[R] Type function names don't match type names

[asfimport]

I noticed this while working on documentation for ARROW-5505, trying to show how you could pass an explicit schema definition to make a table. For a few types, the name of the type that gets printed (and comes from the C++ library) doesn't match the name of the function you use to specify the type in a schema:

> tab <- to_arrow(data.frame(
+   a = 1:10,
+   b = as.numeric(1:10),
+   c = sample(c(TRUE, FALSE, NA), 10, replace = TRUE),
+   d = letters[1:10],
+   stringsAsFactors = FALSE
+ ))
> tab$schema
arrow::Schema 
a: int32
b: double
c: bool
d: string 
# Alright, let's make that schema
> schema(a = int32(), b = double(), c = bool(), d = string())
Error in bool() : could not find function "bool"
# Hmm, ok, so bool --> boolean()
> schema(a = int32(), b = double(), c = boolean(), d = string())
Error in string() : could not find function "string"
# string --> utf8()
> schema(a = int32(), b = double(), c = boolean(), d = utf8())
Error: type does not inherit from class arrow::DataType
# Wha?
> double()
numeric(0)
# Oh. double is a base R function.
> schema(a = int32(), b = float64(), c = boolean(), d = utf8())
arrow::Schema 
a: int32
b: double
c: bool
d: string 

If you believe this switch statement is correct, these three, along with float and half_float, are the only mismatches: https://github.com/apache/arrow/blob/master/r/R/R6.R#L81-L109

> schema(b = float64(), c = boolean(), d = utf8(), e = float32(), f = float16())
arrow::Schema 
b: double
c: bool
d: string
e: float
f: halffloat 

I can add aliases (i.e. another function that does the same thing) for bool, string, float, and halffloat, and I can add some magic so that double() (and even integer()) work inside the schema() function. But in looking into the C++ side to confirm where these alternate type names were coming from, I saw some inconsistencies. For example, https://github.com/apache/arrow/blob/master/cpp/src/arrow/type.h#L773-L788 suggests that the StringType should report its name as "utf8". But the ToString method here https://github.com/apache/arrow/blob/master/cpp/src/arrow/type.cc#L191 has it report as "string". It's unclear why those should report differently.

Reporter: Neal Richardson / @nealrichardson
Assignee: Neal Richardson / @nealrichardson

PRs and other links:

• GitHub Pull Request #5198

_{Note: This issue was originally created as ARROW-6338. Please see the migration documentation for further details.}

[asfimport]

Francois Saint-Jacques / @fsaintjacques:
The names you want to follow are the factory functions:

• https://github.com/apache/arrow/blob/master/cpp/src/arrow/type_fwd.h#L208-L245

• https://github.com/apache/arrow/blob/master/cpp/src/arrow/type.h#L1358-L1509

  As for inconsistencies between ToString(), type_name(), name(), I think it's worth reconciling, @pitrou.

[asfimport]

Ben Kietzman / @bkietz:
@nealrichardson @wesm I don't think those names are used for anything but error logging purposes; IE nothing at protocol level depends on them and it should be safe to just change them in C++

Ran this query for all usages of DataType::name and came up with nothing but error messages (those should be changed to use ToString, which is more informative than name.)

$ for f in `ls cpp/src/arrow/**/*.cc`
do
clang-query-7 -h -p=cpp/build/compile_commands.json -c='match cxxMemberCallExpr(callee(functionDecl(hasName("DataType::name"))))' $f >> locations
done
$ cat locations | grep '"root" binds here' | uniq
/home/bkietz/arrow/cpp/src/arrow/ipc/json_internal.cc:365:70: note: "root" binds here
/home/bkietz/arrow/cpp/src/arrow/ipc/json_internal.cc:1418:70: note: "root" binds here
/home/bkietz/arrow/cpp/src/arrow/python/arrow_to_pandas.cc:1907:70: note: "root" binds here
/home/bkietz/arrow/cpp/src/arrow/record_batch.cc:107:51: note: "root" binds here
/home/bkietz/arrow/cpp/src/arrow/record_batch.cc:108:66: note: "root" binds here

[asfimport]

Neal Richardson / @nealrichardson:
We are using those factory functions currently: https://github.com/apache/arrow/blob/master/r/R/R6.R#L281-L376

The point is that the factory function names don't always match the type names, and I'm curious if there's a good reason for that mismatch.

[asfimport]

Ben Kietzman / @bkietz:
When writing the C API "double" can't be used because it's reserved, so we use "float64" for the DataType factory and "DoubleArray", "DoubleType", "Type::DOUBLE" everywhere else because those variations are not reserved. Similarly, when deciding what string to return from DoubleType::name(), "double" was chosen just because that's what one calls a 64 bit floating point number in C. I don't believe it would break anything to amend DoubleType::name() to return "float64" instead.

[asfimport]

Antoine Pitrou / @pitrou:
Issue resolved by pull request 5198
#5198