From c4c8a0f985e27d1856a724b77c94b9b4dc8ab4ee Mon Sep 17 00:00:00 2001
From: Yang Ming-Tian <1178715749@qq.com>
Date: Wed, 27 Dec 2023 16:51:08 +0800
Subject: [PATCH] Add a REPL (#35)

* Upgrade Geni to v0.0.42

* Upgrade Clojure to v1.11.1

* Fix dependencies for REPL

* Implement Datajure REPL
---
 project.clj           | 18 +++++++++++++++++-
 src/datajure/dsl.clj  | 41 ++++++++++++++++++++++++++++++++++++++++-
 src/datajure/repl.clj | 31 +++++++++++++++++++++++++++++++
 3 files changed, 88 insertions(+), 2 deletions(-)
 create mode 100644 src/datajure/repl.clj

diff --git a/project.clj b/project.clj
index 7eb4aac..ff3ad00 100644
--- a/project.clj
+++ b/project.clj
@@ -21,8 +21,24 @@
              "--add-opens=java.base/java.util.concurrent=ALL-UNNAMED"
              "--add-opens=java.base/sun.nio.ch=ALL-UNNAMED"]
   :profiles {:uberjar {:aot :all
-                       :jvm-opts ["-Dclojure.compiler.direct-linking=true"]}
+                       :plugins [[arctype/log4j2-plugins-cache "1.0.0"]]
+                       :middleware [leiningen.log4j2-plugins-cache/middleware]
+                       :manifest {"Multi-Release" true}
+                       :dependencies [[log4j/log4j "1.2.17"]
+                                      [org.apache.logging.log4j/log4j-core "2.21.0"]
+                                      ;; Arrow
+                                      [org.apache.arrow/arrow-memory-core "4.0.0"]
+                                      [org.apache.arrow/arrow-vector "4.0.0"
+                                       :exclusions [commons-codec com.fasterxml.jackson.core/jackson-databind]]
+                                      ;; Spark
+                                      [org.apache.spark/spark-avro_2.12 "3.3.3"]
+                                      [org.apache.spark/spark-core_2.12 "3.3.3"]
+                                      [org.apache.spark/spark-hive_2.12 "3.3.3"]
+                                      [org.apache.spark/spark-mllib_2.12 "3.3.3"]
+                                      [org.apache.spark/spark-sql_2.12 "3.3.3"]
+                                      [org.apache.spark/spark-streaming_2.12 "3.3.3"]]}
              :test {:dependencies [[org.apache.logging.log4j/log4j-core "2.21.0"]]}
+             :repl {:dependencies [[org.apache.logging.log4j/log4j-core "2.21.0"]]}
              :provided {:dependencies [[com.fasterxml.jackson.core/jackson-core "2.15.3"]
                                        [com.fasterxml.jackson.core/jackson-annotations "2.15.3"]
                                        [org.apache.spark/spark-core_2.12 "3.3.3" :exclusions [org.apache.logging.log4j/log4j-slf4j-impl]]
diff --git a/src/datajure/dsl.clj b/src/datajure/dsl.clj
index b2e6e78..6633493 100644
--- a/src/datajure/dsl.clj
+++ b/src/datajure/dsl.clj
@@ -1,5 +1,7 @@
 (ns datajure.dsl
-  (:refer-clojure :exclude [print]))
+  (:refer-clojure :exclude [print]) 
+  (:require [datajure.repl :as repl])
+  (:gen-class))
 
 (require '[tech.v3.dataset :as ds]
          '[tablecloth.api :as tc]
@@ -130,3 +132,40 @@
   "Choose `back` as the backend implementation."
   [back]
   (reset! backend back))
+
+(def init-eval
+  "The initial form evaluated when the REPL starts up."
+  '(do
+     (require
+      '[tech.v3.dataset :as ds]
+      '[tablecloth.api :as tc]
+      '[clojask.dataframe :as ck]
+      '[zero-one.geni.core :as g]
+      '[datajure.dsl :as dtj])))
+
+(defn- custom-stream [script-path]
+  (try
+    (-> (str (slurp script-path) "\nexit\n")
+        (.getBytes "UTF-8")
+        (java.io.ByteArrayInputStream.))
+    (catch Exception _
+      (println (str "Cannot find file " script-path "!"))
+      (System/exit 1))))
+
+(defn -main
+  "The Datajure CLI entrypoint.
+
+  It does the following:
+  - Prints a welcome note.
+  - Launches an nREPL server, which writes to `.nrepl-port` for a
+    text editor to connect to.
+  - Starts a REPL(-y).
+  "
+  [& args]
+  (println repl/welcome-note)
+  (let [script-path (if (empty? args) nil (first args))]
+    (repl/launch-repl (merge {:port (+ 65001 (rand-int 500))
+                              :custom-eval init-eval}
+                             (when script-path
+                               {:input-stream (custom-stream script-path)}))))
+  (System/exit 0))
\ No newline at end of file
diff --git a/src/datajure/repl.clj b/src/datajure/repl.clj
new file mode 100644
index 0000000..887f9dd
--- /dev/null
+++ b/src/datajure/repl.clj
@@ -0,0 +1,31 @@
+(ns datajure.repl
+  (:require
+   [clojure.string]
+   [clojure.java.io :as io]
+   [nrepl.server]
+   [reply.main]))
+
+(defn- client [opts]
+  (let [port (:port opts)
+        host (or (:host opts) "127.0.0.1")
+        default-opts {:color true :history-file ".nrepl-history"}
+        opts (assoc (merge default-opts opts) :attach (str host ":" port))]
+    (reply.main/launch-nrepl opts)))
+
+(defn datajure-prompt
+  "Custom Datajure REPL prompt."
+  [ns-]
+  (str "datajure-repl (" ns- ")> "))
+
+(defn launch-repl
+  "Starts an nREPL server and steps into a REPL-y."
+  [opts]
+  (let [port   (:port opts)
+        server (nrepl.server/start-server :port port)]
+    (doto (io/file ".nrepl-port") .deleteOnExit (spit port))
+    (client (merge {:custom-prompt datajure-prompt} opts))
+    (nrepl.server/stop-server server)))
+
+(def welcome-note
+  "A REPL welcome note."
+  (str "Datajure " "1.0.1"))
\ No newline at end of file