Improve string escaping for HCL interpolation (#224)

Revised `_escape_interpolated_str` to handle heredoc strings and added more robust escaping for special characters. Added test HCL file to validate various unicode and complex string scenarios, ensuring accurate handling and output.

Author: eranor

hcl2/reconstructor.py

@@ -353,11 +353,21 @@ def _name_to_identifier(name: str) -> Tree:
 
     @staticmethod
     def _escape_interpolated_str(interp_s: str) -> str:
-        # begin by doing basic JSON string escaping, to add backslashes
-        interp_s = json.dumps(interp_s)
-
+        if interp_s.strip().startswith('<<-') or interp_s.strip().startswith('<<'):
+            # For heredoc strings, preserve their format exactly
+            return reverse_quotes_within_interpolation(interp_s)
+        # Escape backslashes first (very important to do this first)
+        escaped = interp_s.replace('\\', '\\\\')
+        # Escape quotes
+        escaped = escaped.replace('"', '\\"')
+        # Escape control characters
+        escaped = escaped.replace('\n', '\\n')
+        escaped = escaped.replace('\r', '\\r')
+        escaped = escaped.replace('\t', '\\t')
+        escaped = escaped.replace('\b', '\\b')
+        escaped = escaped.replace('\f', '\\f')
         # find each interpolation within the string and remove the backslashes
-        interp_s = reverse_quotes_within_interpolation(interp_s)
+        interp_s = reverse_quotes_within_interpolation(f'"{escaped}"')
         return interp_s
 
     @staticmethod

test/helpers/terraform-config-json/unicode_strings.json

@@ -0,0 +1,20 @@
+{
+  "locals": [
+    {
+      "basic_unicode": "Hello, 世界! こんにちは Привет नमस्ते",
+      "unicode_escapes": "© ♥ ♪ ☠ ☺",
+      "emoji_string": "🚀 🌍 🔥 🎉",
+      "rtl_text": "English and العربية text mixed",
+      "complex_unicode": "Python (파이썬) es 很棒的! ♥ αβγδ",
+      "ascii": "ASCII: abc123",
+      "emoji": "Emoji: 🚀🌍🔥🎉",
+      "math": "Math: ∑∫√∞≠≤≥",
+      "currency": "Currency: £€¥₹₽₩",
+      "arrows": "Arrows: ←↑→↓↔↕",
+      "cjk": "CJK: 你好世界안녕하세요こんにちは",
+      "cyrillic": "Cyrillic: Привет мир",
+      "special": "Special: ©®™§¶†‡",
+      "mixed_content": "Line with interpolation: ${var.name}\nLine with emoji: 👨‍👩‍👧‍👦\nLine with quotes: \"quoted text\"\nLine with backslash: \\escaped"
+    }
+  ]
+}

test/helpers/terraform-config/unicode_strings.tf

@@ -0,0 +1,21 @@
+locals {
+  basic_unicode = "Hello, 世界! こんにちは Привет नमस्ते"
+  unicode_escapes = "© ♥ ♪ ☠ ☺"
+  emoji_string = "🚀 🌍 🔥 🎉"
+  rtl_text = "English and العربية text mixed"
+  complex_unicode = "Python (파이썬) es 很棒的! ♥ αβγδ"
+  ascii = "ASCII: abc123"
+  emoji = "Emoji: 🚀🌍🔥🎉"
+  math = "Math: ∑∫√∞≠≤≥"
+  currency = "Currency: £€¥₹₽₩"
+  arrows = "Arrows: ←↑→↓↔↕"
+  cjk = "CJK: 你好世界안녕하세요こんにちは"
+  cyrillic = "Cyrillic: Привет мир"
+  special = "Special: ©®™§¶†‡"
+  mixed_content = <<-EOT
+    Line with interpolation: ${var.name}
+    Line with emoji: 👨‍👩‍👧‍👦
+    Line with quotes: "quoted text"
+    Line with backslash: \escaped
+  EOT
+}