Merge pull request #30 from eddyb/trailing-dots

v0: also support preserving extra suffixes found after mangled symbol.

Author: alexcrichton

src/legacy.rs

@@ -46,64 +46,57 @@ pub struct Demangle<'a> {
 // Note that this demangler isn't quite as fancy as it could be. We have lots
 // of other information in our symbols like hashes, version, type information,
 // etc. Additionally, this doesn't handle glue symbols at all.
-pub fn demangle(s: &str) -> Result<Demangle, ()> {
+pub fn demangle(s: &str) -> Result<(Demangle, &str), ()> {
     // First validate the symbol. If it doesn't look like anything we're
     // expecting, we just print it literally. Note that we must handle non-Rust
     // symbols because we could have any function in the backtrace.
-    let inner;
-    if s.len() > 4 && s.starts_with("_ZN") && s.ends_with('E') {
-        inner = &s[3..s.len() - 1];
-    } else if s.len() > 3 && s.starts_with("ZN") && s.ends_with('E') {
+    let inner = if s.starts_with("_ZN") {
+        &s[3..]
+    } else if s.starts_with("ZN") {
         // On Windows, dbghelp strips leading underscores, so we accept "ZN...E"
         // form too.
-        inner = &s[2..s.len() - 1];
-    } else if s.len() > 5 && s.starts_with("__ZN") && s.ends_with('E') {
+        &s[2..]
+    } else if s.starts_with("__ZN") {
         // On OSX, symbols are prefixed with an extra _
-        inner = &s[4..s.len() - 1];
+        &s[4..]
     } else {
         return Err(());
-    }
+    };
 
     // only work with ascii text
     if inner.bytes().any(|c| c & 0x80 != 0) {
         return Err(());
     }
 
     let mut elements = 0;
-    let mut chars = inner.chars().peekable();
-    loop {
-        let mut i = 0usize;
-        while let Some(&c) = chars.peek() {
-            if !c.is_digit(10) {
-                break
-            }
-            chars.next();
-            let next = i.checked_mul(10)
-                .and_then(|i| i.checked_add(c as usize - '0' as usize));
-            i = match next {
-                Some(i) => i,
-                None => {
-                    return Err(());
-                }
-            };
+    let mut chars = inner.chars();
+    let mut c = try!(chars.next().ok_or(()));
+    while c != 'E' {
+        // Decode an identifier element's length.
+        if !c.is_digit(10) {
+            return Err(());
+        }
+        let mut len = 0usize;
+        while let Some(d) = c.to_digit(10) {
+            len = try!(len.checked_mul(10)
+                .and_then(|len| len.checked_add(d as usize))
+                .ok_or(()));
+            c = try!(chars.next().ok_or(()));
         }
 
-        if i == 0 {
-            if !chars.next().is_none() {
-                return Err(());
-            }
-            break;
-        } else if chars.by_ref().take(i).count() != i {
-            return Err(());
-        } else {
-            elements += 1;
+        // `c` already contains the first character of this identifier, skip it and
+        // all the other characters of this identifier, to reach the next element.
+        for _ in 0..len {
+            c = try!(chars.next().ok_or(()));
         }
+
+        elements += 1;
     }
 
-    Ok(Demangle {
+    Ok((Demangle {
         inner: inner,
         elements: elements,
-    })
+    }, chars.as_str()))
 }
 
 // Rust hashes are hex digits with an `h` prepended.

src/lib.rs

@@ -81,25 +81,33 @@ pub fn demangle(mut s: &str) -> Demangle {
         }
     }
 
-    // Output like LLVM IR adds extra period-delimited words. See if
-    // we are in that case and save the trailing words if so.
     let mut suffix = "";
-    if let Some(i) = s.rfind("E.") {
-        let (head, tail) = s.split_at(i + 1); // After the E, before the period
-
-        if is_symbol_like(tail) {
-            s = head;
-            suffix = tail;
+    let mut style = match legacy::demangle(s) {
+        Ok((d, s)) => {
+            suffix = s;
+            Some(DemangleStyle::Legacy(d))
         }
-    }
-
-    let style = match legacy::demangle(s) {
-        Ok(d) => Some(DemangleStyle::Legacy(d)),
         Err(()) => match v0::demangle(s) {
-            Ok(d) => Some(DemangleStyle::V0(d)),
+            Ok((d, s)) => {
+                suffix = s;
+                Some(DemangleStyle::V0(d))
+            }
             Err(v0::Invalid) => None,
         },
     };
+
+    // Output like LLVM IR adds extra period-delimited words. See if
+    // we are in that case and save the trailing words if so.
+    if !suffix.is_empty() {
+        if suffix.starts_with(".") && is_symbol_like(suffix) {
+            // Keep the suffix.
+        } else {
+            // Reset the suffix and invalidate the demangling.
+            suffix = "";
+            style = None;
+        }
+    }
+
     Demangle {
         style: style,
         original: s,

src/v0.rs

@@ -12,7 +12,7 @@ pub struct Demangle<'a> {
 /// This function will take a **mangled** symbol and return a value. When printed,
 /// the de-mangled version will be written. If the symbol does not look like
 /// a mangled symbol, the original value will be written instead.
-pub fn demangle(s: &str) -> Result<Demangle, Invalid> {
+pub fn demangle(s: &str) -> Result<(Demangle, &str), Invalid> {
     // First validate the symbol. If it doesn't look like anything we're
     // expecting, we just print it literally. Note that we must handle non-Rust
     // symbols because we could have any function in the backtrace.
@@ -47,17 +47,18 @@ pub fn demangle(s: &str) -> Result<Demangle, Invalid> {
         next: 0,
     };
     try!(parser.skip_path());
-    if parser.next < parser.sym.len() {
-        // Instantiating crate.
-        try!(parser.skip_path());
-    }
-    if parser.next != parser.sym.len() {
-        return Err(Invalid);
+
+    // Instantiating crate (paths always start with uppercase characters).
+    match parser.sym.as_bytes().get(parser.next) {
+        Some(&b'A'...b'Z') => {
+            try!(parser.skip_path());
+        }
+        _ => {}
     }
 
-    Ok(Demangle {
+    Ok((Demangle {
         inner: inner,
-    })
+    }, &parser.sym[parser.next..]))
 }
 
 impl<'s> Display for Demangle<'s> {
@@ -1064,4 +1065,20 @@ mod tests {
                ((((_, _), (_, _)), ((_, _), (_, _))), (((_, _), (_, _)), ((_, _), (_, _))))))"
         );
     }
+
+    #[test]
+    fn demangle_thinlto() {
+        t_nohash!("_RC3foo.llvm.9D1C9369", "foo");
+        t_nohash!("_RC3foo.llvm.9D1C9369@@16", "foo");
+        t_nohash!("_RNvC9backtrace3foo.llvm.A5310EB9", "backtrace::foo");
+    }
+
+    #[test]
+    fn demangle_extra_suffix() {
+        // From alexcrichton/rustc-demangle#27:
+        t_nohash!(
+            "_RNvNtNtNtNtCs92dm3009vxr_4rand4rngs7adapter9reseeding4fork23FORK_HANDLER_REGISTERED.0.0",
+            "rand::rngs::adapter::reseeding::fork::FORK_HANDLER_REGISTERED.0.0"
+        );
+    }
 }