pitrou commented on a change in pull request #8621:
URL: https://github.com/apache/arrow/pull/8621#discussion_r520675172
##########
File path: cpp/src/arrow/compute/kernels/scalar_string.cc
##########
@@ -1231,6 +1252,302 @@ Result<ValueDescr> StrptimeResolve(KernelContext* ctx,
const std::vector<ValueDe
return Status::Invalid("strptime does not provide default StrptimeOptions");
}
+#ifdef ARROW_WITH_UTF8PROC
+
+template <typename Type, bool left, bool right, typename Derived>
+struct UTF8TrimWhitespaceBase : StringTransform<Type, Derived> {
+ using Base = StringTransform<Type, Derived>;
+ using offset_type = typename Base::offset_type;
+ bool Transform(const uint8_t* input, offset_type input_string_ncodeunits,
+ uint8_t* output, offset_type* output_written) {
+ const uint8_t* begin = input;
+ const uint8_t* end = input + input_string_ncodeunits;
+ const uint8_t* end_trimmed = end;
+ const uint8_t* begin_trimmed = begin;
+
+ auto predicate = [](uint32_t c) { return !IsSpaceCharacterUnicode(c); };
+ if (left && !ARROW_PREDICT_TRUE(
+ arrow::util::UTF8FindIf(begin, end, predicate,
&begin_trimmed))) {
+ return false;
+ }
+ if (right && (begin_trimmed < end)) {
+ if (!ARROW_PREDICT_TRUE(arrow::util::UTF8FindIfReverse(begin_trimmed,
end,
+ predicate,
&end_trimmed))) {
+ return false;
+ }
+ }
+ std::copy(begin_trimmed, end_trimmed, output);
+ *output_written = static_cast<offset_type>(end_trimmed - begin_trimmed);
+ return true;
+ }
+ void Execute(KernelContext* ctx, const ExecBatch& batch, Datum* out) {
+ EnsureLookupTablesFilled();
+ Base::Execute(ctx, batch, out);
+ }
+};
+
+template <typename Type>
+struct UTF8TrimWhitespace
+ : UTF8TrimWhitespaceBase<Type, true, true, UTF8TrimWhitespace<Type>> {};
+
+template <typename Type>
+struct UTF8LTrimWhitespace
+ : UTF8TrimWhitespaceBase<Type, true, false, UTF8LTrimWhitespace<Type>> {};
+
+template <typename Type>
+struct UTF8RTrimWhitespace
+ : UTF8TrimWhitespaceBase<Type, false, true, UTF8RTrimWhitespace<Type>> {};
+
+template <typename Type, bool left, bool right, typename Derived>
+struct UTF8TrimBase : StringTransform<Type, Derived> {
+ using Base = StringTransform<Type, Derived>;
+ using offset_type = typename Base::offset_type;
+ using State = OptionsWrapper<TrimOptions>;
+ TrimOptions options;
+ std::set<uint32_t> codepoints;
Review comment:
`std::set` doesn't sound like a terrific data structure for this. I
would expect `std::vector<bool>` (which is
[specified](https://en.cppreference.com/w/cpp/container/vector_bool) as
optimizing memory footprint) to give better perf, though benchmarks are
required to confirm this.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
