Auto merge of #369 - rust-lang:ag-fix-353, r=BurntSushi

compiler: fix RegexSet bug

When compiling a RegexSet, it was possible for the jump locations to
become incorrect if the last regex in the set had a starting location
that didn't correspond to the beginning of its program. This can happen
in simple cases like when your set consists of the regexes `a` and `β`.
In particular, the program for `β` is:

    0: Bytes(\xB2) (goto 2)
    1: Bytes(\xCE) (goto 0)
    2: MATCH

Where the entry point is `1` instead of `0`. To fix this, we compile a
set of regexes similarly to how we compile `a|β`, where we handle the
holes produced by sub-expressions correctly.

Fixes #353

Author: bors

src/compile.rs

@@ -178,18 +178,19 @@ impl Compiler {
         }
         self.fill_to_next(dotstar_patch.hole);
 
+        let mut prev_hole = Hole::None;
         for (i, expr) in exprs[0..exprs.len() - 1].iter().enumerate() {
+            self.fill_to_next(prev_hole);
             let split = self.push_split_hole();
             let Patch { hole, entry } = try!(self.c_capture(0, expr));
             self.fill_to_next(hole);
             self.compiled.matches.push(self.insts.len());
             self.push_compiled(Inst::Match(i));
-
-            let next = self.insts.len();
-            self.fill_split(split, Some(entry), Some(next));
+            prev_hole = self.fill_split(split, Some(entry), None);
         }
         let i = exprs.len() - 1;
-        let Patch { hole, .. } = try!(self.c_capture(0, &exprs[i]));
+        let Patch { hole, entry } = try!(self.c_capture(0, &exprs[i]));
+        self.fill(prev_hole, entry);
         self.fill_to_next(hole);
         self.compiled.matches.push(self.insts.len());
         self.push_compiled(Inst::Match(i));

tests/set.rs

@@ -15,6 +15,7 @@ matset!(set14, &[r".*", "a"], "zzzzzz", 0);
 matset!(set15, &[r"\ba\b"], "hello a bye", 0);
 matset!(set16, &["a"], "a", 0);
 matset!(set17, &[".*a"], "a", 0);
+matset!(set18, &["a", "β"], "β", 1);
 
 nomatset!(nset1, &["a", "a"], "b");
 nomatset!(nset2, &["^foo", "bar$"], "bar foo");