From 5de47e3df316c0662115ef9890ea5621bbdd2063 Mon Sep 17 00:00:00 2001
From: Agnik47 <140933190+Agnik47@users.noreply.github.com>
Date: Sat, 3 Oct 2026 13:18:15 +0530
Subject: [PATCH] fix(extract): end a Svelte script block's trailing comment
 and report read errors

Review follow-ups on #3928:

- A `//` comment closing one `<script>` block swallowed a second block on
  the same line, since masking leaves them on one line. Terminate each
  block with U+2028 then `;`: the line terminator ends the comment and
  the `;` ends a statement that had none. Line numbers and byte offsets
  are unchanged.
- An unreadable file returns the `error` key again instead of an empty
  result, matching the behavior before the masking change.
---
 graphify/extract.py             | 15 +++++++++------
 tests/test_svelte_extraction.py | 32 ++++++++++++++++++++++++++++++++
 2 files changed, 41 insertions(+), 6 deletions(-)

diff --git a/graphify/extract.py b/graphify/extract.py
index 8bb0bfcee..24aa1a11e 100644
--- a/graphify/extract.py
+++ b/graphify/extract.py
@@ -2387,10 +2387,13 @@ def _svelte_mask_non_script(src: str) -> tuple[str, str | None]:
             continue
         start, end = m.start(2), m.end(2)
         chars[start:end] = src[start:end]
-        # Terminate the region in place of the following `<`, so two scripts on
-        # one line don't run together into a single statement.
-        if end < len(chars) and chars[end] == " ":
-            chars[end] = ";"
+        # Terminate the region in place of the closing `</script`, so two scripts
+        # on one line don't run together. U+2028 is a JS line terminator, so it
+        # ends a trailing `//` comment that would otherwise swallow the next
+        # block; the `;` after it then ends the statement (inside the comment it
+        # would be inert). U+2028 is not a newline, so line numbers hold, and its
+        # 3 UTF-8 bytes replace 3 ASCII ones, so byte offsets do too.
+        chars[end:end + 4] = ["\u2028", ";", "", ""]
         if lang is None:
             lang_m = _SCRIPT_LANG_RE.search(m.group(1))
             if lang_m:
@@ -2421,8 +2424,8 @@ def extract_svelte(path: Path) -> dict:
     """
     try:
         src = path.read_text(encoding="utf-8", errors="replace")
-    except OSError:
-        return {"nodes": [], "edges": []}
+    except OSError as e:
+        return {"nodes": [], "edges": [], "error": str(e)}
 
     masked, lang = _svelte_mask_non_script(src)
     config = _JS_CONFIG if lang in ("js", "jsx") else _TS_CONFIG
diff --git a/tests/test_svelte_extraction.py b/tests/test_svelte_extraction.py
index bd6595edf..68d8b4b34 100644
--- a/tests/test_svelte_extraction.py
+++ b/tests/test_svelte_extraction.py
@@ -164,6 +164,38 @@ def test_extract_svelte_scripts_on_one_line_do_not_merge(tmp_path):
     assert "b()" in _labels(result)
 
 
+def test_extract_svelte_line_comment_does_not_swallow_next_script(tmp_path):
+    """A `//` comment ending one block must not hide a block on the same line."""
+    component = _write(
+        tmp_path / "src/Commented.svelte",
+        "<script module>// shared</script><script>function b() {}</script>\n"
+        "<script>function c() {}</script>\n",
+    )
+    result = extract_svelte(component)
+    assert result.get("parse_errors") is None
+    labels = _labels(result)
+    assert labels.get("b()") == "L1"
+    assert labels.get("c()") == "L2"
+
+
+def test_extract_svelte_comment_after_unterminated_statement(tmp_path):
+    """A statement with no `;` before the `//` comment still ends at the block."""
+    component = _write(
+        tmp_path / "src/Unterminated.svelte",
+        "<script module>const a = 1 // shared</script><script>function b() {}</script>\n",
+    )
+    result = extract_svelte(component)
+    assert result.get("parse_errors") is None
+    assert _labels(result).get("b()") == "L1"
+
+
+def test_extract_svelte_unreadable_file_reports_error(tmp_path):
+    """A file that cannot be read is reported, not returned as an empty result."""
+    result = extract_svelte(tmp_path / "missing" / "Gone.svelte")
+    assert result["nodes"] == [] and result["edges"] == []
+    assert "error" in result
+
+
 def test_extract_svelte_markup_only_component_does_not_crash(tmp_path):
     """A `.svelte` file need not have a `<script>` block at all."""
     component = _write(tmp_path / "src/Plain.svelte", "<h1>no script here</h1>\n")