rules: early detection of invalid encoding

xkbcommon · Oct 30, 2023 · 28d8e81 · 28d8e81
1 parent 43b13fc
commit 28d8e81
Show file tree

Hide file tree

Showing 5 changed files with 62 additions and 2 deletions.
diff --git a/src/xkbcomp/rules.c b/src/xkbcomp/rules.c
@@ -1101,13 +1101,37 @@ read_rules_file(struct xkb_context *ctx,
 
     /* Skip UTF-8 encoded BOM (U+FEFF) */
     /* See: https://www.unicode.org/faq/utf_bom.html#bom5 */
-    scanner_str(&scanner, "\xef\xbb\xbf", 3);
+    if (scanner_str(&scanner, "\xef\xbb\xbf", 3) || scanner.len < 2) {
+        goto initial;
+    }
 
-    ret = matcher_match(matcher, &scanner, include_depth, string, size, path);
+    /* Early detection of wrong file encoding, e.g. UTF-16 or UTF-32 */
+    if (scanner.s[0] == '\0' || scanner.s[1] == '\0') {
+        if (scanner.s[0] != '\0')
+            scanner.token_column++;
+        scanner_err(&scanner, "unexpected NULL character.");
+        goto encoding_error;
+    }
+    if (!is_ascii(scanner.s[0])) {
+        scanner_err(&scanner, "unexpected non-ASCII character.");
+        goto encoding_error;
+    }
 
+initial:
+    ret = matcher_match(matcher, &scanner, include_depth, string, size, path);
     unmap_file(string, size);
+
 out:
     return ret;
+
+encoding_error:
+    scanner_err(&scanner,
+        "This could be a file encoding issue. "
+        "Supported encodings must be backward compatible with ASCII.");
+    scanner_err(&scanner,
+        "E.g. ISO/CEI 8859 and UTF-8 are supported "
+        "but UTF-16, UTF-32 and CP1026 are not.");
+    return false;
 }
 
 bool

diff --git a/test/data/rules/utf-16be_with_bom b/test/data/rules/utf-16be_with_bom
diff --git a/test/data/rules/utf-16le_with_bom b/test/data/rules/utf-16le_with_bom
diff --git a/test/data/rules/utf-32be b/test/data/rules/utf-32be
diff --git a/test/rules-file.c b/test/rules-file.c
@@ -106,6 +106,42 @@ main(int argc, char *argv[])
     };
     assert(test_rules(ctx, &test_utf_8_with_bom));
 
+    struct test_data test_utf_16le_with_bom = {
+        .rules = "utf-16le_with_bom",
+
+        .model = "my_model", .layout = "my_layout", .variant = "my_variant",
+        .options = "my_option",
+
+        .keycodes = "my_keycodes", .types = "my_types",
+        .compat = "my_compat|some:compat",
+        .symbols = "my_symbols+extra_variant",
+    };
+    assert(!test_rules(ctx, &test_utf_16le_with_bom));
+
+    struct test_data test_utf_16be_with_bom = {
+        .rules = "utf-16be_with_bom",
+
+        .model = "my_model", .layout = "my_layout", .variant = "my_variant",
+        .options = "my_option",
+
+        .keycodes = "my_keycodes", .types = "my_types",
+        .compat = "my_compat|some:compat",
+        .symbols = "my_symbols+extra_variant",
+    };
+    assert(!test_rules(ctx, &test_utf_16be_with_bom));
+
+    struct test_data test_utf_32be = {
+        .rules = "utf-32be",
+
+        .model = "my_model", .layout = "my_layout", .variant = "my_variant",
+        .options = "my_option",
+
+        .keycodes = "my_keycodes", .types = "my_types",
+        .compat = "my_compat|some:compat",
+        .symbols = "my_symbols+extra_variant",
+    };
+    assert(!test_rules(ctx, &test_utf_32be));
+
     struct test_data test1 = {
         .rules = "simple",