paleolimbot opened a new pull request, #458:
URL: https://github.com/apache/arrow-nanoarrow/pull/458
This PR unifies how types are printed, which was different among the
`Array`, `Scalar`, `ArrayStream`, and `Schema`. These now all use the same
function such that type information is communicated in a consistent way for all
user-facing classes.
The general approach is to print all possible information on the first line.
Most types have descriptions like `int32`, so this is a good fit and there's no
need for multiple lines. This makes a list of schemas (e.g., `schema.fields`)
have a nice repr, too. The Schema repr additionally prints out metadata if
present (this is skipped by the other classes).
Some examples:
```python
import nanoarrow as na
# Schema with no parameters
na.int32()
#> <Schema> int32
# Type parameters are already part of the schema's to_string
na.fixed_size_binary(123)
#> <Schema> fixed_size_binary(123)
# non-nullable schema
na.int32(nullable=False)
#> <Schema> non-nullable int32
# named schema
na.Schema(na.int32(), name="some_col")
#> <Schema> 'some_col': int32
# ordered dictionary
na.dictionary(na.int32(), na.string(), True)
#> <Schema> ordered dictionary(int32)<string>
# Schema with metadata
na.Schema(na.int32(), metadata={"some key": "some value"})
#> <Schema> int32
#> - metadata: {b'some key': b'some value'}
# fields are already a part of the schema's to_string
struct = na.struct({"some col": na.int32(), "some other col": na.string()})
struct
#> <Schema> struct<some col: int32, some other col: string>
# ...and you can inspect them more closely with .fields
struct.fields
#> [<Schema> 'some col': int32, <Schema> 'some other col': string]
```
--
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]
