Files
DH/project_notebook.ipynb
2023-04-12 01:00:18 +03:00

82 lines
2.2 KiB
Plaintext

{
"cells": [
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"## Imported modules"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [],
"source": [
"import json\n",
"import scrapping\n",
"import psycopg2\n",
"from psycopg2.extras import execute_batch\n",
"import re\n",
"import logging\n",
"import datetime\n",
"logging.basicConfig(\n",
" level=logging.INFO, filename=f'{datetime.datetime.now().strftime(\"%Y-%m-%d_%H-%M-%S\")}.log', filemode='w',\n",
" format='%(name)s - %(levelname)s - %(message)s', datefmt='%d-%b-%y %H:%M:%S')\n",
"# from psycopg2.extensions import register_adapter\n"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [],
"source": [
"conn = psycopg2.connect(\"dbname='dh' user='dh' host='dh.saret.tk' password='qwerty'\")\n",
"connection = conn.cursor()\n",
"with open('project_list') as f:\n",
" for project in f.read().split('\\n'):\n",
" scrap = scrapping.get_raw_english_texts_of_project(project)\n",
" # connection = connection.execute(\"insert into raw_texts values (%(id_text)s, %(project_name)s, %(raw_text)s)\", scrap)\n",
" try:\n",
" execute_batch(connection, \"insert into raw_texts values (%(id_text)s, %(project_name)s, %(raw_text)s)\", scrap)\n",
" conn.commit()\n",
" except Exception:\n",
" with open(project, 'w') as f:\n",
" f.write(json.dumps(scrap))\n",
" logging.error(f\"Error in {project}:{Exception}\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "venv",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.9.13"
},
"orig_nbformat": 4
},
"nbformat": 4,
"nbformat_minor": 2
}