gitlore/src/cli/commands/embed.rs

use crate::cli::render::Theme;
use serde::Serialize;

use crate::Config;
use crate::cli::robot::RobotMeta;
use crate::core::db::{LATEST_SCHEMA_VERSION, create_connection, get_schema_version};
use crate::core::error::{LoreError, Result};
use crate::core::paths::get_db_path;
use crate::core::shutdown::ShutdownSignal;
use crate::embedding::ollama::{OllamaClient, OllamaConfig};
use crate::embedding::pipeline::{DEFAULT_EMBED_CONCURRENCY, embed_documents};

#[derive(Debug, Default, Serialize)]
pub struct EmbedCommandResult {
    pub docs_embedded: usize,
    pub chunks_embedded: usize,
    pub failed: usize,
    pub skipped: usize,
}

pub async fn run_embed(
    config: &Config,
    full: bool,
    retry_failed: bool,
    progress_callback: Option<Box<dyn Fn(usize, usize)>>,
    signal: &ShutdownSignal,
) -> Result<EmbedCommandResult> {
    let db_path = get_db_path(config.storage.db_path.as_deref());
    let conn = create_connection(&db_path)?;

    let schema_version = get_schema_version(&conn);
    if schema_version < LATEST_SCHEMA_VERSION {
        return Err(LoreError::MigrationFailed {
            version: schema_version,
            message: format!(
                "Database is at schema version {schema_version} but {LATEST_SCHEMA_VERSION} is required. \
                 Run 'lore migrate' first."
            ),
            source: None,
        });
    }

    let ollama_config = OllamaConfig {
        base_url: config.embedding.base_url.clone(),
        model: config.embedding.model.clone(),
        ..OllamaConfig::default()
    };
    let client = OllamaClient::new(ollama_config);

    client.health_check().await?;

    if full {
        conn.execute_batch(
            "BEGIN;
             DELETE FROM embedding_metadata;
             DELETE FROM embeddings;
             COMMIT;",
        )?;
    } else if retry_failed {
        // DELETE (not UPDATE) so the LEFT JOIN in find_pending_documents returns NULL,
        // making the doc appear pending again. UPDATE would leave a matching row that
        // still satisfies the config-param check, making the doc permanently invisible.
        conn.execute_batch(
            "BEGIN;
             DELETE FROM embeddings WHERE rowid / 1000 IN (
               SELECT DISTINCT document_id FROM embedding_metadata
               WHERE last_error IS NOT NULL
             );
             DELETE FROM embedding_metadata WHERE last_error IS NOT NULL;
             COMMIT;",
        )?;
    }

    let model_name = &config.embedding.model;
    let concurrency = if config.embedding.concurrency > 0 {
        config.embedding.concurrency as usize
    } else {
        DEFAULT_EMBED_CONCURRENCY
    };
    let result = embed_documents(
        &conn,
        &client,
        model_name,
        concurrency,
        progress_callback,
        signal,
    )
    .await?;

    Ok(EmbedCommandResult {
        docs_embedded: result.docs_embedded,
        chunks_embedded: result.chunks_embedded,
        failed: result.failed,
        skipped: result.skipped,
    })
}

pub fn print_embed(result: &EmbedCommandResult) {
    if result.docs_embedded == 0 && result.failed == 0 && result.skipped == 0 {
        println!(
            "\n  {} nothing to embed",
            Theme::success().bold().render("Embedding")
        );
        return;
    }

    println!(
        "\n  {} {} documents ({} chunks)",
        Theme::success().bold().render("Embedded"),
        Theme::bold().render(&result.docs_embedded.to_string()),
        result.chunks_embedded
    );
    if result.failed > 0 {
        println!(
            "  {}",
            Theme::error().render(&format!("{} failed", result.failed))
        );
    }
    if result.skipped > 0 {
        println!(
            "  {}",
            Theme::dim().render(&format!("{} skipped", result.skipped))
        );
    }
}

#[derive(Serialize)]
struct EmbedJsonOutput<'a> {
    ok: bool,
    data: &'a EmbedCommandResult,
    meta: RobotMeta,
}

pub fn print_embed_json(result: &EmbedCommandResult, elapsed_ms: u64) {
    let output = EmbedJsonOutput {
        ok: true,
        data: result,
        meta: RobotMeta { elapsed_ms },
    };
    println!("{}", serde_json::to_string(&output).unwrap());
}