dsl b/src/test/dsl/iso-8859.dsl
new file mode 100644
index 0000000..caad37c
--- /dev/null
+++ b/src/test/dsl/iso-8859.dsl
@@ -0,0 +1,5 @@
+workspace {
+    model {
+        softwareSystem "Namé"
+    }
+}
\ No newline at end of file
diff --git a/src/test/java/com/structurizr/dsl/DslTests.java b/src/test/java/com/structurizr/dsl/DslTests.java
index 27bc9a2..321db56 100644
--- a/src/test/java/com/structurizr/dsl/DslTests.java
+++ b/src/test/java/com/structurizr/dsl/DslTests.java
@@ -8,6 +8,7 @@
 import org.junit.jupiter.params.provider.ValueSource;
 
 import java.io.File;
+import java.nio.charset.StandardCharsets;
 import java.nio.file.Files;
 import java.util.ArrayList;
 import java.util.Base64;
@@ -1029,4 +1030,12 @@ void test_GroupWithoutBrace() throws Exception {
         }
     }
 
+    @Test
+    void test_ISO8859Encoding() throws Exception {
+        StructurizrDslParser parser = new StructurizrDslParser();
+        parser.setCharacterEncoding(StandardCharsets.ISO_8859_1);
+        parser.parse(new File("src/test/dsl/iso-8859.dsl"));
+        assertNotNull(parser.getWorkspace().getModel().getSoftwareSystemWithName("Namé"));
+    }
+
 }
\ No newline at end of file