From 0471c74929ae7291749e88cf9197e69505a1ff65 Mon Sep 17 00:00:00 2001
From: Andrew Gallant <jamslam@gmail.com>
Date: Thu, 27 Oct 2016 17:36:12 -0400
Subject: [PATCH] Fixes a bug introduced by the fix in 630049.

This is another variant of incorrectly producing an unambiguous set of
literals.

Fixes #291.
---
 .gitignore                   |  1 +
 regex-syntax/Cargo.toml      |  3 +++
 regex-syntax/src/literals.rs | 12 +++++++++---
 tests/regression.rs          |  6 +++++-
 4 files changed, 18 insertions(+), 4 deletions(-)

diff --git a/.gitignore b/.gitignore
index dc01bfda71..7c25d48667 100644
--- a/.gitignore
+++ b/.gitignore
@@ -3,3 +3,4 @@ Cargo.lock
 bench-log
 .*.swp
 wiki
+tags
diff --git a/regex-syntax/Cargo.toml b/regex-syntax/Cargo.toml
index ee00ec07cf..1f935cff0d 100644
--- a/regex-syntax/Cargo.toml
+++ b/regex-syntax/Cargo.toml
@@ -11,3 +11,6 @@ description = "A regular expression parser."
 [dev-dependencies]
 quickcheck = "0.2"
 rand = "0.3"
+
+[profile.test]
+codegen-units = 8
diff --git a/regex-syntax/src/literals.rs b/regex-syntax/src/literals.rs
index 49de379be4..0b89ccb250 100644
--- a/regex-syntax/src/literals.rs
+++ b/regex-syntax/src/literals.rs
@@ -250,9 +250,12 @@ impl Literals {
                     }
                 } else {
                     if let Some(i) = position(&lit2, &candidate) {
-                        candidate.truncate(i);
-                        candidate.cut();
                         lit2.cut();
+                        let mut new_candidate = candidate.clone();
+                        new_candidate.truncate(i);
+                        new_candidate.cut();
+                        old.push(new_candidate);
+                        candidate.clear();
                     }
                 }
                 // Oops, the candidate is already represented in the set.
@@ -1385,7 +1388,7 @@ mod tests {
                 vec![M("Mo'"), M("Mu'"), M("Mo"), M("Mu")],
                 vec![C("Mo"), C("Mu")]);
     test_unamb!(unambiguous11,
-                vec![M("zazb"), M("azb")], vec![C("azb"), C("z")]);
+                vec![M("zazb"), M("azb")], vec![C("a"), C("z")]);
     test_unamb!(unambiguous12, vec![M("foo"), C("foo")], vec![C("foo")]);
     test_unamb!(unambiguous13,
                 vec![M("ABCX"), M("CDAX"), M("BCX")],
@@ -1393,6 +1396,9 @@ mod tests {
     test_unamb!(unambiguous14,
                 vec![M("IMGX"), M("MVIX"), M("MGX"), M("DSX")],
                 vec![M("DSX"), C("I"), C("MGX"), C("MV")]);
+    test_unamb!(unambiguous15,
+                vec![M("IMG_"), M("MG_"), M("CIMG")],
+                vec![C("C"), C("I"), C("MG_")]);
 
 
     // ************************************************************************
diff --git a/tests/regression.rs b/tests/regression.rs
index fb85507561..d2732467cf 100644
--- a/tests/regression.rs
+++ b/tests/regression.rs
@@ -70,4 +70,8 @@ ismatch!(partial_anchor_alternate_begin, u!(r"^a|z"), "yyyyya", false);
 ismatch!(partial_anchor_alternate_end, u!(r"a$|z"), "ayyyyy", false);
 
 // See: https://github.com/rust-lang-nursery/regex/issues/289
-mat!(lits_unambiguous, u!(r"(ABC|CDA|BC)X"), "CDAX", Some((0, 4)));
+mat!(lits_unambiguous1, u!(r"(ABC|CDA|BC)X"), "CDAX", Some((0, 4)));
+
+// See: https://github.com/rust-lang-nursery/regex/issues/291
+mat!(lits_unambiguous2, u!(r"((IMG|CAM|MG|MB2)_|(DSCN|CIMG))(?P<n>[0-9]+)$"),
+     "CIMG2341", Some((0, 8)), Some((0, 4)), None, Some((0, 4)), Some((4, 8)));