Renaming

[lorenzwalthert]

Naming of data

At the moment, for the parse table returned by utils::getParseData and
processed through various functions, we use many different names in different
functions for essentially this (flat) table:

• pd (all rules in rules.R)
• parse_data (serialize_parse_data)
• parse_data_with_ws (serialize_parse_data_flat)
• data (create_filler)

I think it would simplify things to just use one variable name throughout. I
suggest parse_data. We could use parse_data_flt for the flat version and and
parse_data_nstd for the nested version for situations where the processing
differs.

Naming of functions

As far as I understand, the terminology around ws is more of a legacy thing, i.e. in the functions compute_parse_data_flat_with_ws or add_ws_to_parse_data since it is not really just white space information that is affected by these functions, but rather a bunch of preprocessing Hence, I would suggest to rename the latter to pre_process_parse_data and the former to compute_parse_data_flt for now.

[krlmlr]

A consistent name is a good idea. Can we simply use nested and flat, or perhaps standardize on pd for parse data and use this as a prefix or suffix?

[lorenzwalthert]

Ok yes, let us just use pd then and use flat and nested as suffix.

[lorenzwalthert]

Reference: #146 (comment).

[lorenzwalthert]

So as stated in #146 (comment), should we use nest for "a parse table at one level of nesting, which always represents a complete expression", and pd_nested for the whole nested structure, i.e. the parse table at the top level of nesting?

[lorenzwalthert]

This includes documentation also obviously.

[krlmlr]

Yes, we can keep the terms "parse data" and "nested parse data" for simplicity. The latter consists of multiple interlinked nests, this is the only new term.

[lorenzwalthert]

Reference: #132.

[lorenzwalthert]

Renaming of files

• modify_pd.R -> indent.R
• nested.R -> nest.R
• ws.R -> ui.R
• style_md.R -> transform-code.R.
• transform.R -> transform-files.R

[krlmlr]

Do we need to change anything here, short of renaming the files as you suggested?

[lorenzwalthert]

Seems no, if we want to leave the rest (pd, pd_flat, pd_nested etc.) as is. However, I think we should change it too for consistency. Not all of it though, e.g. flattened_pd is not a nest anymore.

[lorenzwalthert]

• serialized_tests.R to serialized-tests.R
• style_guides.R to style-guides.R
• move stop_insufficient_r_version() to communicate.
• dplyr.R to compat-dplyr.R
• consistency: is_fun_dec() or pd_is_fun_dec() in expr-is.R