Add load feature

[ryo-ma]

Summary

example

import dim
dim.load(name="example-json", file_type="json")
# return example.json data as dict type

method

• load(name, file_type)
  • required
  • name
  • option
  • file_type(default: text)

Solutions

No response

APIs to be deprecated

No response

Additional Context

No response

[lisphilar]

Is this funciton for the following commands of dim-deno?

• $ dim install -f ./path/dim.json
• $ dim install -f https://raw.githubusercontent.com/xxxx/xxxx/main/dim.json

If so, we may implement dim.load(filepath_or_buffer: Union[str, bytes, os.PathLike] = "example-json.json"). filepath_or_buffer is a positional argument. (Please try to avoid type argument because this will over-write build-in type of Python.)

[ryo-ma]

$ dim install -f ./path/dim.json
$ dim install -f https://raw.githubusercontent.com/xxxx/xxxx/main/dim.json

These two correspond to install_from_file.

https://github.com/c-3lab/dim-python/issues/8

[ryo-ma]

This function can load the file by specifying the name described in dim.json

[lisphilar]

Sorry for my incorrect understanding.

Just to confirm, "name" is the dataset name (title) and this function returns each dataset as follows.

• dim.load(name="example-json", file_type="json") -> dict
• dim.load(name="example-csv", file_type="csv") -> pandas.DataFrame (for example)
• dim.load(name="example-xlsx", file_type="xlsx") -> pandas.DataFrame (for example)

Is not possible to save file_type information in dim.json or dim-lock.json? This may require changes of dim-deno commands.

[ryo-ma]

How does the user specify the file_type when saving file_type with dim.json?