Hi, I suggest some improvement for this great library.

In East Asian languages, including Japanese, characters are represented with a width twice that of ASCII. This is expressed as full width in Unicode.

This library seems to calculate the amount of padding simply by character count, so if input that includes full-width characters is given, the table display will be corrupted.

Since there are libraries that determine whether Unicode characters are full-width or not[1], it would be possible to render the table accurately by improving the algorithm that calculates the amount of padding.

Actual output:

Thanks!

[1] https://unicode-org.github.io/icu-docs/apidoc/dev/icu4j/com/ibm/icu/lang/UCharacter.html#getIntPropertyValue-int-int-

I'm using this library mostly for debugging trading bots. The problem is that the prices never go to 0, but instead hang around, let's say, 8000. The charts force the zero to be in range of the Y axis, which flattens my chart to a point where I can't see movements anymore. Can you add an option to not force zero in range on the Y axis? Or better yet, allow us to determine our own ranges?

Could you please port some features of the UnicodePlots project?

• https://github.com/JuliaPlots/UnicodePlots.jl

Especially scatterplot and lineplot

An NPE is thrown when some of the data cells are null. It would be good if these could be rendered in a special way (e.g. as "null" or configurable via a new method withNullValue(String) ) to differentiate them from empty strings.

Example to reproduce the NPE:

System.out.println(ASCIITable.fromData(new String[]{"ID", "Name"}, new String[][]{{"1", null},}).toString());

This results in:

Exception in thread "main" java.lang.NullPointerException
	at com.mitchtalmadge.asciidata.table.ASCIITable.<init>(ASCIITable.java:79)

I appreciate that I could convert the data before passing it into ASCIITable, but having absent values is quite a common scenario and I think your library would be improved by handling this gracefully.

I would like an option to generate a table without a header. Sometimes there is no meaning to the columns of the table, so the header cells are empty. As a workaround, I strip the first two lines of the generated table and fix the corner and column separator characters.

For the tables, it would be really useful to have an option on the ASCIITable class to set a column's justification, so that we can create tables with centered or right-aligned content. What do you think about it?

Examples

Instead of this:

╔═══════════╤══════════╗
║ FirstName │ LastName ║
╠═══════════╪══════════╣
║ Foo       │ Bar      ║
╟───────────┼──────────╢
║ Kit       │ Kat      ║
╚═══════════╧══════════╝

I sometimes prefer this:

╔═══════════╤══════════╗
║ FirstName │ LastName ║
╠═══════════╪══════════╣
║ Foo       │      Bar ║
╟───────────┼──────────╢
║ Kit       │      Kat ║
╚═══════════╧══════════╝

Escpecially when numbers are involved:

╔═══════════╤══════════╗
║ Name      │ Score    ║
╠═══════════╪══════════╣
║ Foo       │        5 ║
╟───────────┼──────────╢
║ Kit       │      123 ║
╚═══════════╧══════════╝