From f18955d0242942fe396c8d4614569b66510009f8 Mon Sep 17 00:00:00 2001
From: Elijah Rippeth <elijah.rippeth@gmail.com>
Date: Sat, 12 Feb 2022 11:31:16 -0500
Subject: [PATCH 1/2] meaningless change to make XML well formed (though it is
 not important).

---
 test/datasets/test_iwslt2016.py | 3 ++-
 test/datasets/test_iwslt2017.py | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/test/datasets/test_iwslt2016.py b/test/datasets/test_iwslt2016.py
index 03b681f6f1..ccd5eb49e6 100644
--- a/test/datasets/test_iwslt2016.py
+++ b/test/datasets/test_iwslt2016.py
@@ -53,7 +53,8 @@ def _generate_uncleaned_train():
         # to ensure cleaning happens appropriately
         if random.random() < 0.1:
             open_tag = random.choice(xml_tags) + ">"
-            close_tag = "</" + open_tag[1:] + ">"
+            # Open tag already contains the closing >
+            close_tag = "</" + open_tag[1:-1] + ">"
             file_contents.append(open_tag + rand_string + close_tag)
         else:
             examples.append(rand_string + "\n")
diff --git a/test/datasets/test_iwslt2017.py b/test/datasets/test_iwslt2017.py
index 375ca4525d..5e373ee8ce 100644
--- a/test/datasets/test_iwslt2017.py
+++ b/test/datasets/test_iwslt2017.py
@@ -46,7 +46,8 @@ def _generate_uncleaned_train():
         # to ensure cleaning happens appropriately
         if random.random() < 0.1:
             open_tag = random.choice(xml_tags) + ">"
-            close_tag = "</" + open_tag[1:] + ">"
+            # Open tag already contains the closing >
+            close_tag = "</" + open_tag[1:-1] + ">"
             file_contents.append(open_tag + rand_string + close_tag)
         else:
             examples.append(rand_string + "\n")

From ab553b768da46c9c3c95f61489de3fafd4a7aa3c Mon Sep 17 00:00:00 2001
From: nayef211 <n63ahmed@edu.uwaterloo.ca>
Date: Tue, 8 Mar 2022 16:19:38 -0800
Subject: [PATCH 2/2] Remove closing doc tag from xml list. Update how close
 tags are created

---
 test/datasets/test_iwslt2016.py | 3 +--
 test/datasets/test_iwslt2017.py | 3 +--
 2 files changed, 2 insertions(+), 4 deletions(-)

diff --git a/test/datasets/test_iwslt2016.py b/test/datasets/test_iwslt2016.py
index ccd5eb49e6..c5afd37664 100644
--- a/test/datasets/test_iwslt2016.py
+++ b/test/datasets/test_iwslt2016.py
@@ -45,7 +45,6 @@ def _generate_uncleaned_train():
         "<title",
         "<speaker",
         "<doc",
-        "</doc",
     ]
     for i in range(100):
         rand_string = " ".join(random.choice(string.ascii_letters) for i in range(10))
@@ -54,7 +53,7 @@ def _generate_uncleaned_train():
         if random.random() < 0.1:
             open_tag = random.choice(xml_tags) + ">"
             # Open tag already contains the closing >
-            close_tag = "</" + open_tag[1:-1] + ">"
+            close_tag = "</" + open_tag[1:]
             file_contents.append(open_tag + rand_string + close_tag)
         else:
             examples.append(rand_string + "\n")
diff --git a/test/datasets/test_iwslt2017.py b/test/datasets/test_iwslt2017.py
index 5e373ee8ce..6f462e8599 100644
--- a/test/datasets/test_iwslt2017.py
+++ b/test/datasets/test_iwslt2017.py
@@ -38,7 +38,6 @@ def _generate_uncleaned_train():
         "<title",
         "<speaker",
         "<doc",
-        "</doc",
     ]
     for i in range(100):
         rand_string = " ".join(random.choice(string.ascii_letters) for i in range(10))
@@ -47,7 +46,7 @@ def _generate_uncleaned_train():
         if random.random() < 0.1:
             open_tag = random.choice(xml_tags) + ">"
             # Open tag already contains the closing >
-            close_tag = "</" + open_tag[1:-1] + ">"
+            close_tag = "</" + open_tag[1:]
             file_contents.append(open_tag + rand_string + close_tag)
         else:
             examples.append(rand_string + "\n")