Inconsistent `zsv help <command>` formatting

[iamazeem]

Environment

• OS: Ubuntu 22.04 LTS
• zsv: v0.3.8-alpha

Description

The help output for some commands seems to be inconsistent in terms of:

• line length
• indentation of the flag description
• space between short and long flags
• placement of colon after flag
• missing description
• duplicate description
• argument format (underscore vs space)
• argument case mismatch (lowercase vs the rest)

Here are such observed instances:

• zsv help (indentation)

  zsv (un)register [<extension_id>]: (un)register an extension
      Registration info is saved in zsv.ini located in a directory determined as:
        ZSV_CONFIG_DIR environment variable value, if set
        otherwise, /usr/local/etc

  -u,--malformed-utf8-replacement <replacement_string>: replacement string (can be empty) in case of malformed UTF8 input
       (default for "desc" commamnd is '?')

• zsv help select (indentation, colon placement)

Usage: select [filename] [options] [-- col_specifier [... col_specifier]]
  where col_specifier is a column name or, if the -n option is used,
   a column index (starting at 1) or index range in the form of n-m
  e.g. select -n myfile.csv -- 1 4-6 50 10
       select myfile.csv -- first_col fiftieth_column "Tenth Column"

  -e <embedded lineend char>  : char to replace embedded lineend. if none provided, embedded lineends are preserved
      If the provided string begins with 0x, it will be interpreted as the hex representation of a string

  -n: provided column indexes are numbers corresponding to column positions (starting with 1), instead of names

  --whitespace-clean-no-newline: clean whitespace and remove embedded newlines
  -W,--no-trim: do not trim whitespace
  -o <output filename>: name of file to save output to

• zsv help sql (indentation, colon placement)

sql: run ad hoc sql on a CSV file
          or join multiple CSV files on one or more common column(s)

Options:
  --join-indexes <n1...>: specify one or more column names to join multiple files by
     each n is treated as an index in the first input file that determines a column
     of the join. For example, if joining two files that, respectively, have columns
     A,B,C,D and X,B,C,A,Y then `--join-indexes 1,3` will join on columns A and C.
     When using this option, do not include an sql statement
  -b: output with BOM

• zsv help count (missing description, one space indentation whereas others have two, redundant square brackets)

Usage: count [options]
Options:
 -h, --help            : show usage
 [-i, --input] <filename>: use specified file input

• zsv help desc (colon placement)

Options:
  -b, --with-bom : output with BOM
  -C <maximum_number_of_columns>: defaults to 1024
  -H: only output header names
  -q, --quick: minimize example counts,
  -a, --all: calculate all metadata (for now, this only adds uniqueness info)
  -o <output filename>: name of file to save output to (defaults to stdout)

• zsv help pretty (indentation, double colons)

  -W, --width <n>: set the max line width to output. if not provided
                            will try to detect automatically
  -p, --rows:             : set the number of (preview) rows to calculated widths from
                            if not provided, defaults to 150

• zsv help flatten (long description, formatting of flags, colon placement)

flatten: flatten a table, based on a single-column key assuming that rows to flatten always appear in contiguous blocks

Usage: flatten [<filename>] [<options>] -- [aggregate_output_spec ...]
Each aggregate output specification is either (i) a single-column aggregation or (future: (ii) the "*" placeholder (in conjunction with -a)).
A single-column aggregation consists of the column name or index, followed by the equal sign (=) and then an aggregation method.
If the equal sign should be part of the column name, it can be escaped with a preceding backslash.

The following aggregation methods may be used:  array (pipe-delimited)
  array_<delim> (user-specified delimiter)

Options:
  -b: output with BOM
  -v, --verbose: display verbose messages
  -C <max columns to output>: maximum number of columns to output
  -m <max rows per aggregation>: defaults to 1024. If this limit is reached for any aggregation,
     an error will be output
  --row-id <Row ID column name>: Required. name of column to group by
  --col-name <Column ID column name>: name of column specifying the output column name
  -V <Value column name>: name of column specifying the output value
  (future: -a <Aggregation method>: aggregation method to use for the select-all placeholder)
  -o <output filename>: name of file to save output to

• zsv help 2json (flags: extra space after comma)

Options:
  -h, --help
  -o, --output <filename>       : output to specified filename

• zsv help 2tsv (indentation)

2tsv: convert CSV to TSV (tab-delimited text) suitable for simple-delimiter
       text processing. By default, embedded tabs or multilines will be escaped
       to \t, \n or \r, respectively

• zsv help serialize (duplicate description)

serialize: Serialize a CSV file into Row/Colname/Value triplets

Usage: serialize [<filename>]
Serializes a CSV file

• zsv help stack (flags)

zsv help stack
stack: stack one or more csv files vertically, aligning columns with the same name

Usage: stack [options] filename [filename...]

Options:
  -o <filename>: output file
  -b: output with BOM
  -q: always add double-quotes
  -T: input is tab-delimited, instead of comma-delimited

• zsv help compare (missing description)

Usage: compare [options] [file1.csv] [file2.csv] [...]

There may be other instances too.