WillAyd commented on issue #597:
URL:
https://github.com/apache/arrow-nanoarrow/issues/597#issuecomment-2313017661
In your mind are you thinking of creating these in plain C just to increase
their use outside of nanoarrow? The challenge I think with plain C extensions
is getting reference counting correct, especially with branches that may return
early. I think an equivalent of the above code in C would look like:
```c
static int AppendBytes(const CArrayBuilder *builder, PyObject *obj) {
ArrowErrorCode code;
ArrowBufferView item
if (!PyIter_Check(obj)) {
PyErr_SetString(PyExc_TypeError, "Object is not iterable");
return -1;
}
PyObject *iterator = PyObject_GetIter(obj);
if (iterator == NULL)
return NULL;
PyObject *item;
while ((item = PyIter_Next(iterator))) {
if (item == Py_None) {
code = ArrowArrayAppendNull(builder->_ptr, 1);
} else {
Py_buffer buffer;
if (PyObject_GetBuffer(py_item, &buffer, PyBUF_ANY_CONTIGUOUS |
PyBUF_FORMAT)) {
Py_DECREF(item);
Py_DECREF(iterator);
return NULL;
}
if (buffer.ndim != 1) {
PyErr_SetString(PyExc_ValueError, "Can't append buffer with
dimensions != 1 to binary array");
// TODO: current implementation might be missing a release?
// PyBuffer_Release(&buffer);
Py_DECREF(item);
Py_DECREF(iterator);
return NULL;
}
if (buffer.itemsize != 1) {
PyErr_SetString(PyExc_ValueError, "Can't append buffer with
itemsize != 1 to binary array");
PyBuffer_Release(&buffer);
Py_DECREF(item);
Py_DECREF(iterator);
return NULL;
}
item.data.data = buffer.buf;
item.size_bytes = buffer.len;
code = ArrowArrayAppendBytes(ptr_, item);
PyBuffer_Release(&buffer);
}
if (code != NANOARROW_OK) {
// TODO: might be a format specifier that doesn't require
PyObject_Str
PyObject *str = PyObject_Str(item);
// TODO: might need to create a custom exception
PyErr_Format(PyExc_ValueError, "append bytes item %s failed %d",
str, code);
Py_DECREF(str);
Py_DECREF(item);
Py_DECREF(iterator);
return -1;
}
Py_DECREF(item);
}
Py_DECREF(iterator);
return 0;
}
```
of course code use goto cleverly to better simulate an RAII cleanup of
resources, but again wanted to keep this as equivalent to current code as
possible
--
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]
