feat: add language synonym mapping and canonicalization

Add LANGUAGE_SYNONYMS dictionary to map language aliases to canonical IDs,
_canonical_language_id() to normalize language identifiers, and
_language_guidance() to provide language-specific instructions for LLM
code generation. This improves language detection and ensures consistent
prompt context across different language format variations.

src/plugins/copilotPlugin.ts

@@ -18,7 +18,7 @@ interface CopilotState {
 }
 
 interface CopilotConfig {
-  fetchSuggestion: (prefix: string, suffix: string, signal?: AbortSignal) => Promise<string>
+  fetchSuggestion: (prefix: string, suffix: string, languageId: string, signal?: AbortSignal) => Promise<string>
   debounceMs?: number
 }
 
@@ -187,6 +187,52 @@ function normalizeSuggestionText(raw: string): string {
   return text
 }
 
+function sanitizeLanguageId(value: string): string {
+  if (!value) return ''
+  const trimmed = value.trim()
+  if (!trimmed) return ''
+  let safe = ''
+  for (const ch of trimmed) {
+    if ((ch >= 'a' && ch <= 'z') || (ch >= 'A' && ch <= 'Z') || (ch >= '0' && ch <= '9') || ch === '-' || ch === '_' || ch === '+' || ch === '.') {
+      safe += ch
+    }
+  }
+  return safe.slice(0, 32).toLowerCase()
+}
+
+function getCodeBlockLanguage(node: ProseNode): string {
+  const candidates = [node.attrs?.language, node.attrs?.lang, node.attrs?.info]
+  for (const value of candidates) {
+    if (typeof value === 'string' && value.trim()) {
+      return value.trim()
+    }
+  }
+  return ''
+}
+
+function getCursorLanguageId(view: EditorView): string {
+  const fallback = 'markdown'
+  const { $from } = view.state.selection
+
+  for (let depth = $from.depth; depth > 0; depth -= 1) {
+    const node = $from.node(depth)
+    const typeName = node.type?.name || ''
+
+    if (typeName === 'math_inline' || typeName === 'math_block' || typeName === 'latex' || typeName === 'math') {
+      return 'latex'
+    }
+
+    if (typeName === 'code_block' || typeName === 'codeBlock' || typeName === 'code_fence' || typeName === 'fence') {
+      const lang = sanitizeLanguageId(getCodeBlockLanguage(node))
+      if (!lang) return fallback
+      if (lang === 'tex' || lang === 'latex' || lang === 'katex') return 'latex'
+      return lang
+    }
+  }
+
+  return fallback
+}
+
 async function insertGhostText(view: EditorView, suggestion: string, from: number, ctx: Ctx) {
   if (!suggestion) return
 
@@ -303,7 +349,8 @@ function doFetchSuggestion(
   const controller = new AbortController()
   runtime.abortController = controller
 
-  config.fetchSuggestion(prefix, suffix, controller.signal)
+  const languageId = getCursorLanguageId(view)
+  config.fetchSuggestion(prefix, suffix, languageId, controller.signal)
     .then((suggestion) => {
       if (!runtime.enabled) return
       if (runtime.requestSeq !== requestSeq) return

src/utils/api.js

@@ -60,7 +60,17 @@ async function getClientIP() {
     }
 }
 
-export async function fetchSuggestion(prefix, suffix, signal, apiUrl = API_URL) {
+export async function fetchSuggestion(prefix, suffix, languageId, signal, apiUrl = API_URL) {
+    let normalizedLanguageId = 'markdown'
+    if (typeof languageId === 'string' && languageId.trim()) {
+        normalizedLanguageId = languageId.trim()
+    } else if (languageId && typeof languageId === 'object' && 'aborted' in languageId) {
+        signal = languageId
+    }
+    if (typeof signal === 'string') {
+        apiUrl = signal
+        signal = undefined
+    }
     const requestId = generateRequestId()
     const cancelUrl = getCancelUrl(apiUrl)
 
@@ -94,7 +104,7 @@ export async function fetchSuggestion(prefix, suffix, signal, apiUrl = API_URL)
         const body = {
             prefix,
             suffix,
-            languageId: 'markdown',
+            languageId: normalizedLanguageId,
             model_thinking: settings.modelThinking,
             privacy_mode: settings.privacyMode,
             user_preferences: {