How to import data asset in R notebook with R function

Question

Hi everyone

When I followed the instructions in Interactive R development. I encountered several problems as follows:
https://learn.microsoft.com/en-us/azure/machine-learning/how-to-r-interactive-development?view=azureml-api-2

1. unable to load data file with R function (not python function)

In the section "Read tabular data from registered data assets or datastores", I created credential as described in steps 1-3. In step 4: I can also successfully load the data asset with Pandas.
pd <- import("pandas")
cc <- pd$read_csv(py$data_uri)
head(cc)

But, if I choose the R functions, I can not load the data:

df <- read.csv(py$data_uri)

The error message is
Warning message in file(file, "rt"): “cannot open file 'azureml://subscriptions/*****/resourcegroups/****/workspaces/****/datastores/****/paths/clean_street_key/*****.csv': No such file or directory” Error in file(file, "rt"): cannot open the connection

In my opinion, this message is irrational, because it recognized already my data file, and I can load it with Python function.

2. unable to use the filestore object

In the following code provided, a filestore object "azureml.fsspec$AzureMachineLearningFileSystem" required twice the uri : once uri, secondly <path>. But after filled in the information, it still can not read data. I doubt if something is wrong in the filestore?

fs <- azureml.fsspec$AzureMachineLearningFileSystem(uri, sep = "")

df <- with(fs$open("<path>)", "r") %as% f, {
 x <- as.character(f$read(), encoding = "utf-8")
 read.csv(textConnection(x), header = TRUE, sep = ",", stringsAsFactors = FALSE)
})
print(df)

Thanks