reorganization
This commit is contained in:
parent
0f1088a066
commit
3a9ba2f320
@ -20,7 +20,7 @@ BASE_DIR = Path(__file__).resolve().parent.parent
|
||||
# See https://docs.djangoproject.com/en/5.1/howto/deployment/checklist/
|
||||
|
||||
# SECURITY WARNING: keep the secret key used in production secret!
|
||||
SECRET_KEY = 'django-insecure-k_$r^3e!9ycqnt0=+ur&sx#hsl44_+v3=al2$_gpnh3u^w!xj$'
|
||||
SECRET_KEY = 'django-insecure-m!o^q^+en&_v%64&m8%d^%_olkzf7$8jbp0^4dph2=1rn=666m'
|
||||
|
||||
# SECURITY WARNING: don't run with debug turned on in production!
|
||||
DEBUG = True
|
||||
@ -31,7 +31,6 @@ ALLOWED_HOSTS = []
|
||||
# Application definition
|
||||
|
||||
INSTALLED_APPS = [
|
||||
#'rest_framework',
|
||||
'sample_app',
|
||||
'django.contrib.admin',
|
||||
'django.contrib.auth',
|
||||
|
||||
@ -15,10 +15,8 @@ Including another URLconf
|
||||
2. Add a URL to urlpatterns: path('blog/', include('blog.urls'))
|
||||
"""
|
||||
from django.contrib import admin
|
||||
from django.urls import path, include
|
||||
|
||||
from django.urls import path
|
||||
|
||||
urlpatterns = [
|
||||
path('admin/', admin.site.urls),
|
||||
path('api/', include('sample_app.urls')), # Include your app's URLs
|
||||
]
|
||||
|
||||
@ -4,36 +4,23 @@ import glob
|
||||
import uuid
|
||||
import asyncio
|
||||
from datetime import datetime, timedelta
|
||||
|
||||
import pandas as pd
|
||||
import django
|
||||
from asgiref.sync import sync_to_async
|
||||
|
||||
import pandas as pd
|
||||
|
||||
|
||||
|
||||
# #environment init for django
|
||||
# current_dir = os.getcwd()
|
||||
# relative_path = os.path.join(current_dir, 'gaia_orm')
|
||||
# os.environ.setdefault('DJANGO_SETTINGS_MODULE', 'gaia_orm.settings')
|
||||
|
||||
# sys.path.append(os.path.normpath(relative_path))
|
||||
# #sys.path.append('/home/kityr/practiceproject/gaia_orm')
|
||||
# django.setup()
|
||||
|
||||
|
||||
|
||||
#import models for both the sources and the files
|
||||
from django.core.management.base import BaseCommand
|
||||
from sample_app.models import GaiaSource, CatalogFile
|
||||
|
||||
|
||||
class Command(BaseCommand):
|
||||
help = 'Ingest CSV files into the database'
|
||||
|
||||
def handle(self, *args, **options):
|
||||
|
||||
#fetching the file list
|
||||
directory = input("Please enter the path to the directory containing the csv files (or leave empty if the files are in the same directory as this script): ")
|
||||
csv_files = glob.glob(os.path.join(directory, '*csv*'))
|
||||
print(f"Files found: {len(csv_files)}.")
|
||||
print("Populating the file database...")
|
||||
self.stdout.write(f"Files found: {len(csv_files)}.")
|
||||
self.stdout.write("Populating the file database...")
|
||||
#initialize the counter
|
||||
new_files_count = 0
|
||||
#add files as catalogfile instances into the database
|
||||
@ -48,7 +35,7 @@ for file_path in csv_files:
|
||||
if created:
|
||||
new_files_count += 1
|
||||
#show how many duplicates were already in db
|
||||
print(f"File database populated. {len(csv_files) - new_files_count} were already in the database.")
|
||||
self.stdout.write(f"File database populated. {len(csv_files) - new_files_count} were already in the database.")
|
||||
|
||||
|
||||
|
||||
@ -81,7 +68,7 @@ def count_ingested_files():
|
||||
|
||||
|
||||
current_time = (datetime.now() + timedelta(hours=3)).strftime("%H:%M:%S")
|
||||
print(f"[{current_time}] Starting the data ingestion.")
|
||||
self.stdout.write(f"[{current_time}] Starting the data ingestion.")
|
||||
|
||||
#function that iterates over all catalog files and ingests sources from them
|
||||
async def ingest_files():
|
||||
@ -93,17 +80,18 @@ async def ingest_files():
|
||||
for catalog_file in catalog_files:
|
||||
|
||||
if catalog_file.status == 'INGESTED':
|
||||
print(f"Skipping {catalog_file.name} as it is already ingested.")
|
||||
self.stdout.write(f"Skipping {catalog_file.name} as it is already ingested.")
|
||||
continue
|
||||
|
||||
file_path = os.path.join(directory, catalog_file.name)
|
||||
|
||||
#print(file_path)
|
||||
#self.stdout.write(file_path)
|
||||
|
||||
if os.path.exists(file_path): #check if the file exists at all just in case
|
||||
pass
|
||||
|
||||
if catalog_file.status == 'IN_PROGRESS':
|
||||
print(f"{catalog_file.name} seems to have been interrupted, starting over.")#if the file status is in_progress, it must've
|
||||
self.stdout.write(f"{catalog_file.name} seems to have been interrupted, starting over.")#if the file status is in_progress, it must've
|
||||
await delete_gaiasources_for_catalogfile(catalog_file) #been interrupted mid-ingest, delete all the
|
||||
#associated sources if there are any and start over
|
||||
|
||||
@ -136,8 +124,8 @@ async def ingest_files():
|
||||
|
||||
ingested_files_count = await count_ingested_files()
|
||||
|
||||
current_time = (datetime.now() + timedelta(hours=3)).strftime("%H:%M:%S") #Timestamp and progress print statement
|
||||
print(f"[{current_time}] {ingested_files_count}/{len(catalog_files)}")
|
||||
current_time = (datetime.now() + timedelta(hours=3)).strftime("%H:%M:%S") #Timestamp and progress self.stdout.write statement
|
||||
self.stdout.write(f"[{current_time}] {ingested_files_count}/{len(catalog_files)}")
|
||||
|
||||
|
||||
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
# Generated by Django 5.1.1 on 2024-09-09 12:01
|
||||
# Generated by Django 5.1.1 on 2024-09-10 11:15
|
||||
|
||||
import django.db.models.deletion
|
||||
import uuid
|
||||
@ -18,7 +18,7 @@ class Migration(migrations.Migration):
|
||||
fields=[
|
||||
('uuid', models.UUIDField(default=uuid.uuid4, editable=False, primary_key=True, serialize=False)),
|
||||
('name', models.CharField(max_length=32)),
|
||||
('status', models.CharField(choices=[('PENDING', 'Pending'), ('INGESTED', 'Ingested'), ('INDEXED', 'Indexed')], default='PENDING', max_length=10)),
|
||||
('status', models.CharField(choices=[('PENDING', 'Pending'), ('IN_PROGRESS', 'In Progress'), ('INGESTED', 'Ingested'), ('INDEXED', 'Indexed')], default='PENDING', max_length=11)),
|
||||
],
|
||||
),
|
||||
migrations.CreateModel(
|
||||
|
||||
@ -1,18 +0,0 @@
|
||||
# Generated by Django 5.1.1 on 2024-09-09 14:26
|
||||
|
||||
from django.db import migrations, models
|
||||
|
||||
|
||||
class Migration(migrations.Migration):
|
||||
|
||||
dependencies = [
|
||||
('sample_app', '0001_initial'),
|
||||
]
|
||||
|
||||
operations = [
|
||||
migrations.AlterField(
|
||||
model_name='catalogfile',
|
||||
name='status',
|
||||
field=models.CharField(choices=[('PENDING', 'Pending'), ('IN_PROGRESS', 'In Progress'), ('INGESTED', 'Ingested'), ('INDEXED', 'Indexed')], default='PENDING', max_length=11),
|
||||
),
|
||||
]
|
||||
@ -1,7 +0,0 @@
|
||||
from rest_framework import serializers
|
||||
from .models import GaiaSource
|
||||
|
||||
class GaiaSourceSerializer(serializers.ModelSerializer):
|
||||
class Meta:
|
||||
model = GaiaSource
|
||||
fields = '__all__'
|
||||
@ -1,7 +1 @@
|
||||
from django.urls import path
|
||||
from .views import GaiaSourceListCreate, GaiaSourceDetail
|
||||
|
||||
urlpatterns = [
|
||||
path('GaiaSource/', GaiaSourceListCreate.as_view(), name='GaiaSource-list-create'),
|
||||
path('GaiaSource/<int:pk>/', GaiaSourceDetail.as_view(), name='GaiaSource-detail'),
|
||||
]
|
||||
|
||||
@ -1,13 +1 @@
|
||||
from django.shortcuts import render
|
||||
|
||||
from rest_framework import generics
|
||||
from .models import GaiaSource
|
||||
from .serializers import GaiaSourceSerializer
|
||||
|
||||
class GaiaSourceListCreate(generics.ListCreateAPIView):
|
||||
queryset = GaiaSource.objects.all()
|
||||
serializer_class = GaiaSourceSerializer
|
||||
|
||||
class GaiaSourceDetail(generics.RetrieveUpdateDestroyAPIView):
|
||||
queryset = GaiaSource.objects.all()
|
||||
serializer_class = GaiaSourceSerializer
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user