From 33fab04c31ad5c287897aeb260a2db455b0be78b Mon Sep 17 00:00:00 2001
From: Stijn Peeters <stijn.peeters@uva.nl>
Date: Wed, 29 Sep 2021 15:48:43 +0200
Subject: [PATCH] Script that does first-time setup

---
 4cat-daemon.py              | 15 +++++++++++++
 docker/docker_setup.py      |  9 --------
 helper-scripts/first-run.py | 42 +++++++++++++++++++++++++++++++++++++
 webtool/__init__.py         | 15 +++++++++++++
 4 files changed, 72 insertions(+), 9 deletions(-)
 create mode 100644 helper-scripts/first-run.py

diff --git a/4cat-daemon.py b/4cat-daemon.py
index 44ea10c4f..4464f77a3 100644
--- a/4cat-daemon.py
+++ b/4cat-daemon.py
@@ -1,3 +1,4 @@
+import subprocess
 import argparse
 import time
 import sys
@@ -18,6 +19,20 @@
 cli.add_argument("command")
 args = cli.parse_args()
 
+# ---------------------------------------------
+#  first-run.py ensures everything is set up
+#  right when running 4CAT for the first time
+# ---------------------------------------------
+first_run = Path(__file__).parent.joinpath("helper-scripts", "first-run.py")
+result = subprocess.run([sys.executable, str(first_run)], stdout=subprocess.PIPE,
+                        stderr=subprocess.PIPE)
+
+if result.returncode != 0:
+    print("Unexpected error while preparing 4CAT. You may need to re-install 4CAT.")
+    print("stdout:\n".join(["  " + line for line in result.stdout.decode("utf-8").split("\n")]))
+    print("stderr:\n".join(["  " + line for line in result.stderr.decode("utf-8").split("\n")]))
+    exit(1)
+
 # ---------------------------------------------
 #     Do not start if migration is required
 # ---------------------------------------------
diff --git a/docker/docker_setup.py b/docker/docker_setup.py
index 91f8a0235..392cd4266 100644
--- a/docker/docker_setup.py
+++ b/docker/docker_setup.py
@@ -30,15 +30,6 @@
 
     # Install specific packages
     if docker_config['DOCKER'].getboolean('new_installation'):
-        # NLTK resources
-        import nltk
-        nltk.download("wordnet")
-        nltk.download("punkt")
-
-        # Spacy lexicon
-        import spacy
-        spacy.cli.download('en_core_web_sm')
-
         # Update config to skip this on future runs
         docker_config['DOCKER']['new_installation'] = 'False'
 
diff --git a/helper-scripts/first-run.py b/helper-scripts/first-run.py
new file mode 100644
index 000000000..ccf3d3dda
--- /dev/null
+++ b/helper-scripts/first-run.py
@@ -0,0 +1,42 @@
+"""
+Actions to execute the first time a 4CAT instance is run
+
+There are some things that should be done once, and only once, in the life of
+a 4CAT instance. These are set-up actions that cannot be handled by e.g.
+running pip or setup.py because they rely on things not available to those
+processes or that could interfere with other processes run at that time.
+
+This file is only fully  executed if the file .current-version does NOT exist
+in the 4CAT root folder. The file is created if it does not exist yet, ensuring
+the set-up actions in this file are only ever run once.
+
+It should not be necessary to run this file directly; it is run automatically
+by 4CAT while starting up.
+"""
+import shutil
+import sys
+from pathlib import Path
+
+# make sure version files are in order
+version_file = Path(__file__).resolve().parent.parent.joinpath("VERSION")
+if not version_file.exists():
+    # this file should ALWAYS exist, because it is part of the repository, and
+    # required by other parts of 4CAT. If it's absent, something has gone
+    # wrong, and the preferred course of action is restarting from scratch
+    print("VERSION file not found. You should re-install 4CAT before continuing.", file=sys.stderr)
+    exit(1)
+
+current_version_file = version_file.with_name(".current-version")
+if current_version_file.exists():
+    # this file does not exist by default, so if it does, that means we don't
+    # need to do further on-boarding since 4CAT has already been run
+    exit(0)
+
+shutil.copy(version_file, current_version_file)
+
+# Now check for presence of required NLTK packages
+import nltk
+nltk_downloads = ("wordnet", "punkt")
+for package in nltk_downloads:
+    # if it already exists, .download() will just NOP
+    nltk.download(package)
\ No newline at end of file
diff --git a/webtool/__init__.py b/webtool/__init__.py
index b6b4ebdbb..deaa4da54 100644
--- a/webtool/__init__.py
+++ b/webtool/__init__.py
@@ -1,3 +1,18 @@
+import subprocess
+import sys
+from pathlib import Path
+
+# first-run.py ensures everything is set up right when running 4CAT for the first time
+first_run = Path(__file__).parent.parent.joinpath("helper-scripts", "first-run.py")
+result = subprocess.run([sys.executable, str(first_run)], stdout=subprocess.PIPE,
+                        stderr=subprocess.PIPE)
+
+if result.returncode != 0:
+    print("Unexpected error while preparing 4CAT. You may need to re-install 4CAT.")
+    print("stdout:\n".join(["  " + line for line in result.stdout.decode("utf-8").split("\n")]))
+    print("stderr:\n".join(["  " + line for line in result.stderr.decode("utf-8").split("\n")]))
+    exit(1)
+
 from flask import Flask
 from flask_login import LoginManager
 from flask_limiter import Limiter