From ad126065109aaa72b6eb324ba5abd555b70bb4ae Mon Sep 17 00:00:00 2001
From: David Wendt <45795991+davidwendt@users.noreply.github.com>
Date: Tue, 3 May 2022 07:44:19 -0400
Subject: [PATCH] Fix replace error when regex has only zero match quantifiers
 (#10760)

Closes #10753

Fixes `cudf::strings::replace_re` logic that was reading past the end of a string when given a regex that contained net zero match quantifier pattern (e.g. 'D*' or 'D?s?' both can match to nothing).

Authors:
  - David Wendt (https://github.com/davidwendt)

Approvers:
  - Nghia Truong (https://github.com/ttnghia)
  - Bradley Dice (https://github.com/bdice)
  - Mark Harris (https://github.com/harrism)

URL: https://github.com/rapidsai/cudf/pull/10760
---
 cpp/src/strings/replace/replace_re.cu     | 15 ++++++++-------
 cpp/tests/strings/replace_regex_tests.cpp | 12 ++++++++++++
 python/cudf/cudf/tests/test_string.py     | 10 ++++++++++
 3 files changed, 30 insertions(+), 7 deletions(-)

diff --git a/cpp/src/strings/replace/replace_re.cu b/cpp/src/strings/replace/replace_re.cu
index d42359deeac..af74d8bdb92 100644
--- a/cpp/src/strings/replace/replace_re.cu
+++ b/cpp/src/strings/replace/replace_re.cu
@@ -54,18 +54,19 @@ struct replace_regex_fn {
       return;
     }
 
-    auto const d_str = d_strings.element<string_view>(idx);
-    auto nbytes      = d_str.size_bytes();                  // number of bytes in input string
-    auto mxn = maxrepl < 0 ? d_str.length() + 1 : maxrepl;  // max possible replaces for this string
-    auto in_ptr        = d_str.data();                      // input pointer (i)
-    auto out_ptr       = d_chars ? d_chars + d_offsets[idx]  // output pointer (o)
-                                 : nullptr;
+    auto const d_str  = d_strings.element<string_view>(idx);
+    auto const nchars = d_str.length();
+    auto nbytes       = d_str.size_bytes();             // number of bytes in input string
+    auto mxn     = maxrepl < 0 ? nchars + 1 : maxrepl;  // max possible replaces for this string
+    auto in_ptr  = d_str.data();                        // input pointer (i)
+    auto out_ptr = d_chars ? d_chars + d_offsets[idx]   // output pointer (o)
+                           : nullptr;
     size_type last_pos = 0;
     int32_t begin      = 0;   // these are for calling prog.find
     int32_t end        = -1;  // matches final word-boundary if at the end of the string
 
     // copy input to output replacing strings as we go
-    while (mxn-- > 0) {  // maximum number of replaces
+    while (mxn-- > 0 && begin <= nchars) {  // maximum number of replaces
 
       if (prog.is_empty() || prog.find<stack_size>(idx, d_str, begin, end) <= 0) {
         break;  // no more matches
diff --git a/cpp/tests/strings/replace_regex_tests.cpp b/cpp/tests/strings/replace_regex_tests.cpp
index 2b9e8b7aae7..1ccbc6fa676 100644
--- a/cpp/tests/strings/replace_regex_tests.cpp
+++ b/cpp/tests/strings/replace_regex_tests.cpp
@@ -157,6 +157,18 @@ TEST_F(StringsReplaceRegexTest, WordBoundary)
   CUDF_TEST_EXPECT_COLUMNS_EQUIVALENT(*results, expected);
 }
 
+TEST_F(StringsReplaceRegexTest, ZeroLengthMatch)
+{
+  cudf::test::strings_column_wrapper input({"DD", "zéz", "DsDs", ""});
+  auto repl     = cudf::string_scalar("_");
+  auto results  = cudf::strings::replace_re(cudf::strings_column_view(input), "D*", repl);
+  auto expected = cudf::test::strings_column_wrapper({"__", "_z_é_z_", "__s__s_", "_"});
+  CUDF_TEST_EXPECT_COLUMNS_EQUIVALENT(*results, expected);
+  results  = cudf::strings::replace_re(cudf::strings_column_view(input), "D?s?", repl);
+  expected = cudf::test::strings_column_wrapper({"___", "_z_é_z_", "___", "_"});
+  CUDF_TEST_EXPECT_COLUMNS_EQUIVALENT(*results, expected);
+}
+
 TEST_F(StringsReplaceRegexTest, Multiline)
 {
   auto const multiline = cudf::strings::regex_flags::MULTILINE;
diff --git a/python/cudf/cudf/tests/test_string.py b/python/cudf/cudf/tests/test_string.py
index d600fdeee27..d212c6b2072 100644
--- a/python/cudf/cudf/tests/test_string.py
+++ b/python/cudf/cudf/tests/test_string.py
@@ -923,6 +923,16 @@ def test_string_replace(
         assert_eq(expect, got)
 
 
+@pytest.mark.parametrize("pat", ["A*", "F?H?"])
+def test_string_replace_zero_length(ps_gs, pat):
+    ps, gs = ps_gs
+
+    expect = ps.str.replace(pat, "_", regex=True)
+    got = gs.str.replace(pat, "_", regex=True)
+
+    assert_eq(expect, got)
+
+
 def test_string_lower(ps_gs):
     ps, gs = ps_gs