[arrow] branch master updated: ARROW-9099: [C++][Gandiva] Implement trim function for string

[pr...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

praveenbingo pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/arrow.git


The following commit(s) were added to refs/heads/master by this push:
     new 4b8cbee  ARROW-9099: [C++][Gandiva] Implement trim function for string
4b8cbee is described below

commit 4b8cbee11735d7a97a6150938fbee8fba95d8d9e
Author: Sagnik Chakraborty <sa...@dremio.com>
AuthorDate: Wed Jun 24 11:33:42 2020 +0530

    ARROW-9099: [C++][Gandiva] Implement trim function for string
    
    Closes #7402 from sagnikc-dremio/master and squashes the following commits:
    
    60923b82a <Sagnik Chakraborty> ARROW-9099:  Implement trim function for string
    
    Authored-by: Sagnik Chakraborty <sa...@dremio.com>
    Signed-off-by: Praveen <pr...@dremio.com>
---
 cpp/src/gandiva/function_registry_string.cc    |  1 +
 cpp/src/gandiva/precompiled/string_ops.cc      | 36 ++++++++++++++++++++++++++
 cpp/src/gandiva/precompiled/string_ops_test.cc | 27 +++++++++++++++++++
 cpp/src/gandiva/precompiled/types.h            |  3 +++
 4 files changed, 67 insertions(+)

diff --git a/cpp/src/gandiva/function_registry_string.cc b/cpp/src/gandiva/function_registry_string.cc
index b4b02ef..6f2d54d 100644
--- a/cpp/src/gandiva/function_registry_string.cc
+++ b/cpp/src/gandiva/function_registry_string.cc
@@ -55,6 +55,7 @@ std::vector<NativeFunction> GetStringFunctionRegistry() {
       UNARY_UNSAFE_NULL_IF_NULL(length, {}, utf8, int32),
       UNARY_UNSAFE_NULL_IF_NULL(lengthUtf8, {}, binary, int32),
       UNARY_UNSAFE_NULL_IF_NULL(reverse, {}, utf8, utf8),
+      UNARY_UNSAFE_NULL_IF_NULL(trim, {}, utf8, utf8),
 
       UNARY_SAFE_NULL_NEVER_BOOL_FN(isnull, {}),
       UNARY_SAFE_NULL_NEVER_BOOL_FN(isnotnull, {}),
diff --git a/cpp/src/gandiva/precompiled/string_ops.cc b/cpp/src/gandiva/precompiled/string_ops.cc
index 0b2d934..6ee02dd 100644
--- a/cpp/src/gandiva/precompiled/string_ops.cc
+++ b/cpp/src/gandiva/precompiled/string_ops.cc
@@ -284,6 +284,42 @@ const char* reverse_utf8(gdv_int64 context, const char* data, gdv_int32 data_len
   return ret;
 }
 
+// Trim a utf8 sequence
+FORCE_INLINE
+const char* trim_utf8(gdv_int64 context, const char* data, gdv_int32 data_len,
+                      int32_t* out_len) {
+  if (data_len == 0) {
+    *out_len = 0;
+    return "";
+  }
+
+  gdv_int32 start = 0, end = data_len - 1;
+  // start and end denote the first and last positions of non-space
+  // characters in the input string respectively
+  while (start <= end && data[start] == ' ') {
+    ++start;
+  }
+  while (end >= start && data[end] == ' ') {
+    --end;
+  }
+
+  // string with no leading/trailing spaces, return original string
+  if (start == 0 && end == data_len - 1) {
+    *out_len = data_len;
+    return data;
+  }
+
+  // string with all spaces
+  if (start > end) {
+    *out_len = 0;
+    return "";
+  }
+
+  // string has some leading/trailing spaces and some non-space characters
+  *out_len = end - start + 1;
+  return data + start;
+}
+
 // Truncates the string to given length
 FORCE_INLINE
 const char* castVARCHAR_utf8_int64(gdv_int64 context, const char* data,
diff --git a/cpp/src/gandiva/precompiled/string_ops_test.cc b/cpp/src/gandiva/precompiled/string_ops_test.cc
index 75be26d..20a3d27 100644
--- a/cpp/src/gandiva/precompiled/string_ops_test.cc
+++ b/cpp/src/gandiva/precompiled/string_ops_test.cc
@@ -426,6 +426,33 @@ TEST(TestStringOps, TestReverse) {
   ctx.Reset();
 }
 
+TEST(TestStringOps, TestTrim) {
+  gandiva::ExecutionContext ctx;
+  uint64_t ctx_ptr = reinterpret_cast<gdv_int64>(&ctx);
+  gdv_int32 out_len = 0;
+  const char* out_str;
+
+  out_str = trim_utf8(ctx_ptr, "TestString", 10, &out_len);
+  EXPECT_EQ(std::string(out_str, out_len), "TestString");
+  EXPECT_FALSE(ctx.has_error());
+
+  out_str = trim_utf8(ctx_ptr, "      TestString  ", 18, &out_len);
+  EXPECT_EQ(std::string(out_str, out_len), "TestString");
+  EXPECT_FALSE(ctx.has_error());
+
+  out_str = trim_utf8(ctx_ptr, " Test  çåå†bD   ", 21, &out_len);
+  EXPECT_EQ(std::string(out_str, out_len), "Test  çåå†bD");
+  EXPECT_FALSE(ctx.has_error());
+
+  out_str = trim_utf8(ctx_ptr, "", 0, &out_len);
+  EXPECT_EQ(std::string(out_str, out_len), "");
+  EXPECT_FALSE(ctx.has_error());
+
+  out_str = trim_utf8(ctx_ptr, "      ", 6, &out_len);
+  EXPECT_EQ(std::string(out_str, out_len), "");
+  EXPECT_FALSE(ctx.has_error());
+}
+
 TEST(TestStringOps, TestLocate) {
   gandiva::ExecutionContext ctx;
   uint64_t ctx_ptr = reinterpret_cast<gdv_int64>(&ctx);
diff --git a/cpp/src/gandiva/precompiled/types.h b/cpp/src/gandiva/precompiled/types.h
index 09ee596..051f522 100644
--- a/cpp/src/gandiva/precompiled/types.h
+++ b/cpp/src/gandiva/precompiled/types.h
@@ -199,6 +199,9 @@ const char* lower_utf8(gdv_int64 context, const char* data, gdv_int32 data_len,
 const char* reverse_utf8(gdv_int64 context, const char* data, gdv_int32 data_len,
                          int32_t* out_len);
 
+const char* trim_utf8(gdv_int64 context, const char* data, gdv_int32 data_len,
+                      int32_t* out_len);
+
 gdv_int32 locate_utf8_utf8(gdv_int64 context, const char* sub_str, gdv_int32 sub_str_len,
                            const char* str, gdv_int32 str_len);