CODEWITHSUNDEEP
rdeserialization
habana
habana

Reputation: 317

Decoded Base64 to dataframe

From an API I get a Base64 encoded dataset. I use RCurl::base64 to decode it, but it's serialized. How do I convert it to a dataframe?

After decoding my return, I get a long textstring with semi colon separated data and column names. Looks like this:

[1] "\"lfdn\";\"pseudonym\";\"external_lfdn\";\"tester\"\r\n\"50\";\"434444345\";\"0\";\"0\"\r\n\"91\";\"454444748\";\"0\";\"0\"\r\n\

You can see the structure with a simple cat(x):

"lfdn";"pseudonym";"external_lfdn";"tester"
"50";"434444345";"0";"0"
"91";"454444748";"0";"0"
"111";"444444141";"0";"0"

I've tried the obvious unserialize(x), but I get:

R> Error in unserialize(enc) : 
R> character vectors are no longer accepted by unserialize()

Whatever I throw at it... I can write the object to disk, and read it back in, but I prefer to avoid that.

Getting the data from the serialized textstring into a dataframe with column names would be great!

Upvotes: 1

Views: 303

Answers (1)

thothal
thothal

Reputation: 20379

This should do the trick:

read.table(text=j, header = TRUE, sep = ";")
#   lfdn pseudonym external_lfdn tester
# 1   50 434444345             0      0
# 2   91 454444748             0      0

Note. I copied your string from above, it does not contain the last row with 111 in it.

Upvotes: 1

Related Questions