Adventure score screen and conditional question polish

[clpetersonucf]

• Adds a custom score screen. Supports all node types and accurately displays inventory item changes and conditional questions (hopefully!)

• In support of the score screen, the following changes to the score module (and scoring generally) were made:

  • Adjusted the formatting of data sent to the score screen via the get_score_details method.
  • Previously, score logs were matched to the corresponding qset item by a log text comparison, which is prone to inaccuracy. Now, logs of type FINAL_SCORE_FROM_CLIENT will provide the end node's node id as the log's item_id. Older adventure score logs always have an item id of 0 for FINAL_SCORE_FROM_CLIENT log types.
  • Narrative node visits are now included in logs, which was not the case previously. This ensures proper tracking of inventory deltas.

• Adds partial matches option for Short Answer node types. Updates the short answer node UI to be slightly better formatted.

• Minor adjustments to conditional question UI. No changes were made to how conditional questions are actually selected.

• Condensed give/take item and required item dialogs into a single variant, instead of two duplicate UIs depending on layout.

• Significant rework of the "Add a node in between these two nodes" UI. The in-between button is always visible as a small dot on the midpoint of links where both the source and target are not blank. Additionally, these in-between nodes will shift position when too close to another node to prevent occlusion.

• A handful of miscellaneous bug fixes and styling tweaks.

Remaining todos:

• Adjust the advanced question editor button to be positioned adjacent to question text and media options. Did not implement
• Fix sizing of question text area in the node manager.
• Handle truncating of excessively long question and answer text in the score screen DNI

[FrenjaminBanklin]

The streamlining and score screen modifications seem to be working properly. Can't get anything to start misbehaving.

[clpetersonucf]

Short Answer updates:

• Special character sensitivity settings work again.
• Score screen now displays the actual response text the user submitted, instead of the matched value
• Updates to partial string detection: The set of potential matches for a particular response is now collected and ranked based on specificity. A match is selected based on the following: exact match (from an answer set that requires exact matches) > exact match (from an answer set that allows partial matches) > partial matches ranked by specificity (length of matched string)

[cayb0rg]

Just noticed we need to reset response to the original response at the start of each loop iteration. There's an edge case here where if a preceding answer doesn't require case sensitivity, then response will be lowercase forever, breaking the case sensitivity option for following answers. Same happens for character sensitivity.

[clpetersonucf]

Great catch, fixed!

[cayb0rg]

This is another error irrelevant to the PR's changes, but would be good to include. The following code is removing special characters in addition to whitespace, leaving only (a-z, A-Z), digits (0-9), and underscores (_), so the character sensitivity option doesn't work properly. We should probably use the replace() function instead with the \s metacharacter.

[clpetersonucf]

Your suggestion appears to do the trick, fixed.

[cayb0rg]

New changes look great, and partial match priority selection seems to work correctly. Apart from the issues with case/character sensitivity, this is looking good to me!

[clpetersonucf]

Updated again: Short Answer matches should be more robust based on fixing issues @cayb0rg identified.

[cayb0rg]

Case and special character sensitivity works great! Not spotting any other outstanding bugs, apart from what Matt found. Nice work 💪🏻

[FrenjaminBanklin]

All of the listed features appear to be working well.

[dmols]

Tested it before the new changes, then after. Widget works as intended and the changes look good. Great work!