fix: Columns not auto sizing correctly

[ericlln]

• Improved logic for calculating the range to truncate a string, based on the width of the assumed largest character and smallest character for a given font. This fixes the observed issue of wide strings not being truncated.
• Changed column width calculation to calculate text width by iterating through the text and adding up individual character widths which are cached. This approach size columns more precisely than estimating based on a single character's width and the length of the string, and does not appear to have a significant performance impact.

Test snippet used to see if column width calculation works as expected; resize columns to observe truncation behaviour

from deephaven import new_table
from deephaven.column import string_col

a = new_table([string_col("allllllllllllllllllllllllllllllllllllllllllllllll", ["a"])])
b = new_table([string_col("ammmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm", ["b"])])
c = new_table([string_col("c", ["allllllllllllllllllllllllllllllllllllllllllllllll"])])
d = new_table([string_col("d", ["ammmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm"])])

Closes #2288

[codecov]

Codecov Report

Attention: Patch coverage is 50.00000% with 39 lines in your changes missing coverage. Please review.

Project coverage is 46.80%. Comparing base (ae544fa) to head (da41aaf).
Report is 6 commits behind head on main.

Files with missing lines	Patch %	Lines
packages/grid/src/GridMetricCalculator.ts	45.09%	28 Missing ⚠️
packages/iris-grid/src/IrisGridRenderer.ts	0.00%	6 Missing ⚠️
packages/grid/src/DataBarCellRenderer.ts	0.00%	2 Missing ⚠️
packages/iris-grid/src/IrisGridTextCellRenderer.ts	0.00%	2 Missing ⚠️
packages/grid/src/TextCellRenderer.ts	85.71%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2359      +/-   ##
==========================================
+ Coverage   46.76%   46.80%   +0.03%     
==========================================
  Files         710      710              
  Lines       39107    39141      +34     
  Branches     9773     9956     +183     
==========================================
+ Hits        18288    18318      +30     
- Misses      20808    20812       +4     
  Partials       11       11              

Flag	Coverage Δ
unit	46.80% <50.00%> (+0.03%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[dgodinez-dh]

In the ticket, the complaint was about a capital 'M' overflowing the width. Is lowercase 'm' going to have the same width?

[ericlln]

When I was initially investigating this issue I measured all alphanumeric characters using our canvas context, and found that a lowercase ‘m’ is a wider than an upper case one by about 8%, so I ended up taking that as the largest value. This might not be true for all fonts but it seems to be more common based on my limited research.

The Width of Lowercase ´m´

[dgodinez-dh]

I see in a lot of places that we no longer check the default font. What happens if the font is not the map? Do we get a default somehow, or is it guaranteed to be calculated?

[ericlln]

I saw that this constant was being used in a lot of different places, since CellRenderer.getCachedTruncatedString doesn’t accept an undefined value for the font widths. The function GridRenderer.truncateToWidth that it calls does though, and already uses that constant for default values, so I made getCachedTruncatedString accept undefined for the font widths, and the constant can be removed from other places safely.

If you did happen to need the font widths for something else in those methods, you would have to add back that undefined check.

[dgodinez-dh]