Skip to content

Instantly share code, notes, and snippets.

@michael-erasmus

michael-erasmus/Audit Views.ipynb

Created December 11, 2015 23:53
Show Gist options
  • Save michael-erasmus/209ec081d5bd2831c149 to your computer and use it in GitHub Desktop.
Save michael-erasmus/209ec081d5bd2831c149 to your computer and use it in GitHub Desktop.
Download ZIP
Display the source blob
Display the rendered blob
Raw
Audit Views.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"import os\n",
"import glob\n",
"import pandas as pd\n",
"from StringIO import StringIO"
]
},
{
"cell_type": "code",
"execution_count": 53,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"def get_file_blame_stats(fname):\n",
" \n",
" raw = os.popen('git ls-tree -r -z --name-only HEAD -- %s | xargs -0 -n1 git blame \\\n",
" --line-porcelain HEAD |grep \"^author \"|sort|uniq -c|sort -nr' % fname).read()\n",
"\n",
" data = pd.read_csv(StringIO(raw), names=['raw'])\n",
" data['lines'] = data.raw.map(lambda r: int(r.partition('author')[0]))\n",
" data['author'] = data.raw.map(lambda r: r.partition('author')[2])\n",
" return data.drop('raw',1)\n",
"\n",
"def get_file_owner(fname):\n",
" #\"Getting a file 'owner' is done by getting the user who've changed the most lines in the file :)\"\n",
" stats = get_file_blame_stats(fname)\n",
" return stats['author'][0]\n",
"\n",
"def get_file_last_updated(fname):\n",
" return os.popen('git log -1 --format=\"%ad\" -- ' + fname).read().rstrip()\n",
"\n",
"def get_number_of_lines_in_file(fname):\n",
" return sum(1 for line in open(fname))\n",
"\n",
"def get_number_of_commits_to_file(fname):\n",
" raw = os.popen('git log --name-only --pretty=format: %s | sort | uniq -c | sort -nr | tail -n 1' % fname).read()\n",
" return [int(s) for s in raw.split() if s.isdigit()][0]"
]
},
{
"cell_type": "code",
"execution_count": 55,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"def get_data_for_files(path):\n",
" data = []\n",
" for fname in glob.glob(path):\n",
" owner = get_file_owner(fname)\n",
" number_of_lines = get_number_of_lines_in_file(fname)\n",
" number_of_commits = get_number_of_commits_to_file(fname)\n",
" last_updated = get_file_last_updated(fname)\n",
" data.append({\n",
" 'file' : fname, \n",
" 'owner' : owner, \n",
" 'last_updated': last_updated,\n",
" 'number_of_lines': number_of_lines,\n",
" 'number_of_commits': number_of_commits\n",
" })\n",
"\n",
" return pd.DataFrame(data)"
]
},
{
"cell_type": "code",
"execution_count": 56,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>file</th>\n",
" <th>last_updated</th>\n",
" <th>number_of_commits</th>\n",
" <th>number_of_lines</th>\n",
" <th>owner</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>actions_taken.view.lookml</td>\n",
" <td>Wed Nov 25 11:24:06 2015 +0000</td>\n",
" <td>79</td>\n",
" <td>876</td>\n",
" <td>Maxime Berthelot</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>actions_taken_by_customer.view.lookml</td>\n",
" <td>Fri Oct 24 22:22:50 2014 +0000</td>\n",
" <td>1</td>\n",
" <td>37</td>\n",
" <td>Michael Erasmus</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>actions_taken_by_user_by_day.view.lookml</td>\n",
" <td>Fri Nov 28 14:36:04 2014 +0000</td>\n",
" <td>1</td>\n",
" <td>76</td>\n",
" <td>Michael Erasmus</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>actions_taken_cohorts.view.lookml</td>\n",
" <td>Thu Nov 12 15:12:04 2015 +0000</td>\n",
" <td>3</td>\n",
" <td>71</td>\n",
" <td>Sunil Sadasivan</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>actions_taken_mrr.view.lookml</td>\n",
" <td>Mon Oct 27 08:38:48 2014 +0000</td>\n",
" <td>1</td>\n",
" <td>54</td>\n",
" <td>Michael Erasmus</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" file last_updated \\\n",
"0 actions_taken.view.lookml Wed Nov 25 11:24:06 2015 +0000 \n",
"1 actions_taken_by_customer.view.lookml Fri Oct 24 22:22:50 2014 +0000 \n",
"2 actions_taken_by_user_by_day.view.lookml Fri Nov 28 14:36:04 2014 +0000 \n",
"3 actions_taken_cohorts.view.lookml Thu Nov 12 15:12:04 2015 +0000 \n",
"4 actions_taken_mrr.view.lookml Mon Oct 27 08:38:48 2014 +0000 \n",
"\n",
" number_of_commits number_of_lines owner \n",
"0 79 876 Maxime Berthelot \n",
"1 1 37 Michael Erasmus \n",
"2 1 76 Michael Erasmus \n",
"3 3 71 Sunil Sadasivan \n",
"4 1 54 Michael Erasmus "
]
},
"execution_count": 56,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"view_data = get_data_for_files(\"*.view.lookml\")\n",
"view_data.head()"
]
},
{
"cell_type": "code",
"execution_count": 170,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>file</th>\n",
" <th>last_updated</th>\n",
" <th>owner</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>awesome.dashboard.lookml</td>\n",
" <td>Tue Nov 3 02:40:34 2015 +0000</td>\n",
" <td>Sunil Sadasivan</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>churn_dashboard.dashboard.lookml</td>\n",
" <td>Wed Nov 26 09:19:36 2014 +0000</td>\n",
" <td>Steven Cheng</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>client_dashboard.dashboard.lookml</td>\n",
" <td>Fri Jun 12 06:17:23 2015 +0000</td>\n",
" <td>Sunil Sadasivan</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>daily_pulse.dashboard.lookml</td>\n",
" <td>Mon Mar 23 15:46:43 2015 +0000</td>\n",
" <td>Steven Cheng</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>experiment_results.dashboard.lookml</td>\n",
" <td>Mon Mar 23 15:46:43 2015 +0000</td>\n",
" <td>Sunil Sadasivan</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" file last_updated \\\n",
"0 awesome.dashboard.lookml Tue Nov 3 02:40:34 2015 +0000 \n",
"1 churn_dashboard.dashboard.lookml Wed Nov 26 09:19:36 2014 +0000 \n",
"2 client_dashboard.dashboard.lookml Fri Jun 12 06:17:23 2015 +0000 \n",
"3 daily_pulse.dashboard.lookml Mon Mar 23 15:46:43 2015 +0000 \n",
"4 experiment_results.dashboard.lookml Mon Mar 23 15:46:43 2015 +0000 \n",
"\n",
" owner \n",
"0 Sunil Sadasivan \n",
"1 Steven Cheng \n",
"2 Sunil Sadasivan \n",
"3 Steven Cheng \n",
"4 Sunil Sadasivan "
]
},
"execution_count": 170,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"dashboard_data = get_data_for_files(\"*.dashboard.lookml\")\n",
"dashboard_data.head()"
]
},
{
"cell_type": "code",
"execution_count": 178,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>file</th>\n",
" <th>last_updated</th>\n",
" <th>owner</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>admin.model.lookml</td>\n",
" <td>Fri May 22 11:48:54 2015 +0000</td>\n",
" <td>Michael Erasmus</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>buffer.model.lookml</td>\n",
" <td>Thu Dec 3 15:02:50 2015 +0000</td>\n",
" <td>Sunil Sadasivan</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>business_usage.model.lookml</td>\n",
" <td>Tue Oct 27 19:59:32 2015 +0000</td>\n",
" <td>Michael Erasmus</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>customer_success.model.lookml</td>\n",
" <td>Wed Nov 25 00:53:49 2015 +0000</td>\n",
" <td>Michael Erasmus</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>event_data.model.lookml</td>\n",
" <td>Thu Oct 15 12:39:45 2015 +0000</td>\n",
" <td>Michael Erasmus</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" file last_updated \\\n",
"0 admin.model.lookml Fri May 22 11:48:54 2015 +0000 \n",
"1 buffer.model.lookml Thu Dec 3 15:02:50 2015 +0000 \n",
"2 business_usage.model.lookml Tue Oct 27 19:59:32 2015 +0000 \n",
"3 customer_success.model.lookml Wed Nov 25 00:53:49 2015 +0000 \n",
"4 event_data.model.lookml Thu Oct 15 12:39:45 2015 +0000 \n",
"\n",
" owner \n",
"0 Michael Erasmus \n",
"1 Sunil Sadasivan \n",
"2 Michael Erasmus \n",
"3 Michael Erasmus \n",
"4 Michael Erasmus "
]
},
"execution_count": 178,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"model_data = get_data_for_files(\"*.model.lookml\")\n",
"model_data.head()"
]
},
{
"cell_type": "code",
"execution_count": 179,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"view_data.to_csv(\"view_lookml_files.csv\", index=False)\n",
"dashboard_data.to_csv(\"dashboard_lookml_files.csv\", index=False)\n",
"model_data.to_csv(\"model_lookml_files.csv\", index=False)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 2",
"language": "python",
"name": "python2"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 2
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython2",
"version": "2.7.10"
}
},
"nbformat": 4,
"nbformat_minor": 0
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment