fix(code-editor): trigger completion for unicode identifiers

[aofei]

Allow completion to start when the typed character is a Go identifier letter or decimal digit outside ASCII. This lets resource names such as Chinese sprite names request suggestions after IME input.

[Copilot]

Pull request overview

Updates the XGo code editor’s completion trigger logic to recognize non-ASCII Go identifier characters, improving completion behavior after IME input (e.g., Chinese sprite/resource names).

Changes:

• Expand shouldTriggerCompletion from an ASCII-focused \w regex to a Unicode-aware regex using \p{L} / \p{Nd} with the u flag.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[fennoai]

The fix correctly addresses the stated goal — Unicode letters (e.g. Chinese sprite names) now trigger completion after IME input. A few observations:

Regex not hoisted to module scope
The regex literal inside shouldTriggerCompletion is re-evaluated on every call. Unicode property escapes (\p{L}) involve constructing internal Unicode character-category tables, which is more expensive than the ASCII \w shorthand. Hoisting the regex to module scope ensures it is compiled exactly once.

wordPattern not configured for Unicode
shouldTriggerCompletion now returns true for Chinese characters, but the subsequent call to textDocument.getWordAtPosition(position) (line 43) falls back to Monaco's default word pattern, which only recognises ASCII \w characters. As a result, when the cursor is after a Unicode identifier, wordStart will equal the current cursor position rather than the start of the identifier — completion is triggered, but the filter prefix and replacement range will be wrong. A custom wordPattern: /[\p{L}\p{Nd}_]+/u (or equivalent) should be added to xgoLanguageConfiguration to fully realise this fix.

\p{Nd} over-matches non-ASCII decimal digits
Go identifiers only permit ASCII decimal digits (0-9); the Go spec uses unicode_letter | unicode_digit but decimal_digit = "0" … "9". \p{Nd} also matches Arabic-Indic digits (٠١٢…), Devanagari digits (०१२…), etc. Using [0-9] instead of \p{Nd} would be a more faithful representation of the Go spec, though the current over-match is harmless (it triggers completion on characters that would be invalid in a Go identifier).

[gemini-code-assist]

Code Review

This pull request updates the shouldTriggerCompletion function in the code editor's completion logic to use a Unicode-aware regular expression. This change ensures that completion is correctly triggered for non-ASCII letters and digits, improving support for international characters. I have no feedback to provide.