zed_extension_api: Return structured slash command completions (#13879)

This PR updates the extension API to use structured slash command
completions instead of plain strings.

This allows slash commands defined in extensions to take advantage of
the improvements made in #13876.

Release Notes:

- N/A

extensions/gleam/src/gleam.rs

@@ -4,7 +4,7 @@ use std::fs;
 use zed::lsp::CompletionKind;
 use zed::{
     CodeLabel, CodeLabelSpan, HttpRequest, KeyValueStore, LanguageServerId, SlashCommand,
-    SlashCommandOutput, SlashCommandOutputSection,
+    SlashCommandArgumentCompletion, SlashCommandOutput, SlashCommandOutputSection,
 };
 use zed_extension_api::{self as zed, Result};
 
@@ -154,12 +154,24 @@ impl zed::Extension for GleamExtension {
         &self,
         command: SlashCommand,
         _query: String,
-    ) -> Result<Vec<String>, String> {
+    ) -> Result<Vec<SlashCommandArgumentCompletion>, String> {
         match command.name.as_str() {
             "gleam-project" => Ok(vec![
-                "apple".to_string(),
-                "banana".to_string(),
-                "cherry".to_string(),
+                SlashCommandArgumentCompletion {
+                    label: "apple".to_string(),
+                    new_text: "Apple".to_string(),
+                    run_command: false,
+                },
+                SlashCommandArgumentCompletion {
+                    label: "banana".to_string(),
+                    new_text: "Banana".to_string(),
+                    run_command: false,
+                },
+                SlashCommandArgumentCompletion {
+                    label: "cherry".to_string(),
+                    new_text: "Cherry".to_string(),
+                    run_command: true,
+                },
             ]),
             _ => Ok(Vec::new()),
         }