feat(clp-tdl-package): Add task signatures and shared I/O types for non-aggregation queries.

components/clp-rust-utils/src/job_config/search.rs

@@ -5,6 +5,8 @@ use serde::Serialize;
 
 pub const QUERY_JOBS_TABLE_NAME: &str = "query_jobs";
 
+pub type QueryJobId = i32;
+
 /// Mirror of `job_orchestration.scheduler.job_config.AggregationConfig`. Must be kept in sync.
 #[derive(Clone, Debug, Default, Deserialize, Eq, PartialEq, Serialize)]
 #[serde(default)]

components/clp-rust-utils/src/task_io.rs

@@ -1 +1,2 @@
 pub mod compression;
+pub mod query;

components/clp-rust-utils/src/task_io/query.rs

@@ -0,0 +1,34 @@
+//! Protocol types exchanged with the Spider (Huntsman) tasks that run CLP query jobs.
+
+use std::num::NonZeroU32;
+
+use non_empty_string::NonEmptyString;
+use serde::Deserialize;
+use serde::Serialize;
+
+/// `clp-s` options for a query job.
+#[derive(Clone, Debug, Deserialize, Eq, PartialEq, Serialize)]
+pub struct ClpSQueryOption {
+    /// The query string passed positionally to `clp-s`.
+    pub query_string: NonEmptyString,
+
+    /// The per-archive result limit. When absent, the task omits `--max-num-results` and uses the
+    /// `clp-s` default.
+    pub max_num_results: Option<NonZeroU32>,
+
+    /// Inclusive `--tge` bound in Unix epoch milliseconds.
+    pub begin_timestamp_millisecs: Option<i64>,
+
+    /// Inclusive `--tle` bound in Unix epoch milliseconds.
+    pub end_timestamp_millisecs: Option<i64>,
+
+    /// Whether `clp-s` performs a case-insensitive search.
+    pub ignore_case: bool,
+}
+
+/// The output handler that `clp-s` writes a query task's results to.
+#[derive(Clone, Debug, Deserialize, Eq, PartialEq, Serialize)]
+pub enum QueryOutputHandle {
+    /// Writes results to the results-cache collection named after the query job's ID.
+    ResultsCache,
+}

components/clp-tdl-package/src/lib.rs

@@ -1,4 +1,4 @@
-//! Spider TDL task package `clp`: the CLP compression tasks the Spider task executor loads.
+//! Spider TDL package `clp`, providing CLP compression and query tasks for Spider task executors.
 
 pub mod common;
 mod task;
@@ -28,5 +28,9 @@ fn package_init() -> Result<(), TdlError> {
 spider_tdl::register_tdl_package! {
     package_name: "clp",
     init: package_init,
-    tasks: [task::compression::s3_compress_task, task::compression::commit_task],
+    tasks: [
+        task::compression::s3_compress_task,
+        task::compression::commit_task,
+        task::query::clp_s_search_task,
+    ],
 }

components/clp-tdl-package/src/task/mod.rs

@@ -1,3 +1,4 @@
 //! The task implementations this package registers with Spider.
 
 pub mod compression;
+pub mod query;

components/clp-tdl-package/src/task/query/mod.rs

@@ -0,0 +1,20 @@
+//! The query-task signatures registered with Spider.
+
+use clp_rust_utils::job_config::QueryJobId;
+use clp_rust_utils::task_io::query::ClpSQueryOption;
+use clp_rust_utils::task_io::query::QueryOutputHandle;
+use non_empty_string::NonEmptyString;
+use spider_tdl::TaskContext;
+use spider_tdl::task;
+
+#[task(name = "query::clp_s_search")]
+pub(crate) fn clp_s_search_task(
+    _ctx: TaskContext,
+    _query_job_id: QueryJobId,
+    _clp_s_query_option: ClpSQueryOption,
+    _dataset: Option<NonEmptyString>,
+    _archive_id: NonEmptyString,
+    _query_output_handle: QueryOutputHandle,
+) -> Result<(), spider_tdl::TdlError> {
+    todo!("query task is not implemented")
+}