feat: integrate Google semantic search and branded desktop launcher

crates/searchi-cli/src/main.rs

@@ -5,6 +5,7 @@ use std::path::PathBuf;
 use anyhow::{bail, Context, Result};
 use clap::{Parser, Subcommand};
 use searchi_core::{IndexOptions, SearchIndex, SearchOptions};
+use searchi_service::HybridSearch;
 
 #[derive(Debug, Parser)]
 #[command(name = "searchi", version, about = "Search your local files offline")]
@@ -15,6 +16,12 @@ struct Cli {
 
 #[derive(Debug, Subcommand)]
 enum Command {
+    /// Set up the required Google model in the default local cache.
+    Setup {
+        /// Override the platform-local model cache directory.
+        #[arg(long)]
+        model_dir: Option<PathBuf>,
+    },
     /// Discover files and refresh an index for one folder.
     Index {
         /// SQLite index file; parent directory must exist.
@@ -54,6 +61,25 @@ enum Command {
         modified_after_ms: Option<i64>,
         #[arg(long)]
         modified_before_ms: Option<i64>,
+        /// Enable hybrid search with explicitly provisioned local model assets.
+        #[arg(long, requires = "embedding_worker")]
+        model_dir: Option<PathBuf>,
+        /// Local native Google embedding worker, required with --model-dir.
+        #[arg(long, requires = "model_dir")]
+        embedding_worker: Option<PathBuf>,
+        #[arg(long, default_value_t = 1, value_parser = clap::value_parser!(u8).range(1..=8))]
+        cpu_threads: u8,
+    },
+    /// Build a persistent semantic generation from already indexed passages.
+    Embed {
+        #[arg(long)]
+        db: PathBuf,
+        #[arg(long)]
+        model_dir: PathBuf,
+        #[arg(long)]
+        embedding_worker: PathBuf,
+        #[arg(long, default_value_t = 1, value_parser = clap::value_parser!(u8).range(1..=8))]
+        cpu_threads: u8,
     },
     /// Show the indexed file and content counts.
     Stats {
@@ -66,6 +92,15 @@ enum Command {
 
 fn main() -> Result<()> {
     match Cli::parse().command {
+        Command::Setup { model_dir } => {
+            let directory = model_dir
+                .map(Ok)
+                .unwrap_or_else(searchi_assets::default_model_directory)?;
+            let progress = std::sync::atomic::AtomicU64::new(0);
+            eprintln!("Setting up Google EmbeddingGemma 2; cached assets are reused (865 MB maximum download).");
+            searchi_assets::provision(&directory, &progress)?;
+            println!("Google model ready: {}", directory.display());
+        }
         Command::Index { db, root, exclude } => {
             let mut options = IndexOptions::default();
             options.excluded_directory_names.extend(exclude);
@@ -74,6 +109,9 @@ fn main() -> Result<()> {
             let report = index
                 .index_folder_with_options(&root, &options)
                 .with_context(|| format!("Cannot index folder {}", root.display()))?;
+            index
+                .prewarm()
+                .context("Cannot prepare search statistics")?;
             println!(
                 "Discovered: {}  Indexed: {}  Unchanged: {}  Removed: {}  Skipped: {}  Extraction failures: {}",
                 report.discovered, report.indexed, report.unchanged, report.removed, report.skipped, report.failures
@@ -90,27 +128,32 @@ fn main() -> Result<()> {
             max_size,
             modified_after_ms,
             modified_before_ms,
+            model_dir,
+            embedding_worker,
+            cpu_threads,
         } => {
             if query.trim().is_empty() {
                 bail!("Search text cannot be empty");
             }
             require_existing_index(&db)?;
-            let index = SearchIndex::open(&db)
-                .with_context(|| format!("Cannot open index {}", db.display()))?;
-            let results = index
-                .search_with_options(
-                    &query,
-                    usize::from(limit),
-                    &SearchOptions {
-                        folder,
-                        extension,
-                        min_size,
-                        max_size,
-                        modified_after_unix_ms: modified_after_ms,
-                        modified_before_unix_ms: modified_before_ms,
-                    },
-                )
-                .context("Search failed")?;
+            let options = SearchOptions {
+                folder,
+                extension,
+                min_size,
+                max_size,
+                modified_after_unix_ms: modified_after_ms,
+                modified_before_unix_ms: modified_before_ms,
+            };
+            let results = if let (Some(model), Some(runtime)) = (model_dir, embedding_worker) {
+                HybridSearch::open(&db, model, runtime, usize::from(cpu_threads))?
+                    .search(&query, usize::from(limit), &options)
+                    .context("Hybrid search failed")?
+            } else {
+                SearchIndex::open(&db)
+                    .with_context(|| format!("Cannot open index {}", db.display()))?
+                    .search_with_options(&query, usize::from(limit), &options)
+                    .context("Search failed")?
+            };
             if json {
                 println!("{}", serde_json::to_string_pretty(&results)?);
             } else if results.is_empty() {
@@ -130,6 +173,19 @@ fn main() -> Result<()> {
                 }
             }
         }
+        Command::Embed {
+            db,
+            model_dir,
+            embedding_worker,
+            cpu_threads,
+        } => {
+            require_existing_index(&db)?;
+            let report =
+                HybridSearch::open(&db, model_dir, embedding_worker, usize::from(cpu_threads))?
+                    .index_embeddings()
+                    .context("Semantic indexing failed; prior generation remains active")?;
+            println!("{}", serde_json::to_string_pretty(&report)?);
+        }
         Command::Stats { db, json } => {
             require_existing_index(&db)?;
             let index = SearchIndex::open(&db)
@@ -173,6 +229,30 @@ mod tests {
         }
     }
 
+    #[test]
+    fn requires_complete_semantic_configuration_and_bounded_cpu_budget() {
+        assert!(Cli::try_parse_from([
+            "searchi",
+            "search",
+            "--db",
+            "index.db",
+            "query",
+            "--model-dir",
+            "models"
+        ])
+        .is_err());
+        assert!(Cli::try_parse_from([
+            "searchi",
+            "search",
+            "--db",
+            "index.db",
+            "query",
+            "--cpu-threads",
+            "0"
+        ])
+        .is_err());
+    }
+
     #[test]
     fn supports_json_search_and_quoted_query() {
         let cli = Cli::try_parse_from([