paleolimbot commented on a change in pull request #12030:
URL: https://github.com/apache/arrow/pull/12030#discussion_r778353013
##########
File path: r/src/io.cpp
##########
@@ -178,4 +180,134 @@ void io___BufferOutputStream__Write(
StopIfNotOk(stream->Write(RAW(bytes), bytes.size()));
}
+// TransformInputStream::TransformFunc wrapper
+
+class RIconvWrapper {
+ public:
+ RIconvWrapper(std::string to, std::string from)
+ : handle_(Riconv_open(to.c_str(), from.c_str())) {
+ if (handle_ == ((void*)-1)) {
+ cpp11::stop("Can't convert encoding from '%s' to '%s'", from.c_str(),
to.c_str());
+ }
+ }
+
+ size_t iconv(const char** inbuf, size_t* inbytesleft, char** outbuf,
+ size_t* outbytesleft) {
+ return Riconv(handle_, inbuf, inbytesleft, outbuf, outbytesleft);
+ }
+
+ ~RIconvWrapper() {
+ if (handle_ != ((void*)-1)) {
+ Riconv_close(handle_);
+ }
+ }
+
+ protected:
+ void* handle_;
+};
+
+struct ReencodeUTF8TransformFunctionWrapper {
+ explicit ReencodeUTF8TransformFunctionWrapper(std::string from)
+ : from_(from), iconv_("UTF-8", from), n_pending_(0) {}
+
+ // This may get copied and we need a fresh RIconvWrapper for each copy.
+ ReencodeUTF8TransformFunctionWrapper(const
ReencodeUTF8TransformFunctionWrapper& ref)
+ : ReencodeUTF8TransformFunctionWrapper(ref.from_) {}
+
+ arrow::Result<std::shared_ptr<arrow::Buffer>> operator()(
+ const std::shared_ptr<arrow::Buffer>& src) {
+ ARROW_ASSIGN_OR_RAISE(auto dest, arrow::AllocateResizableBuffer(32));
Review comment:
I'm totally game to use it, but in the current version I only interact
with the buffer object 3 times:
```cpp
ARROW_ASSIGN_OR_RAISE(auto dest,
arrow::AllocateResizableBuffer(initial_size));
```
```cpp
if (out_bytes_left < in_bytes_left) {
RETURN_NOT_OK(dest->Resize(dest->size() * 1.2));
out_buf = dest->mutable_data() + out_bytes_used;
out_bytes_left = dest->size() - out_bytes_used;
}
```
```cpp
RETURN_NOT_OK(dest->Resize(out_bytes_used, false));
```
I found that with the `BufferBuiler` I needed an extra line for `Finish()`
and the rest was almost the same (`Reserve()` instead of `Resize()`). Is there
anything that I'm missing that would be safer or more readable using the
`BufferBuilder`?
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
