fix(table): correctly truncate escaped cells

[Broderick-Westrope]

Fixes #502.

This pull request introduces new utility functions for handling terminal escape sequences and updates the table rendering logic to use these utilities. It also includes comprehensive unit tests for the new functions. Much of the new utility functions is derived from here (I made the assumption that this project is aiming for minimal dependencies).

New Utility Functions for Escape Sequences:

• table/escape_utils.go: Added functions to handle terminal escape sequences, including truncate, lengthWithoutEscapeSequences, extractEscapeSequences, and applyEscapeSequences.

Tests for New Utility Functions:

• table/escape_utils_test.go: Added tests for truncate, lengthWithoutEscapeSequences, extractEscapeSequences, and applyEscapeSequences to ensure they work correctly with various types of strings, including those with escape sequences, emojis, and combining characters.

Updates to Table Rendering Logic:

• table/table.go: Replaced runewidth.Truncate with the new truncate function in the headersView and renderRow methods to properly handle escape sequences. [1] [2]

Assumptions:

• When truncating a string an ellipsis (...) is used. When truncating, any escape sequences from the cut-off portion of the string are instead placed at the end of the truncated string. Closing these is import. The assumption I've made is that when the ellipsis is appended to the truncated string it will not be placed within these closing ellipsis. The consequences this are:
  • If for example the original string was to be colored red. The truncated string will still be colored red however the ellipsis will be the default text color (likely either white or black depending on terminal theme).
  • If the user has used a nonsensical combination of escape sequences which prevents the terminal from rendering the string, the ellipsis will still be shown. This is only true if the user included closing sequences (which is the users' responsibility).

[Broderick-Westrope]

May be redundant due to #246
Needs further investigation.

[bashbunni]

Hey @Broderick-Westrope, thank you so much for your help in improving bubbles. This functionality already exists in x/ansi, but can probably be improved with the additional test cases you wrote. If you're interested in contributing these tests over there, that would be nice! I do think that text truncation shouldn't be in the scope of the bubbles library. #502 should be fixed when we migrate the bubbles table to use the Lip Gloss table for rendering. If it's not, then we would likely want to fix that in Lip Gloss with x/ansi

[Broderick-Westrope]

Thanks @bashbunni, I'll raise a new PR with some test cases in the ANSI package. I'd be happy to help with the migration towards using lipgloss table, if you need the extra hands.

EDIT: After looking at the existing x/ansi tests I don't think my tests would add any value.