AstroResearch/src/services/observation/cache.rs

// src/services/observation/cache.rs
//
// 观测数据缓存层 + 落盘辅助
//
// observation_cache 表设计：
//   - UNIQUE(source, product, source_id)：同一源的同一标识符可同时缓存不同产品
//     （如 Gaia 同一 source_id 的光谱与光变分别独立缓存）
//   - artifacts_json：一个逻辑产物可对应多个物理文件（Gaia 光变 G/BP/RP 三波段）

use crate::services::observation::types::{Artifact, ProductSpec, Source};
use anyhow::{anyhow, Result};
use serde::{Deserialize, Serialize};
use sqlx::SqlitePool;
use std::path::Path;
use tracing::error;

// ═══════════════════════════════════════════════════════════════
// 1. 缓存命中行类型
// ═══════════════════════════════════════════════════════════════

/// observation_cache 中单条 artifacts 的 JSON 结构
#[derive(Debug, Clone, Serialize, Deserialize)]
struct ArtifactRecord {
    path: String,
    format: String,
    #[serde(default)]
    size: usize,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    band: Option<String>,
    #[serde(default, skip_serializing_if = "Option::is_none")]
    original_name: Option<String>,
}

impl From<&ArtifactRecord> for Artifact {
    fn from(r: &ArtifactRecord) -> Self {
        Artifact {
            band: r.band.clone(),
            original_name: r.original_name.clone(),
            file_path: r.path.clone(),
            file_url: crate::services::observation::cache::file_url_from_path(&r.path),
            file_format: r.format.clone(),
            size_bytes: r.size,
            cached: true,
        }
    }
}

/// 缓存查询结果
#[derive(sqlx::FromRow)]
struct CacheHitRow {
    artifacts_json: String,
    meta_json: Option<String>,
}

/// observation_cache 列表查询行（对外暴露）
#[derive(Debug, Clone, Serialize, sqlx::FromRow)]
pub struct ObservationCacheRow {
    pub source: String,
    pub product: String,
    pub source_id: String,
    pub ra: Option<f64>,
    pub dec: Option<f64>,
    pub artifacts_json: String,
    pub meta_json: Option<String>,
    pub created_at: String,
}

// ═══════════════════════════════════════════════════════════════
// 2. 缓存读写
// ═══════════════════════════════════════════════════════════════

/// 查询单条缓存（按 source + product + source_id），返回命中的 artifacts
pub async fn fetch_observation_cache(
    pool: &SqlitePool,
    source: Source,
    product: &ProductSpec,
    source_id: &str,
) -> Result<Option<(Vec<Artifact>, Option<serde_json::Value>)>> {
    let row = sqlx::query_as::<_, CacheHitRow>(
        "SELECT artifacts_json, meta_json FROM observation_cache \
         WHERE source = ? AND product = ? AND source_id = ? LIMIT 1",
    )
    .bind(source.as_str())
    .bind(product.product.as_str())
    .bind(source_id)
    .fetch_optional(pool)
    .await
    .map_err(|e| anyhow!("查询 observation_cache 失败: {}", e))?;

    match row {
        Some(r) => {
            let records: Vec<ArtifactRecord> = serde_json::from_str(&r.artifacts_json)
                .map_err(|e| anyhow!("反序列化 artifacts_json 失败: {}", e))?;
            let artifacts: Vec<Artifact> = records.iter().map(Artifact::from).collect();
            let meta = r
                .meta_json
                .as_deref()
                .and_then(|s| serde_json::from_str(s).ok());
            Ok(Some((artifacts, meta)))
        }
        None => Ok(None),
    }
}

/// 写入/更新缓存条目
pub async fn write_observation_cache(
    pool: &SqlitePool,
    source: Source,
    product: &ProductSpec,
    source_id: &str,
    ra: Option<f64>,
    dec: Option<f64>,
    artifacts: &[Artifact],
    meta_json: Option<&str>,
) -> Result<()> {
    let records: Vec<ArtifactRecord> = artifacts
        .iter()
        .map(|a| ArtifactRecord {
            path: a.file_path.clone(),
            format: a.file_format.clone(),
            size: a.size_bytes,
            band: a.band.clone(),
            original_name: a.original_name.clone(),
        })
        .collect();
    let artifacts_json = serde_json::to_string(&records)?;
    sqlx::query(
        "INSERT OR REPLACE INTO observation_cache \
         (source, product, source_id, ra, dec, artifacts_json, meta_json, created_at) \
         VALUES (?, ?, ?, ?, ?, ?, ?, CURRENT_TIMESTAMP)",
    )
    .bind(source.as_str())
    .bind(product.product.as_str())
    .bind(source_id)
    .bind(ra)
    .bind(dec)
    .bind(&artifacts_json)
    .bind(meta_json)
    .execute(pool)
    .await
    .map_err(|e| anyhow!("写入 observation_cache 失败: {}", e))?;
    Ok(())
}

/// 列出某源某产品的全部已缓存产物
pub async fn list_cached(
    pool: &SqlitePool,
    source: Source,
    product: &ProductSpec,
) -> Result<Vec<ObservationCacheRow>> {
    let rows = sqlx::query_as::<_, ObservationCacheRow>(
        "SELECT source, product, source_id, ra, dec, artifacts_json, meta_json, created_at \
         FROM observation_cache WHERE source = ? AND product = ? \
         ORDER BY created_at DESC LIMIT 500",
    )
    .bind(source.as_str())
    .bind(product.product.as_str())
    .fetch_all(pool)
    .await
    .map_err(|e| anyhow!("查询 observation_cache 列表失败: {}", e))?;
    Ok(rows)
}

/// 列出全部已缓存产物（不限源/产品）
pub async fn list_all_cached(pool: &SqlitePool) -> Result<Vec<ObservationCacheRow>> {
    let rows = sqlx::query_as::<_, ObservationCacheRow>(
        "SELECT source, product, source_id, ra, dec, artifacts_json, meta_json, created_at \
         FROM observation_cache ORDER BY created_at DESC LIMIT 500",
    )
    .fetch_all(pool)
    .await
    .map_err(|e| anyhow!("查询 observation_cache 列表失败: {}", e))?;
    Ok(rows)
}

/// 缓存库列表筛选 + 排序参数（服务端分页）
#[derive(Debug, Clone, Default)]
pub struct ListCachedParams {
    /// 按数据源过滤（lamost/gaia/sdss/desi）
    pub source: Option<String>,
    /// 按产品类型过滤（spectrum/lightcurve/photometry/image）
    pub product: Option<String>,
    /// source_id 模糊匹配（LIKE %q%）
    pub q: Option<String>,
    /// 排序：created（默认，DESC）/ source / product
    pub sort: Option<String>,
    /// 每页条数（默认 50，上限 200）
    pub limit: Option<i64>,
    /// 偏移量
    pub offset: Option<i64>,
}

/// 缓存库分页查询结果
#[derive(Debug, Clone, Serialize)]
pub struct CachedPage {
    pub items: Vec<ObservationCacheRow>,
    pub total: i64,
}

/// 列出已缓存产物（服务端分页 + 筛选 + 排序）
pub async fn list_cached_paged(pool: &SqlitePool, params: &ListCachedParams) -> Result<CachedPage> {
    let limit = params.limit.unwrap_or(50).clamp(1, 200);
    let offset = params.offset.unwrap_or(0).max(0);

    // 动态构造 SQL（参数绑定防注入；列名/排序方向用白名单）
    let mut where_clauses: Vec<String> = Vec::new();
    if let Some(s) = &params.source {
        where_clauses.push(format!("source = '{}'", s.replace('\'', "''")));
    }
    if let Some(p) = &params.product {
        where_clauses.push(format!("product = '{}'", p.replace('\'', "''")));
    }
    if let Some(q) = &params.q {
        let q = q.replace('\'', "''");
        where_clauses.push(format!("source_id LIKE '%{}%'", q));
    }
    let where_sql = if where_clauses.is_empty() {
        String::new()
    } else {
        format!(" WHERE {}", where_clauses.join(" AND "))
    };

    let order_sql = match params.sort.as_deref() {
        Some("source") => " ORDER BY source ASC, created_at DESC",
        Some("product") => " ORDER BY product ASC, created_at DESC",
        _ => " ORDER BY created_at DESC",
    };

    // count
    let count_sql = format!("SELECT COUNT(*) as cnt FROM observation_cache{}", where_sql);
    let total: i64 = sqlx::query_scalar(&count_sql)
        .fetch_one(pool)
        .await
        .map_err(|e| anyhow!("统计 observation_cache 总数失败: {}", e))?;

    // items
    let items_sql = format!(
        "SELECT source, product, source_id, ra, dec, artifacts_json, meta_json, created_at \
         FROM observation_cache{}{} LIMIT ? OFFSET ?",
        where_sql, order_sql
    );
    let items = sqlx::query_as::<_, ObservationCacheRow>(&items_sql)
        .bind(limit)
        .bind(offset)
        .fetch_all(pool)
        .await
        .map_err(|e| anyhow!("查询 observation_cache 分页列表失败: {}", e))?;

    Ok(CachedPage { items, total })
}

// ═══════════════════════════════════════════════════════════════
// 3. 落盘 + URL 辅助
// ═══════════════════════════════════════════════════════════════

/// 经 /api/files 前缀生成访问 URL
pub fn file_url_from_path(rel_path: &str) -> String {
    format!("/api/files/{}", rel_path)
}

/// 落盘字节到 library_dir/<rel_path>，返回字节数
pub fn persist_bytes(library_dir: &Path, rel_path: &str, bytes: &[u8]) -> Result<usize> {
    let abs_path = library_dir.join(rel_path);
    if let Some(parent) = abs_path.parent() {
        std::fs::create_dir_all(parent)
            .map_err(|e| anyhow!("创建观测数据目录失败 {:?}: {}", parent, e))?;
    }
    std::fs::write(&abs_path, bytes)
        .map_err(|e| anyhow!("写入观测数据文件失败 {:?}: {}", abs_path, e))?;
    Ok(bytes.len())
}

/// 校验缓存产物中至少有一个文件真实存在；存在则返回其总大小
pub fn cached_files_total_size(library_dir: &Path, artifacts: &[Artifact]) -> Option<usize> {
    let mut total = 0;
    for a in artifacts {
        let abs_path = library_dir.join(&a.file_path);
        match std::fs::metadata(&abs_path) {
            Ok(m) => total += m.len() as usize,
            Err(_) => {
                // 任一文件缺失即视为缓存不完整
                return None;
            }
        }
    }
    if total == 0 {
        None
    } else {
        Some(total)
    }
}

/// 记录缓存写入失败但不中断流程
pub fn log_write_failure(source: &str, product: &str, e: anyhow::Error) {
    error!("[Observation] 写入 observation_cache 失败 ({}/{}): {}", source, product, e);
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_file_url_from_path() {
        assert_eq!(
            file_url_from_path("Telescope/lamost/spectrum/123.fits"),
            "/api/files/Telescope/lamost/spectrum/123.fits"
        );
    }

    #[test]
    fn test_artifact_record_roundtrip() {
        let rec = ArtifactRecord {
            path: "Telescope/gaia/lightcurve/123/123_G.fits".into(),
            format: "fits".into(),
            size: 1024,
            band: Some("G".into()),
            original_name: Some("123_EPOCH_PHOTOMETRY_G.fits".into()),
        };
        let json = serde_json::to_string(&rec).unwrap();
        let back: ArtifactRecord = serde_json::from_str(&json).unwrap();
        assert_eq!(back.path, rec.path);
        assert_eq!(back.band.as_deref(), Some("G"));
    }
}