Rollup merge of #79809 - Eric-Arellano:split-once, r=matklad

Dogfood `str_split_once()`

Part of #74773.

Beyond increased clarity, this fixes some instances of a common confusion with how `splitn(2)` behaves: the first element will always be `Some()`, regardless of the delimiter, and even if the value is empty.

Given this code:

```rust
fn main() {
    let val = "...";
    let mut iter = val.splitn(2, '=');
    println!("Input: {:?}, first: {:?}, second: {:?}", val, iter.next(), iter.next());
}
```

We get:

```
Input: "no_delimiter", first: Some("no_delimiter"), second: None
Input: "k=v", first: Some("k"), second: Some("v")
Input: "=", first: Some(""), second: Some("")
```

Using `str_split_once()` makes more clear what happens when the delimiter is not found.

Author: tmandry

compiler/rustc_mir/src/lib.rs

@@ -28,6 +28,7 @@ Rust MIR: a lowered representation of Rust.
 #![feature(or_patterns)]
 #![feature(once_cell)]
 #![feature(control_flow_enum)]
+#![feature(str_split_once)]
 #![recursion_limit = "256"]
 
 #[macro_use]

compiler/rustc_mir/src/transform/coverage/debug.rs

@@ -148,40 +148,46 @@ impl DebugOptions {
 
         if let Ok(env_debug_options) = std::env::var(RUSTC_COVERAGE_DEBUG_OPTIONS) {
             for setting_str in env_debug_options.replace(" ", "").replace("-", "_").split(',') {
-                let mut setting = setting_str.splitn(2, '=');
-                match setting.next() {
-                    Some(option) if option == "allow_unused_expressions" => {
-                        allow_unused_expressions = bool_option_val(option, setting.next());
+                let (option, value) = match setting_str.split_once('=') {
+                    None => (setting_str, None),
+                    Some((k, v)) => (k, Some(v)),
+                };
+                match option {
+                    "allow_unused_expressions" => {
+                        allow_unused_expressions = bool_option_val(option, value);
                         debug!(
                             "{} env option `allow_unused_expressions` is set to {}",
                             RUSTC_COVERAGE_DEBUG_OPTIONS, allow_unused_expressions
                         );
                     }
-                    Some(option) if option == "counter_format" => {
-                        if let Some(strval) = setting.next() {
-                            counter_format = counter_format_option_val(strval);
-                            debug!(
-                                "{} env option `counter_format` is set to {:?}",
-                                RUSTC_COVERAGE_DEBUG_OPTIONS, counter_format
-                            );
-                        } else {
-                            bug!(
-                                "`{}` option in environment variable {} requires one or more \
-                                plus-separated choices (a non-empty subset of \
-                                `id+block+operation`)",
-                                option,
-                                RUSTC_COVERAGE_DEBUG_OPTIONS
-                            );
-                        }
+                    "counter_format" => {
+                        match value {
+                            None => {
+                                bug!(
+                                    "`{}` option in environment variable {} requires one or more \
+                                    plus-separated choices (a non-empty subset of \
+                                    `id+block+operation`)",
+                                    option,
+                                    RUSTC_COVERAGE_DEBUG_OPTIONS
+                                );
+                            }
+                            Some(val) => {
+                                counter_format = counter_format_option_val(val);
+                                debug!(
+                                    "{} env option `counter_format` is set to {:?}",
+                                    RUSTC_COVERAGE_DEBUG_OPTIONS, counter_format
+                                );
+                            }
+                        };
                     }
-                    Some("") => {}
-                    Some(invalid) => bug!(
-                        "Unsupported setting `{}` in environment variable {}",
-                        invalid,
-                        RUSTC_COVERAGE_DEBUG_OPTIONS
-                    ),
-                    None => {}
-                }
+                    _ => {
+                        bug!(
+                            "Unsupported setting `{}` in environment variable {}",
+                            option,
+                            RUSTC_COVERAGE_DEBUG_OPTIONS
+                        )
+                    }
+                };
             }
         }
 

compiler/rustc_session/src/config.rs

@@ -1296,8 +1296,10 @@ fn parse_output_types(
     if !debugging_opts.parse_only {
         for list in matches.opt_strs("emit") {
             for output_type in list.split(',') {
-                let mut parts = output_type.splitn(2, '=');
-                let shorthand = parts.next().unwrap();
+                let (shorthand, path) = match output_type.split_once('=') {
+                    None => (output_type, None),
+                    Some((shorthand, path)) => (shorthand, Some(PathBuf::from(path))),
+                };
                 let output_type = OutputType::from_shorthand(shorthand).unwrap_or_else(|| {
                     early_error(
                         error_format,
@@ -1308,7 +1310,6 @@ fn parse_output_types(
                         ),
                     )
                 });
-                let path = parts.next().map(PathBuf::from);
                 output_types.insert(output_type, path);
             }
         }
@@ -1452,11 +1453,10 @@ fn parse_opt_level(
     let max_c = matches
         .opt_strs_pos("C")
         .into_iter()
-        .flat_map(
-            |(i, s)| {
-                if let Some("opt-level") = s.splitn(2, '=').next() { Some(i) } else { None }
-            },
-        )
+        .flat_map(|(i, s)| {
+            // NB: This can match a string without `=`.
+            if let Some("opt-level") = s.splitn(2, '=').next() { Some(i) } else { None }
+        })
         .max();
     if max_o > max_c {
         OptLevel::Default
@@ -1491,11 +1491,10 @@ fn select_debuginfo(
     let max_c = matches
         .opt_strs_pos("C")
         .into_iter()
-        .flat_map(
-            |(i, s)| {
-                if let Some("debuginfo") = s.splitn(2, '=').next() { Some(i) } else { None }
-            },
-        )
+        .flat_map(|(i, s)| {
+            // NB: This can match a string without `=`.
+            if let Some("debuginfo") = s.splitn(2, '=').next() { Some(i) } else { None }
+        })
         .max();
     if max_g > max_c {
         DebugInfo::Full
@@ -1528,23 +1527,26 @@ fn parse_libs(
         .map(|s| {
             // Parse string of the form "[KIND=]lib[:new_name]",
             // where KIND is one of "dylib", "framework", "static".
-            let mut parts = s.splitn(2, '=');
-            let kind = parts.next().unwrap();
-            let (name, kind) = match (parts.next(), kind) {
-                (None, name) => (name, NativeLibKind::Unspecified),
-                (Some(name), "dylib") => (name, NativeLibKind::Dylib),
-                (Some(name), "framework") => (name, NativeLibKind::Framework),
-                (Some(name), "static") => (name, NativeLibKind::StaticBundle),
-                (Some(name), "static-nobundle") => (name, NativeLibKind::StaticNoBundle),
-                (_, s) => {
-                    early_error(
-                        error_format,
-                        &format!(
-                            "unknown library kind `{}`, expected \
-                             one of dylib, framework, or static",
-                            s
-                        ),
-                    );
+            let (name, kind) = match s.split_once('=') {
+                None => (s, NativeLibKind::Unspecified),
+                Some((kind, name)) => {
+                    let kind = match kind {
+                        "dylib" => NativeLibKind::Dylib,
+                        "framework" => NativeLibKind::Framework,
+                        "static" => NativeLibKind::StaticBundle,
+                        "static-nobundle" => NativeLibKind::StaticNoBundle,
+                        s => {
+                            early_error(
+                                error_format,
+                                &format!(
+                                    "unknown library kind `{}`, expected \
+                                     one of dylib, framework, or static",
+                                    s
+                                ),
+                            );
+                        }
+                    };
+                    (name.to_string(), kind)
                 }
             };
             if kind == NativeLibKind::StaticNoBundle
@@ -1556,10 +1558,11 @@ fn parse_libs(
                      accepted on the nightly compiler",
                 );
             }
-            let mut name_parts = name.splitn(2, ':');
-            let name = name_parts.next().unwrap();
-            let new_name = name_parts.next();
-            (name.to_owned(), new_name.map(|n| n.to_owned()), kind)
+            let (name, new_name) = match name.split_once(':') {
+                None => (name, None),
+                Some((name, new_name)) => (name.to_string(), Some(new_name.to_owned())),
+            };
+            (name, new_name, kind)
         })
         .collect()
 }
@@ -1580,20 +1583,13 @@ pub fn parse_externs(
     let is_unstable_enabled = debugging_opts.unstable_options;
     let mut externs: BTreeMap<String, ExternEntry> = BTreeMap::new();
     for arg in matches.opt_strs("extern") {
-        let mut parts = arg.splitn(2, '=');
-        let name = parts
-            .next()
-            .unwrap_or_else(|| early_error(error_format, "--extern value must not be empty"));
-        let path = parts.next().map(|s| s.to_string());
-
-        let mut name_parts = name.splitn(2, ':');
-        let first_part = name_parts.next();
-        let second_part = name_parts.next();
-        let (options, name) = match (first_part, second_part) {
-            (Some(opts), Some(name)) => (Some(opts), name),
-            (Some(name), None) => (None, name),
-            (None, None) => early_error(error_format, "--extern name must not be empty"),
-            _ => unreachable!(),
+        let (name, path) = match arg.split_once('=') {
+            None => (arg, None),
+            Some((name, path)) => (name.to_string(), Some(path.to_string())),
+        };
+        let (options, name) = match name.split_once(':') {
+            None => (None, name),
+            Some((opts, name)) => (Some(opts), name.to_string()),
         };
 
         let entry = externs.entry(name.to_owned());
@@ -1682,17 +1678,12 @@ fn parse_remap_path_prefix(
     matches
         .opt_strs("remap-path-prefix")
         .into_iter()
-        .map(|remap| {
-            let mut parts = remap.rsplitn(2, '='); // reverse iterator
-            let to = parts.next();
-            let from = parts.next();
-            match (from, to) {
-                (Some(from), Some(to)) => (PathBuf::from(from), PathBuf::from(to)),
-                _ => early_error(
-                    error_format,
-                    "--remap-path-prefix must contain '=' between FROM and TO",
-                ),
-            }
+        .map(|remap| match remap.rsplit_once('=') {
+            None => early_error(
+                error_format,
+                "--remap-path-prefix must contain '=' between FROM and TO",
+            ),
+            Some((from, to)) => (PathBuf::from(from), PathBuf::from(to)),
         })
         .collect()
 }

compiler/rustc_session/src/lib.rs

@@ -1,6 +1,7 @@
 #![feature(crate_visibility_modifier)]
 #![feature(once_cell)]
 #![feature(or_patterns)]
+#![feature(str_split_once)]
 
 #[macro_use]
 extern crate bitflags;

compiler/rustc_session/src/options.rs

@@ -179,9 +179,10 @@ macro_rules! options {
     {
         let mut op = $defaultfn();
         for option in matches.opt_strs($prefix) {
-            let mut iter = option.splitn(2, '=');
-            let key = iter.next().unwrap();
-            let value = iter.next();
+            let (key, value) = match option.split_once('=') {
+                None => (option, None),
+                Some((k, v)) => (k.to_string(), Some(v)),
+            };
             let option_to_lookup = key.replace("-", "_");
             let mut found = false;
             for &(candidate, setter, type_desc, _) in $stat {

compiler/rustc_target/src/lib.rs

@@ -15,6 +15,7 @@
 #![feature(never_type)]
 #![feature(associated_type_bounds)]
 #![feature(exhaustive_patterns)]
+#![feature(str_split_once)]
 
 #[macro_use]
 extern crate rustc_macros;

compiler/rustc_target/src/spec/apple_base.rs

@@ -54,10 +54,7 @@ fn macos_deployment_target() -> (u32, u32) {
     let deployment_target = env::var("MACOSX_DEPLOYMENT_TARGET").ok();
     let version = deployment_target
         .as_ref()
-        .and_then(|s| {
-            let mut i = s.splitn(2, '.');
-            i.next().and_then(|a| i.next().map(|b| (a, b)))
-        })
+        .and_then(|s| s.split_once('.'))
         .and_then(|(a, b)| a.parse::<u32>().and_then(|a| b.parse::<u32>().map(|b| (a, b))).ok());
 
     version.unwrap_or((10, 7))

library/std/src/lib.rs

@@ -314,6 +314,7 @@
 #![feature(stdsimd)]
 #![feature(stmt_expr_attributes)]
 #![feature(str_internals)]
+#![feature(str_split_once)]
 #![feature(test)]
 #![feature(thread_local)]
 #![feature(thread_local_internals)]

library/std/src/sys_common/net.rs

@@ -177,11 +177,8 @@ impl TryFrom<&str> for LookupHost {
         }
 
         // split the string by ':' and convert the second part to u16
-        let mut parts_iter = s.rsplitn(2, ':');
-        let port_str = try_opt!(parts_iter.next(), "invalid socket address");
-        let host = try_opt!(parts_iter.next(), "invalid socket address");
+        let (host, port_str) = try_opt!(s.rsplit_once(':'), "invalid socket address");
         let port: u16 = try_opt!(port_str.parse().ok(), "invalid port value");
-
         (host, port).try_into()
     }
 }

library/test/src/lib.rs

@@ -30,6 +30,7 @@
 #![feature(termination_trait_lib)]
 #![feature(test)]
 #![feature(total_cmp)]
+#![feature(str_split_once)]
 
 // Public reexports
 pub use self::bench::{black_box, Bencher};

library/test/src/time.rs

@@ -105,30 +105,24 @@ impl TimeThreshold {
     /// value.
     pub fn from_env_var(env_var_name: &str) -> Option<Self> {
         let durations_str = env::var(env_var_name).ok()?;
+        let (warn_str, critical_str) = durations_str.split_once(',').unwrap_or_else(|| {
+            panic!(
+                "Duration variable {} expected to have 2 numbers separated by comma, but got {}",
+                env_var_name, durations_str
+            )
+        });
 
-        // Split string into 2 substrings by comma and try to parse numbers.
-        let mut durations = durations_str.splitn(2, ',').map(|v| {
+        let parse_u64 = |v| {
             u64::from_str(v).unwrap_or_else(|_| {
                 panic!(
                     "Duration value in variable {} is expected to be a number, but got {}",
                     env_var_name, v
                 )
             })
-        });
-
-        // Callback to be called if the environment variable has unexpected structure.
-        let panic_on_incorrect_value = || {
-            panic!(
-                "Duration variable {} expected to have 2 numbers separated by comma, but got {}",
-                env_var_name, durations_str
-            );
         };
 
-        let (warn, critical) = (
-            durations.next().unwrap_or_else(panic_on_incorrect_value),
-            durations.next().unwrap_or_else(panic_on_incorrect_value),
-        );
-
+        let warn = parse_u64(warn_str);
+        let critical = parse_u64(critical_str);
         if warn > critical {
             panic!("Test execution warn time should be less or equal to the critical time");
         }

src/librustdoc/config.rs

@@ -397,12 +397,9 @@ impl Options {
             matches
                 .opt_strs("default-setting")
                 .iter()
-                .map(|s| {
-                    let mut kv = s.splitn(2, '=');
-                    // never panics because `splitn` always returns at least one element
-                    let k = kv.next().unwrap().to_string();
-                    let v = kv.next().unwrap_or("true").to_string();
-                    (k, v)
+                .map(|s| match s.split_once('=') {
+                    None => (s.clone(), "true".to_string()),
+                    Some((k, v)) => (k.to_string(), v.to_string()),
                 })
                 .collect(),
         ];
@@ -707,11 +704,9 @@ fn parse_extern_html_roots(
 ) -> Result<BTreeMap<String, String>, &'static str> {
     let mut externs = BTreeMap::new();
     for arg in &matches.opt_strs("extern-html-root-url") {
-        let mut parts = arg.splitn(2, '=');
-        let name = parts.next().ok_or("--extern-html-root-url must not be empty")?;
-        let url = parts.next().ok_or("--extern-html-root-url must be of the form name=url")?;
+        let (name, url) =
+            arg.split_once('=').ok_or("--extern-html-root-url must be of the form name=url")?;
         externs.insert(name.to_string(), url.to_string());
     }
-
     Ok(externs)
 }