Auto merge of rust-lang#7873 - xFrednet:7869-string-index-ice, r=Manishearth

Update `str` utils to prevent ICEs and FNs

This PR reworks some string handling for lints regarding enum naming. I hope the refactoring will prevent future ICEs and help with new bug free implementations.

It might be better to review this PR by going through the commits, as `clippy_utils::camel_case` was renamed to `clippy_utils::str_utils` and then changed further. GH sadly doesn't really make the changes that obvious 🙃

Not too much more to say. Have a nice day 🌞

---

Fixes: rust-lang/rust-clippy#7869

changelog: ICE Fix: [`enum_variant_names`] rust-lang#7869

Author: bors

clippy_lints/src/enum_variants.rs

@@ -1,8 +1,8 @@
 //! lint on enum variants that are prefixed or suffixed by the same characters
 
-use clippy_utils::camel_case;
 use clippy_utils::diagnostics::{span_lint, span_lint_and_help};
 use clippy_utils::source::is_present_in_source;
+use clippy_utils::str_utils::{self, count_match_end, count_match_start};
 use rustc_hir::{EnumDef, Item, ItemKind};
 use rustc_lint::{LateContext, LateLintPass};
 use rustc_session::{declare_tool_lint, impl_lint_pass};
@@ -117,26 +117,6 @@ impl_lint_pass!(EnumVariantNames => [
     MODULE_INCEPTION
 ]);
 
-/// Returns the number of chars that match from the start
-#[must_use]
-fn partial_match(pre: &str, name: &str) -> usize {
-    let mut name_iter = name.chars();
-    let _ = name_iter.next_back(); // make sure the name is never fully matched
-    pre.chars().zip(name_iter).take_while(|&(l, r)| l == r).count()
-}
-
-/// Returns the number of chars that match from the end
-#[must_use]
-fn partial_rmatch(post: &str, name: &str) -> usize {
-    let mut name_iter = name.chars();
-    let _ = name_iter.next(); // make sure the name is never fully matched
-    post.chars()
-        .rev()
-        .zip(name_iter.rev())
-        .take_while(|&(l, r)| l == r)
-        .count()
-}
-
 fn check_variant(
     cx: &LateContext<'_>,
     threshold: u64,
@@ -150,7 +130,7 @@ fn check_variant(
     }
     for var in def.variants {
         let name = var.ident.name.as_str();
-        if partial_match(item_name, &name) == item_name_chars
+        if count_match_start(item_name, &name).char_count == item_name_chars
             && name.chars().nth(item_name_chars).map_or(false, |c| !c.is_lowercase())
             && name.chars().nth(item_name_chars + 1).map_or(false, |c| !c.is_numeric())
         {
@@ -161,7 +141,7 @@ fn check_variant(
                 "variant name starts with the enum's name",
             );
         }
-        if partial_rmatch(item_name, &name) == item_name_chars {
+        if count_match_end(item_name, &name).char_count == item_name_chars {
             span_lint(
                 cx,
                 ENUM_VARIANT_NAMES,
@@ -171,33 +151,33 @@ fn check_variant(
         }
     }
     let first = &def.variants[0].ident.name.as_str();
-    let mut pre = &first[..camel_case::until(&*first)];
-    let mut post = &first[camel_case::from(&*first)..];
+    let mut pre = &first[..str_utils::camel_case_until(&*first).byte_index];
+    let mut post = &first[str_utils::camel_case_start(&*first).byte_index..];
     for var in def.variants {
         let name = var.ident.name.as_str();
 
-        let pre_match = partial_match(pre, &name);
+        let pre_match = count_match_start(pre, &name).byte_count;
         pre = &pre[..pre_match];
-        let pre_camel = camel_case::until(pre);
+        let pre_camel = str_utils::camel_case_until(pre).byte_index;
         pre = &pre[..pre_camel];
         while let Some((next, last)) = name[pre.len()..].chars().zip(pre.chars().rev()).next() {
             if next.is_numeric() {
                 return;
             }
             if next.is_lowercase() {
                 let last = pre.len() - last.len_utf8();
-                let last_camel = camel_case::until(&pre[..last]);
-                pre = &pre[..last_camel];
+                let last_camel = str_utils::camel_case_until(&pre[..last]);
+                pre = &pre[..last_camel.byte_index];
             } else {
                 break;
             }
         }
 
-        let post_match = partial_rmatch(post, &name);
-        let post_end = post.len() - post_match;
+        let post_match = count_match_end(post, &name);
+        let post_end = post.len() - post_match.byte_count;
         post = &post[post_end..];
-        let post_camel = camel_case::from(post);
-        post = &post[post_camel..];
+        let post_camel = str_utils::camel_case_start(post);
+        post = &post[post_camel.byte_index..];
     }
     let (what, value) = match (pre.is_empty(), post.is_empty()) {
         (true, true) => return,
@@ -266,14 +246,16 @@ impl LateLintPass<'_> for EnumVariantNames {
                             );
                         }
                     }
-                    if item.vis.node.is_pub() {
-                        let matching = partial_match(mod_camel, &item_camel);
-                        let rmatching = partial_rmatch(mod_camel, &item_camel);
+                    // The `module_name_repetitions` lint should only trigger if the item has the module in its
+                    // name. Having the same name is accepted.
+                    if item.vis.node.is_pub() && item_camel.len() > mod_camel.len() {
+                        let matching = count_match_start(mod_camel, &item_camel);
+                        let rmatching = count_match_end(mod_camel, &item_camel);
                         let nchars = mod_camel.chars().count();
 
                         let is_word_beginning = |c: char| c == '_' || c.is_uppercase() || c.is_numeric();
 
-                        if matching == nchars {
+                        if matching.char_count == nchars {
                             match item_camel.chars().nth(nchars) {
                                 Some(c) if is_word_beginning(c) => span_lint(
                                     cx,
@@ -284,7 +266,7 @@ impl LateLintPass<'_> for EnumVariantNames {
                                 _ => (),
                             }
                         }
-                        if rmatching == nchars {
+                        if rmatching.char_count == nchars {
                             span_lint(
                                 cx,
                                 MODULE_NAME_REPETITIONS,

clippy_utils/src/camel_case.rs

@@ -37,7 +37,6 @@ pub mod sym_helper;
 #[allow(clippy::module_name_repetitions)]
 pub mod ast_utils;
 pub mod attrs;
-pub mod camel_case;
 pub mod comparisons;
 pub mod consts;
 pub mod diagnostics;
@@ -50,6 +49,7 @@ pub mod paths;
 pub mod ptr;
 pub mod qualify_min_const_fn;
 pub mod source;
+pub mod str_utils;
 pub mod sugg;
 pub mod ty;
 pub mod usage;