Make tidy aware of Rust multiline strings

As a result of tighter and more correct handling of character
literals, this now catches a few kinds of syntax involving lifetimes
that were previously missed, so those have been updated.

components/layout/generated_content.rs

@@ -137,7 +137,7 @@ impl<'a> InorderFlowTraversal for ResolveGeneratedContent<'a> {
 }
 
 /// The object that mutates the generated content fragments.
-struct ResolveGeneratedContentFragmentMutator<'a,'b:'a> {
+struct ResolveGeneratedContentFragmentMutator<'a, 'b: 'a> {
     /// The traversal.
     traversal: &'a mut ResolveGeneratedContent<'b>,
     /// The level we're at in the flow tree.
@@ -148,7 +148,7 @@ struct ResolveGeneratedContentFragmentMutator<'a,'b:'a> {
     incremented: bool,
 }
 
-impl<'a,'b> ResolveGeneratedContentFragmentMutator<'a,'b> {
+impl<'a, 'b> ResolveGeneratedContentFragmentMutator<'a, 'b> {
     fn mutate_fragment(&mut self, fragment: &mut Fragment) {
         // We only reset and/or increment counters once per flow. This avoids double-incrementing
         // counters on list items (once for the main fragment and once for the marker).

components/layout/persistent_list.rs

@@ -69,7 +69,7 @@ impl<T> Clone for PersistentList<T> where T: Send + Sync {
     }
 }
 
-pub struct PersistentListIterator<'a,T> where T: 'a + Send + Sync {
+pub struct PersistentListIterator<'a, T> where T: 'a + Send + Sync {
     entry: Option<&'a PersistentListEntry<T>>,
 }
 

components/script_plugins/unrooted_must_root.rs

@@ -163,7 +163,7 @@ impl<'a, 'tcx> LateLintPass<'a, 'tcx> for UnrootedPass {
     }
 }
 
-struct FnDefVisitor<'a, 'b: 'a, 'tcx: 'a+'b> {
+struct FnDefVisitor<'a, 'b: 'a, 'tcx: 'a + 'b> {
     cx: &'a LateContext<'b, 'tcx>,
     in_new_function: bool,
 }

python/tidy/servo_tidy/tidy.py

@@ -467,15 +467,6 @@ def check_rust(file_name, lines):
         prev_indent = indent
         indent = len(original_line) - len(line)
 
-        # Hack for components/selectors/build.rs
-        if multi_line_string:
-            if line.startswith('"#'):
-                multi_line_string = False
-            else:
-                continue
-        if line.endswith('r#"'):
-            multi_line_string = True
-
         is_attribute = re.search(r"#\[.*\]", line)
         is_comment = re.search(r"^//|^/\*|^\*", line)
 
@@ -494,6 +485,14 @@ def check_rust(file_name, lines):
             line = merged_lines + line
             merged_lines = ''
 
+        if multi_line_string:
+            line, count = re.subn(
+                r'^(\\.|[^"\\])*?"', '', line, count=1)
+            if count == 1:
+                multi_line_string = False
+            else:
+                continue
+
         # Ignore attributes, comments, and imports
         # Keep track of whitespace to enable checking for a merged import block
         if import_block:
@@ -504,9 +503,17 @@ def check_rust(file_name, lines):
                     import_block = False
 
         # get rid of strings and chars because cases like regex expression, keep attributes
-        if not is_attribute:
+        if not is_attribute and not is_comment:
             line = re.sub(r'"(\\.|[^\\"])*?"', '""', line)
-            line = re.sub(r"'(\\.|[^\\'])*?'", "''", line)
+            line = re.sub(
+                r"'(\\.|[^\\']|(\\x[0-9a-fA-F]{2})|(\\u{[0-9a-fA-F]{1,6}}))'",
+                "''", line)
+            # If, after parsing all single-line strings, we still have
+            # an odd number of double quotes, this line starts a
+            # multiline string
+            if line.count('"') % 2 == 1:
+                line = re.sub(r'"(\\.|[^\\"])*?$', '""', line)
+                multi_line_string = True
 
         # get rid of comments
         line = re.sub('//.*?$|/\*.*?$|^\*.*?$', '//', line)

python/tidy/servo_tidy_tests/multiline_string.rs

@@ -0,0 +1,28 @@
+/* This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this
+ * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
+
+
+// This puts a "multi-line string
+// inside of a comment" and then subsequently has a hyphenated-phrase
+
+
+const FOO: &'static str = "Do not confuse 'apostrophes',
+    They can be 'lifetimes' or 'characters'";
+
+
+fn main() {
+    assert!(foo("test
+                 foo-bar"));
+
+    assert!(foo("test
+                 test2 \"
+                 foo-bar"));
+
+    assert!(foo("test
+                 test2 \
+                 foo-bar"));
+
+    println!("This is a multiline string with a URL, which kinda, \
+             sorta looks like a comment https://github.com/servo/servo/");
+}

python/tidy/servo_tidy_tests/test_tidy.py

@@ -266,6 +266,11 @@ class CheckTidiness(unittest.TestCase):
         lst = list(file_list)
         self.assertEqual([os.path.join(base_path, 'whee', 'test.rs')], lst)
 
+    def test_multiline_string(self):
+        errors = tidy.collect_errors_for_files(iterFile('multiline_string.rs'), [], [tidy.check_rust], print_text=True)
+        self.assertNoMoreErrors(errors)
+
+
 def do_tests():
     suite = unittest.TestLoader().loadTestsFromTestCase(CheckTidiness)
     return 0 if unittest.TextTestRunner(verbosity=2).run(suite).wasSuccessful() else 1