2017-10-23 00:41:19 +00:00
|
|
|
from sanic import Sanic
|
|
|
|
from sanic import response
|
2017-10-24 02:00:37 +00:00
|
|
|
from sanic.exceptions import NotFound
|
2017-10-24 02:25:48 +00:00
|
|
|
from sanic.views import HTTPMethodView
|
2017-11-23 04:33:55 +00:00
|
|
|
from sanic.request import RequestParameters
|
2017-11-30 16:30:10 +00:00
|
|
|
from jinja2 import Environment, FileSystemLoader, ChoiceLoader, PrefixLoader
|
2017-11-16 01:32:48 +00:00
|
|
|
import re
|
2017-10-23 00:41:19 +00:00
|
|
|
import sqlite3
|
2017-10-23 16:02:40 +00:00
|
|
|
from pathlib import Path
|
2017-11-05 02:21:44 +00:00
|
|
|
from concurrent import futures
|
|
|
|
import asyncio
|
|
|
|
import threading
|
2017-10-24 05:54:58 +00:00
|
|
|
import urllib.parse
|
2017-10-23 00:41:19 +00:00
|
|
|
import json
|
2017-11-17 15:39:36 +00:00
|
|
|
import jinja2
|
2017-10-23 16:02:40 +00:00
|
|
|
import hashlib
|
2017-10-24 23:55:53 +00:00
|
|
|
import time
|
2017-11-10 19:25:54 +00:00
|
|
|
from .utils import (
|
2017-11-20 06:03:24 +00:00
|
|
|
Filters,
|
2017-11-16 01:34:32 +00:00
|
|
|
compound_pks_from_path,
|
2017-11-10 19:25:54 +00:00
|
|
|
CustomJSONEncoder,
|
2017-11-19 16:59:26 +00:00
|
|
|
detect_fts_sql,
|
2017-11-11 05:55:50 +00:00
|
|
|
escape_css_string,
|
2017-11-12 23:17:00 +00:00
|
|
|
escape_sqlite_table_name,
|
2017-11-23 04:03:46 +00:00
|
|
|
filters_should_redirect,
|
2017-11-16 01:34:32 +00:00
|
|
|
get_all_foreign_keys,
|
2017-11-29 17:05:24 +00:00
|
|
|
is_url,
|
2017-11-10 19:25:54 +00:00
|
|
|
InvalidSql,
|
|
|
|
path_from_row_pks,
|
2017-11-10 20:41:14 +00:00
|
|
|
path_with_added_args,
|
2017-11-11 22:40:44 +00:00
|
|
|
path_with_ext,
|
2017-11-10 19:25:54 +00:00
|
|
|
sqlite_timelimit,
|
2017-11-30 07:09:54 +00:00
|
|
|
to_css_class,
|
2017-11-10 19:25:54 +00:00
|
|
|
validate_sql_select,
|
|
|
|
)
|
2017-11-16 15:20:54 +00:00
|
|
|
from .version import __version__
|
2017-10-23 16:02:40 +00:00
|
|
|
|
2017-10-27 07:08:24 +00:00
|
|
|
app_root = Path(__file__).parent.parent
|
2017-10-23 16:02:40 +00:00
|
|
|
|
|
|
|
HASH_BLOCK_SIZE = 1024 * 1024
|
|
|
|
|
2017-11-05 02:21:44 +00:00
|
|
|
connections = threading.local()
|
2017-10-24 02:00:37 +00:00
|
|
|
|
2017-10-23 16:02:40 +00:00
|
|
|
|
2017-11-30 15:51:40 +00:00
|
|
|
class RenderMixin(HTTPMethodView):
|
2017-11-30 16:30:10 +00:00
|
|
|
def render(self, templates, **context):
|
2017-11-30 15:51:40 +00:00
|
|
|
return response.html(
|
2017-11-30 16:30:10 +00:00
|
|
|
self.jinja_env.select_template(templates).render(**context)
|
2017-11-30 15:51:40 +00:00
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
class BaseView(RenderMixin):
|
2017-11-10 19:05:57 +00:00
|
|
|
def __init__(self, datasette):
|
2017-11-11 20:10:51 +00:00
|
|
|
self.ds = datasette
|
2017-11-10 19:05:57 +00:00
|
|
|
self.files = datasette.files
|
2017-11-30 15:51:40 +00:00
|
|
|
self.jinja_env = datasette.jinja_env
|
2017-11-10 19:05:57 +00:00
|
|
|
self.executor = datasette.executor
|
2017-11-10 20:41:14 +00:00
|
|
|
self.page_size = datasette.page_size
|
2017-11-13 19:33:01 +00:00
|
|
|
self.max_returned_rows = datasette.max_returned_rows
|
2017-11-05 02:13:44 +00:00
|
|
|
|
2017-11-13 02:11:52 +00:00
|
|
|
def options(self, request, *args, **kwargs):
|
|
|
|
r = response.text('ok')
|
2017-11-13 18:17:42 +00:00
|
|
|
if self.ds.cors:
|
|
|
|
r.headers['Access-Control-Allow-Origin'] = '*'
|
2017-11-13 02:11:52 +00:00
|
|
|
return r
|
|
|
|
|
2017-11-23 20:32:54 +00:00
|
|
|
def redirect(self, request, path, forward_querystring=True):
|
|
|
|
if request.query_string and '?' not in path and forward_querystring:
|
2017-10-24 23:54:26 +00:00
|
|
|
path = '{}?{}'.format(
|
|
|
|
path, request.query_string
|
|
|
|
)
|
|
|
|
r = response.redirect(path)
|
|
|
|
r.headers['Link'] = '<{}>; rel=preload'.format(path)
|
2017-11-13 18:17:42 +00:00
|
|
|
if self.ds.cors:
|
|
|
|
r.headers['Access-Control-Allow-Origin'] = '*'
|
2017-10-24 23:54:26 +00:00
|
|
|
return r
|
|
|
|
|
2017-11-05 02:21:44 +00:00
|
|
|
async def pks_for_table(self, name, table):
|
|
|
|
rows = [
|
|
|
|
row for row in await self.execute(
|
|
|
|
name,
|
|
|
|
'PRAGMA table_info("{}")'.format(table)
|
|
|
|
)
|
|
|
|
if row[-1]
|
|
|
|
]
|
|
|
|
rows.sort(key=lambda row: row[-1])
|
|
|
|
return [str(r[1]) for r in rows]
|
|
|
|
|
2017-11-11 20:10:51 +00:00
|
|
|
def resolve_db_name(self, db_name, **kwargs):
|
2017-11-13 15:20:02 +00:00
|
|
|
databases = self.ds.inspect()
|
2017-11-11 20:10:51 +00:00
|
|
|
hash = None
|
|
|
|
name = None
|
|
|
|
if '-' in db_name:
|
|
|
|
# Might be name-and-hash, or might just be
|
|
|
|
# a name with a hyphen in it
|
|
|
|
name, hash = db_name.rsplit('-', 1)
|
|
|
|
if name not in databases:
|
|
|
|
# Try the whole name
|
|
|
|
name = db_name
|
|
|
|
hash = None
|
|
|
|
else:
|
|
|
|
name = db_name
|
|
|
|
# Verify the hash
|
|
|
|
try:
|
|
|
|
info = databases[name]
|
|
|
|
except KeyError:
|
|
|
|
raise NotFound('Database not found: {}'.format(name))
|
|
|
|
expected = info['hash'][:7]
|
|
|
|
if expected != hash:
|
|
|
|
should_redirect = '/{}-{}'.format(
|
|
|
|
name, expected,
|
|
|
|
)
|
|
|
|
if 'table' in kwargs:
|
|
|
|
should_redirect += '/' + kwargs['table']
|
2017-11-14 00:44:08 +00:00
|
|
|
if 'pk_path' in kwargs:
|
|
|
|
should_redirect += '/' + kwargs['pk_path']
|
2017-11-11 20:10:51 +00:00
|
|
|
if 'as_json' in kwargs:
|
|
|
|
should_redirect += kwargs['as_json']
|
|
|
|
if 'as_db' in kwargs:
|
|
|
|
should_redirect += kwargs['as_db']
|
|
|
|
return name, expected, should_redirect
|
|
|
|
return name, expected, None
|
|
|
|
|
2017-11-15 02:55:10 +00:00
|
|
|
async def execute(self, db_name, sql, params=None, truncate=False, custom_time_limit=None):
|
2017-11-05 02:21:44 +00:00
|
|
|
"""Executes sql against db_name in a thread"""
|
|
|
|
def sql_operation_in_thread():
|
|
|
|
conn = getattr(connections, db_name, None)
|
|
|
|
if not conn:
|
2017-11-13 15:20:02 +00:00
|
|
|
info = self.ds.inspect()[db_name]
|
2017-11-05 02:21:44 +00:00
|
|
|
conn = sqlite3.connect(
|
|
|
|
'file:{}?immutable=1'.format(info['file']),
|
|
|
|
uri=True,
|
|
|
|
check_same_thread=False,
|
|
|
|
)
|
2017-11-26 22:51:42 +00:00
|
|
|
self.ds.prepare_connection(conn)
|
2017-11-05 02:21:44 +00:00
|
|
|
setattr(connections, db_name, conn)
|
|
|
|
|
2017-11-15 02:55:10 +00:00
|
|
|
time_limit_ms = self.ds.sql_time_limit_ms
|
|
|
|
if custom_time_limit and custom_time_limit < self.ds.sql_time_limit_ms:
|
|
|
|
time_limit_ms = custom_time_limit
|
|
|
|
|
|
|
|
with sqlite_timelimit(conn, time_limit_ms):
|
2017-11-09 16:09:55 +00:00
|
|
|
try:
|
2017-11-13 19:33:01 +00:00
|
|
|
cursor = conn.cursor()
|
|
|
|
cursor.execute(sql, params or {})
|
|
|
|
if self.max_returned_rows and truncate:
|
|
|
|
rows = cursor.fetchmany(self.max_returned_rows + 1)
|
|
|
|
truncated = len(rows) > self.max_returned_rows
|
|
|
|
rows = rows[:self.max_returned_rows]
|
|
|
|
else:
|
|
|
|
rows = cursor.fetchall()
|
|
|
|
truncated = False
|
2017-11-09 16:09:55 +00:00
|
|
|
except Exception:
|
2017-11-12 23:17:00 +00:00
|
|
|
print('ERROR: conn={}, sql = {}, params = {}'.format(
|
|
|
|
conn, repr(sql), params
|
2017-11-09 16:09:55 +00:00
|
|
|
))
|
|
|
|
raise
|
2017-11-13 19:33:01 +00:00
|
|
|
if truncate:
|
|
|
|
return rows, truncated, cursor.description
|
|
|
|
else:
|
|
|
|
return rows
|
2017-11-05 02:21:44 +00:00
|
|
|
|
|
|
|
return await asyncio.get_event_loop().run_in_executor(
|
|
|
|
self.executor, sql_operation_in_thread
|
|
|
|
)
|
|
|
|
|
2017-11-30 16:30:10 +00:00
|
|
|
def get_templates(self, database, table=None):
|
|
|
|
assert NotImplemented
|
|
|
|
|
2017-10-24 02:25:48 +00:00
|
|
|
async def get(self, request, db_name, **kwargs):
|
2017-11-11 20:10:51 +00:00
|
|
|
name, hash, should_redirect = self.resolve_db_name(db_name, **kwargs)
|
2017-10-24 02:25:48 +00:00
|
|
|
if should_redirect:
|
2017-10-24 23:54:26 +00:00
|
|
|
return self.redirect(request, should_redirect)
|
2017-10-25 15:19:32 +00:00
|
|
|
return await self.view_get(request, name, hash, **kwargs)
|
|
|
|
|
|
|
|
async def view_get(self, request, name, hash, **kwargs):
|
2017-10-24 02:25:48 +00:00
|
|
|
try:
|
|
|
|
as_json = kwargs.pop('as_json')
|
|
|
|
except KeyError:
|
|
|
|
as_json = False
|
2017-10-24 14:10:58 +00:00
|
|
|
extra_template_data = {}
|
2017-10-24 23:55:53 +00:00
|
|
|
start = time.time()
|
2017-11-12 21:16:15 +00:00
|
|
|
status_code = 200
|
2017-11-30 16:30:10 +00:00
|
|
|
templates = []
|
2017-10-24 02:56:27 +00:00
|
|
|
try:
|
2017-11-19 20:25:29 +00:00
|
|
|
response_or_template_contexts = await self.data(
|
2017-10-24 14:10:58 +00:00
|
|
|
request, name, hash, **kwargs
|
|
|
|
)
|
2017-11-19 20:25:29 +00:00
|
|
|
if isinstance(response_or_template_contexts, response.HTTPResponse):
|
|
|
|
return response_or_template_contexts
|
|
|
|
else:
|
2017-11-30 16:30:10 +00:00
|
|
|
data, extra_template_data, templates = response_or_template_contexts
|
2017-11-05 02:49:18 +00:00
|
|
|
except (sqlite3.OperationalError, InvalidSql) as e:
|
2017-10-24 02:56:27 +00:00
|
|
|
data = {
|
|
|
|
'ok': False,
|
|
|
|
'error': str(e),
|
2017-11-12 21:16:15 +00:00
|
|
|
'database': name,
|
|
|
|
'database_hash': hash,
|
2017-10-24 02:56:27 +00:00
|
|
|
}
|
2017-11-12 21:16:15 +00:00
|
|
|
status_code = 400
|
2017-11-30 16:30:10 +00:00
|
|
|
templates = ['error.html']
|
2017-10-24 23:55:53 +00:00
|
|
|
end = time.time()
|
2017-11-11 22:20:00 +00:00
|
|
|
data['query_ms'] = (end - start) * 1000
|
2017-11-13 18:39:25 +00:00
|
|
|
for key in ('source', 'source_url', 'license', 'license_url'):
|
|
|
|
value = self.ds.metadata.get(key)
|
|
|
|
if value:
|
|
|
|
data[key] = value
|
2017-10-24 02:25:48 +00:00
|
|
|
if as_json:
|
2017-10-24 15:07:52 +00:00
|
|
|
# Special case for .jsono extension
|
|
|
|
if as_json == '.jsono':
|
|
|
|
columns = data.get('columns')
|
|
|
|
rows = data.get('rows')
|
|
|
|
if rows and columns:
|
|
|
|
data['rows'] = [
|
|
|
|
dict(zip(columns, row))
|
|
|
|
for row in rows
|
|
|
|
]
|
2017-11-13 18:17:42 +00:00
|
|
|
headers = {}
|
|
|
|
if self.ds.cors:
|
|
|
|
headers['Access-Control-Allow-Origin'] = '*'
|
2017-10-24 14:58:41 +00:00
|
|
|
r = response.HTTPResponse(
|
|
|
|
json.dumps(
|
|
|
|
data, cls=CustomJSONEncoder
|
|
|
|
),
|
2017-11-12 21:16:15 +00:00
|
|
|
status=status_code,
|
2017-10-24 14:58:41 +00:00
|
|
|
content_type='application/json',
|
2017-11-13 18:17:42 +00:00
|
|
|
headers=headers,
|
2017-10-24 14:58:41 +00:00
|
|
|
)
|
2017-10-24 02:25:48 +00:00
|
|
|
else:
|
2017-11-18 03:09:32 +00:00
|
|
|
extras = {}
|
|
|
|
if callable(extra_template_data):
|
|
|
|
extras = extra_template_data()
|
|
|
|
if asyncio.iscoroutine(extras):
|
|
|
|
extras = await extras
|
|
|
|
else:
|
|
|
|
extras = extra_template_data
|
|
|
|
context = {
|
|
|
|
**data,
|
|
|
|
**extras,
|
|
|
|
**{
|
|
|
|
'url_json': path_with_ext(request, '.json'),
|
|
|
|
'url_jsono': path_with_ext(request, '.jsono'),
|
|
|
|
'metadata': self.ds.metadata,
|
2017-11-29 02:38:15 +00:00
|
|
|
'extra_css_urls': self.ds.extra_css_urls(),
|
|
|
|
'extra_js_urls': self.ds.extra_js_urls(),
|
2017-11-18 03:09:32 +00:00
|
|
|
'datasette_version': __version__,
|
|
|
|
}
|
|
|
|
}
|
2017-11-30 15:51:40 +00:00
|
|
|
r = self.render(
|
2017-11-30 16:30:10 +00:00
|
|
|
templates,
|
2017-10-24 14:10:58 +00:00
|
|
|
**context,
|
2017-10-24 02:25:48 +00:00
|
|
|
)
|
2017-11-12 21:16:15 +00:00
|
|
|
r.status = status_code
|
2017-10-24 02:36:44 +00:00
|
|
|
# Set far-future cache expiry
|
2017-11-13 19:33:01 +00:00
|
|
|
if self.ds.cache_headers:
|
2017-11-10 20:26:37 +00:00
|
|
|
r.headers['Cache-Control'] = 'max-age={}'.format(
|
|
|
|
365 * 24 * 60 * 60
|
|
|
|
)
|
2017-10-24 02:36:44 +00:00
|
|
|
return r
|
2017-10-23 00:41:19 +00:00
|
|
|
|
|
|
|
|
2017-11-30 15:51:40 +00:00
|
|
|
class IndexView(RenderMixin):
|
2017-11-10 19:05:57 +00:00
|
|
|
def __init__(self, datasette):
|
2017-11-11 20:10:51 +00:00
|
|
|
self.ds = datasette
|
2017-11-10 19:05:57 +00:00
|
|
|
self.files = datasette.files
|
2017-11-30 15:51:40 +00:00
|
|
|
self.jinja_env = datasette.jinja_env
|
2017-11-10 19:05:57 +00:00
|
|
|
self.executor = datasette.executor
|
2017-11-05 02:13:44 +00:00
|
|
|
|
2017-11-11 17:49:47 +00:00
|
|
|
async def get(self, request, as_json):
|
2017-11-05 02:13:44 +00:00
|
|
|
databases = []
|
2017-11-13 15:20:02 +00:00
|
|
|
for key, info in sorted(self.ds.inspect().items()):
|
2017-11-22 20:12:15 +00:00
|
|
|
tables = [t for t in info['tables'].values() if not t['hidden']]
|
|
|
|
hidden_tables = [t for t in info['tables'].values() if t['hidden']]
|
2017-11-05 02:13:44 +00:00
|
|
|
database = {
|
|
|
|
'name': key,
|
|
|
|
'hash': info['hash'],
|
|
|
|
'path': '{}-{}'.format(key, info['hash'][:7]),
|
|
|
|
'tables_truncated': sorted(
|
2017-11-22 20:12:15 +00:00
|
|
|
tables,
|
|
|
|
key=lambda t: t['count'],
|
2017-11-05 02:13:44 +00:00
|
|
|
reverse=True
|
|
|
|
)[:5],
|
2017-11-22 20:12:15 +00:00
|
|
|
'tables_count': len(tables),
|
|
|
|
'tables_more': len(tables) > 5,
|
|
|
|
'table_rows': sum(t['count'] for t in tables),
|
|
|
|
'hidden_table_rows': sum(t['count'] for t in hidden_tables),
|
|
|
|
'hidden_tables_count': len(hidden_tables),
|
|
|
|
'views_count': len(info['views']),
|
2017-11-05 02:13:44 +00:00
|
|
|
}
|
|
|
|
databases.append(database)
|
2017-11-11 17:49:47 +00:00
|
|
|
if as_json:
|
|
|
|
return response.HTTPResponse(
|
|
|
|
json.dumps(
|
|
|
|
{db['name']: db for db in databases},
|
|
|
|
cls=CustomJSONEncoder
|
|
|
|
),
|
|
|
|
content_type='application/json',
|
|
|
|
headers={
|
|
|
|
'Access-Control-Allow-Origin': '*'
|
|
|
|
}
|
|
|
|
)
|
|
|
|
else:
|
2017-11-30 15:51:40 +00:00
|
|
|
return self.render(
|
2017-11-30 16:30:10 +00:00
|
|
|
['index.html'],
|
2017-11-11 17:49:47 +00:00
|
|
|
databases=databases,
|
2017-11-13 15:20:02 +00:00
|
|
|
metadata=self.ds.metadata,
|
2017-11-16 15:20:54 +00:00
|
|
|
datasette_version=__version__,
|
2017-11-29 02:38:15 +00:00
|
|
|
extra_css_urls=self.ds.extra_css_urls(),
|
|
|
|
extra_js_urls=self.ds.extra_js_urls(),
|
2017-11-11 17:49:47 +00:00
|
|
|
)
|
2017-10-24 02:00:37 +00:00
|
|
|
|
|
|
|
|
|
|
|
async def favicon(request):
|
|
|
|
return response.text('')
|
|
|
|
|
|
|
|
|
2017-10-24 02:25:48 +00:00
|
|
|
class DatabaseView(BaseView):
|
2017-11-24 21:55:00 +00:00
|
|
|
re_named_parameter = re.compile(':([a-zA-Z0-9_]+)')
|
2017-10-23 00:41:19 +00:00
|
|
|
|
2017-11-05 02:21:44 +00:00
|
|
|
async def data(self, request, name, hash):
|
2017-11-12 02:35:35 +00:00
|
|
|
if request.args.get('sql'):
|
|
|
|
return await self.custom_sql(request, name, hash)
|
2017-11-22 20:12:15 +00:00
|
|
|
info = self.ds.inspect()[name]
|
|
|
|
tables = list(info['tables'].values())
|
|
|
|
tables.sort(key=lambda t: (t['hidden'], t['name']))
|
2017-10-24 02:25:48 +00:00
|
|
|
return {
|
|
|
|
'database': name,
|
2017-11-12 01:50:21 +00:00
|
|
|
'tables': tables,
|
2017-11-22 20:12:15 +00:00
|
|
|
'hidden_count': len([t for t in tables if t['hidden']]),
|
|
|
|
'views': info['views'],
|
2017-10-24 14:10:58 +00:00
|
|
|
}, {
|
|
|
|
'database_hash': hash,
|
2017-11-22 20:12:15 +00:00
|
|
|
'show_hidden': request.args.get('_show_hidden'),
|
2017-11-30 16:30:10 +00:00
|
|
|
}, ('database-{}.html'.format(to_css_class(name)), 'database.html')
|
2017-10-23 00:41:19 +00:00
|
|
|
|
2017-11-12 02:35:35 +00:00
|
|
|
async def custom_sql(self, request, name, hash):
|
|
|
|
params = request.raw_args
|
|
|
|
sql = params.pop('sql')
|
|
|
|
validate_sql_select(sql)
|
2017-11-16 01:32:48 +00:00
|
|
|
|
|
|
|
# Extract any :named parameters
|
|
|
|
named_parameters = self.re_named_parameter.findall(sql)
|
|
|
|
named_parameter_values = {
|
|
|
|
named_parameter: params.get(named_parameter) or ''
|
|
|
|
for named_parameter in named_parameters
|
|
|
|
}
|
|
|
|
|
|
|
|
# Set to blank string if missing from params
|
|
|
|
for named_parameter in named_parameters:
|
|
|
|
if named_parameter not in params:
|
|
|
|
params[named_parameter] = ''
|
|
|
|
|
2017-11-15 02:55:10 +00:00
|
|
|
extra_args = {}
|
|
|
|
if params.get('_sql_time_limit_ms'):
|
|
|
|
extra_args['custom_time_limit'] = int(params['_sql_time_limit_ms'])
|
|
|
|
rows, truncated, description = await self.execute(
|
|
|
|
name, sql, params, truncate=True, **extra_args
|
|
|
|
)
|
2017-11-13 19:33:01 +00:00
|
|
|
columns = [r[0] for r in description]
|
2017-11-12 02:35:35 +00:00
|
|
|
return {
|
|
|
|
'database': name,
|
|
|
|
'rows': rows,
|
2017-11-13 19:33:01 +00:00
|
|
|
'truncated': truncated,
|
2017-11-12 02:35:35 +00:00
|
|
|
'columns': columns,
|
|
|
|
'query': {
|
|
|
|
'sql': sql,
|
|
|
|
'params': params,
|
|
|
|
}
|
|
|
|
}, {
|
|
|
|
'database_hash': hash,
|
|
|
|
'custom_sql': True,
|
2017-11-16 01:32:48 +00:00
|
|
|
'named_parameter_values': named_parameter_values,
|
2017-11-30 16:30:10 +00:00
|
|
|
}, ('database-{}.html'.format(to_css_class(name)), 'database.html')
|
2017-11-12 02:35:35 +00:00
|
|
|
|
2017-10-24 02:25:48 +00:00
|
|
|
|
2017-10-25 15:19:32 +00:00
|
|
|
class DatabaseDownload(BaseView):
|
|
|
|
async def view_get(self, request, name, hash, **kwargs):
|
2017-11-13 15:20:02 +00:00
|
|
|
filepath = self.ds.inspect()[name]['file']
|
2017-10-25 15:19:32 +00:00
|
|
|
return await response.file_stream(
|
|
|
|
filepath, headers={
|
|
|
|
'Content-Disposition': 'attachment; filename="{}"'.format(filepath)
|
|
|
|
}
|
|
|
|
)
|
|
|
|
|
|
|
|
|
2017-11-23 21:51:16 +00:00
|
|
|
class RowTableShared(BaseView):
|
|
|
|
async def make_display_rows(self, database, database_hash, table, rows, display_columns, pks, is_view, use_rowid):
|
|
|
|
# Get fancy with foreign keys
|
|
|
|
expanded = {}
|
|
|
|
tables = self.ds.inspect()[database]['tables']
|
|
|
|
table_info = tables.get(table) or {}
|
|
|
|
if table_info and not is_view:
|
|
|
|
foreign_keys = table_info['foreign_keys']['outgoing']
|
|
|
|
for fk in foreign_keys:
|
|
|
|
label_column = tables.get(fk['other_table'], {}).get('label_column')
|
|
|
|
if not label_column:
|
|
|
|
# We only link cells to other tables with label columns defined
|
|
|
|
continue
|
|
|
|
ids_to_lookup = set([row[fk['column']] for row in rows])
|
|
|
|
sql = 'select "{other_column}", "{label_column}" from {other_table} where "{other_column}" in ({placeholders})'.format(
|
|
|
|
other_column=fk['other_column'],
|
|
|
|
label_column=label_column,
|
|
|
|
other_table=escape_sqlite_table_name(fk['other_table']),
|
|
|
|
placeholders=', '.join(['?'] * len(ids_to_lookup)),
|
|
|
|
)
|
|
|
|
try:
|
|
|
|
results = await self.execute(database, sql, list(set(ids_to_lookup)))
|
|
|
|
except sqlite3.OperationalError:
|
|
|
|
# Probably hit the timelimit
|
|
|
|
pass
|
|
|
|
else:
|
|
|
|
for id, value in results:
|
|
|
|
expanded[(fk['column'], id)] = (fk['other_table'], value)
|
|
|
|
|
|
|
|
to_return = []
|
|
|
|
for row in rows:
|
|
|
|
cells = []
|
|
|
|
# Unless we are a view, the first column is a link - either to the rowid
|
|
|
|
# or to the simple or compound primary key
|
|
|
|
if not is_view:
|
|
|
|
display_value = jinja2.Markup(
|
|
|
|
'<a href="/{database}-{database_hash}/{table}/{flat_pks}">{flat_pks}</a>'.format(
|
|
|
|
database=database,
|
|
|
|
database_hash=database_hash,
|
|
|
|
table=urllib.parse.quote_plus(table),
|
|
|
|
flat_pks=path_from_row_pks(row, pks, use_rowid),
|
|
|
|
)
|
|
|
|
)
|
|
|
|
cells.append({
|
|
|
|
'column': 'rowid' if use_rowid else 'Link',
|
|
|
|
'value': display_value,
|
|
|
|
})
|
|
|
|
|
|
|
|
for value, column in zip(row, display_columns):
|
|
|
|
if use_rowid and column == 'rowid':
|
|
|
|
# We already showed this in the linked first column
|
|
|
|
continue
|
|
|
|
elif (column, value) in expanded:
|
|
|
|
other_table, label = expanded[(column, value)]
|
|
|
|
display_value = jinja2.Markup(
|
|
|
|
# TODO: Escape id/label/etc so no XSS here
|
|
|
|
'<a href="/{database}-{database_hash}/{table}/{id}">{label}</a> <em>{id}</em>'.format(
|
|
|
|
database=database,
|
|
|
|
database_hash=database_hash,
|
2017-11-25 00:05:28 +00:00
|
|
|
table=urllib.parse.quote_plus(other_table),
|
2017-11-23 21:51:16 +00:00
|
|
|
id=value,
|
|
|
|
label=label,
|
|
|
|
)
|
|
|
|
)
|
|
|
|
elif value is None:
|
|
|
|
display_value = jinja2.Markup(' ')
|
2017-11-29 17:05:24 +00:00
|
|
|
elif is_url(str(value).strip()):
|
|
|
|
display_value = jinja2.Markup(
|
|
|
|
'<a href="{url}">{url}</a>'.format(url=value.strip())
|
|
|
|
)
|
2017-11-23 21:51:16 +00:00
|
|
|
else:
|
|
|
|
display_value = str(value)
|
|
|
|
cells.append({
|
|
|
|
'column': column,
|
|
|
|
'value': display_value,
|
|
|
|
})
|
|
|
|
to_return.append(cells)
|
|
|
|
return to_return
|
|
|
|
|
|
|
|
|
|
|
|
class TableView(RowTableShared):
|
2017-11-05 02:21:44 +00:00
|
|
|
async def data(self, request, name, hash, table):
|
2017-10-25 01:33:12 +00:00
|
|
|
table = urllib.parse.unquote_plus(table)
|
2017-11-09 14:39:50 +00:00
|
|
|
pks = await self.pks_for_table(name, table)
|
2017-11-12 20:31:46 +00:00
|
|
|
is_view = bool(list(await self.execute(name, "SELECT count(*) from sqlite_master WHERE type = 'view' and name=:n", {
|
|
|
|
'n': table,
|
|
|
|
}))[0][0])
|
|
|
|
view_definition = None
|
2017-11-12 21:16:59 +00:00
|
|
|
table_definition = None
|
2017-11-12 20:31:46 +00:00
|
|
|
if is_view:
|
|
|
|
view_definition = list(await self.execute(name, 'select sql from sqlite_master where name = :n and type="view"', {
|
|
|
|
'n': table,
|
2017-11-12 21:16:59 +00:00
|
|
|
}))[0][0]
|
|
|
|
else:
|
|
|
|
table_definition = list(await self.execute(name, 'select sql from sqlite_master where name = :n and type="table"', {
|
|
|
|
'n': table,
|
|
|
|
}))[0][0]
|
2017-11-12 20:31:46 +00:00
|
|
|
use_rowid = not pks and not is_view
|
2017-11-09 14:39:50 +00:00
|
|
|
if use_rowid:
|
|
|
|
select = 'rowid, *'
|
2017-11-09 16:11:14 +00:00
|
|
|
order_by = 'rowid'
|
|
|
|
else:
|
|
|
|
select = '*'
|
|
|
|
order_by = ', '.join(pks)
|
|
|
|
|
2017-11-12 20:31:46 +00:00
|
|
|
if is_view:
|
|
|
|
order_by = ''
|
|
|
|
|
2017-11-23 04:33:55 +00:00
|
|
|
# We roll our own query_string decoder because by default Sanic
|
|
|
|
# drops anything with an empty value e.g. ?name__exact=
|
|
|
|
args = RequestParameters(
|
|
|
|
urllib.parse.parse_qs(request.query_string, keep_blank_values=True)
|
|
|
|
)
|
|
|
|
|
2017-11-10 20:41:14 +00:00
|
|
|
# Special args start with _ and do not contain a __
|
|
|
|
# That's so if there is a column that starts with _
|
|
|
|
# it can still be queried using ?_col__exact=blah
|
|
|
|
special_args = {}
|
2017-11-20 06:04:13 +00:00
|
|
|
special_args_lists = {}
|
2017-11-10 20:41:14 +00:00
|
|
|
other_args = {}
|
2017-11-23 04:33:55 +00:00
|
|
|
for key, value in args.items():
|
2017-11-10 20:41:14 +00:00
|
|
|
if key.startswith('_') and '__' not in key:
|
|
|
|
special_args[key] = value[0]
|
2017-11-20 06:04:13 +00:00
|
|
|
special_args_lists[key] = value
|
2017-11-10 20:41:14 +00:00
|
|
|
else:
|
|
|
|
other_args[key] = value[0]
|
|
|
|
|
2017-11-19 20:25:29 +00:00
|
|
|
# Handle ?_filter_column and redirect, if present
|
2017-11-23 04:03:46 +00:00
|
|
|
redirect_params = filters_should_redirect(special_args)
|
|
|
|
if redirect_params:
|
2017-11-23 20:32:54 +00:00
|
|
|
return self.redirect(
|
|
|
|
request,
|
|
|
|
path_with_added_args(request, redirect_params),
|
|
|
|
forward_querystring=False
|
|
|
|
)
|
2017-11-19 20:25:29 +00:00
|
|
|
|
2017-11-20 06:03:24 +00:00
|
|
|
filters = Filters(sorted(other_args.items()))
|
|
|
|
where_clauses, params = filters.build_where_clauses()
|
2017-11-10 20:41:14 +00:00
|
|
|
|
2017-11-19 16:59:26 +00:00
|
|
|
# _search support:
|
|
|
|
fts_table = None
|
|
|
|
fts_sql = detect_fts_sql(table)
|
|
|
|
fts_rows = list(await self.execute(name, fts_sql))
|
|
|
|
if fts_rows:
|
2017-11-19 20:25:29 +00:00
|
|
|
fts_table = fts_rows[0][0]
|
2017-11-19 16:59:26 +00:00
|
|
|
|
|
|
|
search = special_args.get('_search')
|
2017-11-24 22:22:57 +00:00
|
|
|
search_description = None
|
2017-11-19 16:59:26 +00:00
|
|
|
if search and fts_table:
|
|
|
|
where_clauses.append(
|
2017-11-20 05:59:53 +00:00
|
|
|
'rowid in (select rowid from [{fts_table}] where [{fts_table}] match :search)'.format(
|
2017-11-19 16:59:26 +00:00
|
|
|
fts_table=fts_table
|
|
|
|
)
|
|
|
|
)
|
2017-11-24 22:22:57 +00:00
|
|
|
search_description = 'search matches "{}"'.format(search)
|
2017-11-19 16:59:26 +00:00
|
|
|
params['search'] = search
|
|
|
|
|
2017-11-13 20:34:56 +00:00
|
|
|
next = special_args.get('_next')
|
2017-11-13 21:10:55 +00:00
|
|
|
offset = ''
|
2017-11-13 20:34:56 +00:00
|
|
|
if next:
|
2017-11-13 21:10:55 +00:00
|
|
|
if is_view:
|
|
|
|
# _next is an offset
|
|
|
|
offset = ' offset {}'.format(int(next))
|
|
|
|
elif use_rowid:
|
2017-11-10 20:41:14 +00:00
|
|
|
where_clauses.append(
|
|
|
|
'rowid > :p{}'.format(
|
|
|
|
len(params),
|
|
|
|
)
|
|
|
|
)
|
2017-11-13 20:34:56 +00:00
|
|
|
params['p{}'.format(len(params))] = next
|
2017-11-10 20:41:14 +00:00
|
|
|
else:
|
2017-11-13 20:34:56 +00:00
|
|
|
pk_values = compound_pks_from_path(next)
|
2017-11-10 20:41:14 +00:00
|
|
|
if len(pk_values) == len(pks):
|
|
|
|
param_counter = len(params)
|
|
|
|
for pk, value in zip(pks, pk_values):
|
|
|
|
where_clauses.append(
|
|
|
|
'"{}" > :p{}'.format(
|
|
|
|
pk, param_counter,
|
|
|
|
)
|
|
|
|
)
|
|
|
|
params['p{}'.format(param_counter)] = value
|
|
|
|
param_counter += 1
|
|
|
|
|
|
|
|
where_clause = ''
|
|
|
|
if where_clauses:
|
2017-11-11 22:20:00 +00:00
|
|
|
where_clause = 'where {} '.format(' and '.join(where_clauses))
|
2017-11-10 20:41:14 +00:00
|
|
|
|
2017-11-12 20:31:46 +00:00
|
|
|
if order_by:
|
|
|
|
order_by = 'order by {} '.format(order_by)
|
|
|
|
|
2017-11-20 06:04:13 +00:00
|
|
|
# _group_count=col1&_group_count=col2
|
|
|
|
group_count = special_args_lists.get('_group_count') or []
|
|
|
|
if group_count:
|
2017-11-24 22:01:53 +00:00
|
|
|
count_sql = None
|
2017-11-20 06:04:13 +00:00
|
|
|
sql = 'select {group_cols}, count(*) as "count" from {table_name} {where} group by {group_cols} order by "count" desc limit 100'.format(
|
|
|
|
group_cols=', '.join('"{}"'.format(group_count_col) for group_count_col in group_count),
|
|
|
|
table_name=escape_sqlite_table_name(table),
|
|
|
|
where=where_clause,
|
|
|
|
)
|
|
|
|
is_view = True
|
|
|
|
else:
|
2017-11-24 22:01:53 +00:00
|
|
|
count_sql = 'select count(*) from {table_name} {where}'.format(
|
|
|
|
table_name=escape_sqlite_table_name(table),
|
|
|
|
where=where_clause,
|
|
|
|
)
|
2017-11-20 06:04:13 +00:00
|
|
|
sql = 'select {select} from {table_name} {where}{order_by}limit {limit}{offset}'.format(
|
|
|
|
select=select,
|
|
|
|
table_name=escape_sqlite_table_name(table),
|
|
|
|
where=where_clause,
|
|
|
|
order_by=order_by,
|
|
|
|
limit=self.page_size + 1,
|
|
|
|
offset=offset,
|
|
|
|
)
|
2017-10-25 01:34:54 +00:00
|
|
|
|
2017-11-15 02:55:10 +00:00
|
|
|
extra_args = {}
|
|
|
|
if request.raw_args.get('_sql_time_limit_ms'):
|
|
|
|
extra_args['custom_time_limit'] = int(request.raw_args['_sql_time_limit_ms'])
|
|
|
|
|
|
|
|
rows, truncated, description = await self.execute(
|
|
|
|
name, sql, params, truncate=True, **extra_args
|
|
|
|
)
|
2017-10-25 01:34:54 +00:00
|
|
|
|
2017-11-13 19:33:01 +00:00
|
|
|
columns = [r[0] for r in description]
|
2017-10-24 14:58:41 +00:00
|
|
|
rows = list(rows)
|
2017-11-17 15:39:36 +00:00
|
|
|
|
|
|
|
display_columns = columns
|
|
|
|
if not use_rowid and not is_view:
|
|
|
|
display_columns = ['Link'] + display_columns
|
|
|
|
|
2017-11-13 15:20:02 +00:00
|
|
|
info = self.ds.inspect()
|
2017-11-16 14:46:49 +00:00
|
|
|
table_rows = None
|
|
|
|
if not is_view:
|
|
|
|
table_rows = info[name]['tables'][table]['count']
|
2017-11-18 03:09:32 +00:00
|
|
|
|
|
|
|
# Pagination next link
|
2017-11-13 21:10:55 +00:00
|
|
|
next_value = None
|
2017-11-13 20:34:56 +00:00
|
|
|
next_url = None
|
2017-11-10 20:41:14 +00:00
|
|
|
if len(rows) > self.page_size:
|
2017-11-13 21:10:55 +00:00
|
|
|
if is_view:
|
|
|
|
next_value = int(next or 0) + self.page_size
|
|
|
|
else:
|
|
|
|
next_value = path_from_row_pks(rows[-2], pks, use_rowid)
|
|
|
|
next_url = urllib.parse.urljoin(request.url, path_with_added_args(request, {
|
|
|
|
'_next': next_value,
|
|
|
|
}))
|
2017-11-17 15:39:36 +00:00
|
|
|
|
2017-11-24 22:01:53 +00:00
|
|
|
# Number of filtered rows in whole set:
|
|
|
|
filtered_table_rows = None
|
|
|
|
if not where_clauses and not is_view:
|
|
|
|
# Use the pre-calculated total
|
|
|
|
filtered_table_rows = table_rows
|
|
|
|
elif not truncated and len(rows) < self.page_size:
|
|
|
|
filtered_table_rows = len(rows)
|
|
|
|
else:
|
|
|
|
# Attempt a full count, if we can do it in < X ms
|
|
|
|
if count_sql:
|
|
|
|
try:
|
|
|
|
count_rows = list(await self.execute(name, count_sql, params))
|
|
|
|
filtered_table_rows = count_rows[0][0]
|
|
|
|
except sqlite3.OperationalError:
|
|
|
|
# Almost certainly hit the timeout
|
|
|
|
pass
|
|
|
|
|
2017-11-24 22:22:57 +00:00
|
|
|
# human_filter_description combines filters AND search, if provided
|
|
|
|
human_description = filters.human_description(extra=search_description)
|
|
|
|
|
2017-11-18 03:09:32 +00:00
|
|
|
async def extra_template():
|
|
|
|
return {
|
|
|
|
'database_hash': hash,
|
2017-11-24 22:22:57 +00:00
|
|
|
'human_filter_description': human_description,
|
2017-11-19 16:59:26 +00:00
|
|
|
'supports_search': bool(fts_table),
|
|
|
|
'search': search or '',
|
2017-11-18 03:09:32 +00:00
|
|
|
'use_rowid': use_rowid,
|
2017-11-23 04:33:55 +00:00
|
|
|
'filters': filters,
|
2017-11-18 03:09:32 +00:00
|
|
|
'display_columns': display_columns,
|
|
|
|
'display_rows': await self.make_display_rows(name, hash, table, rows, display_columns, pks, is_view, use_rowid),
|
|
|
|
}
|
|
|
|
|
2017-10-24 02:25:48 +00:00
|
|
|
return {
|
|
|
|
'database': name,
|
|
|
|
'table': table,
|
2017-11-12 20:31:46 +00:00
|
|
|
'is_view': is_view,
|
|
|
|
'view_definition': view_definition,
|
2017-11-12 21:16:59 +00:00
|
|
|
'table_definition': table_definition,
|
2017-11-10 20:41:14 +00:00
|
|
|
'rows': rows[:self.page_size],
|
2017-11-13 19:33:01 +00:00
|
|
|
'truncated': truncated,
|
2017-11-11 22:20:00 +00:00
|
|
|
'table_rows': table_rows,
|
2017-11-24 22:01:53 +00:00
|
|
|
'filtered_table_rows': filtered_table_rows,
|
2017-10-24 02:25:48 +00:00
|
|
|
'columns': columns,
|
2017-10-24 14:10:58 +00:00
|
|
|
'primary_keys': pks,
|
2017-11-11 22:20:00 +00:00
|
|
|
'query': {
|
|
|
|
'sql': sql,
|
|
|
|
'params': params,
|
|
|
|
},
|
2017-11-13 21:10:55 +00:00
|
|
|
'next': next_value and str(next_value) or None,
|
2017-11-13 20:34:56 +00:00
|
|
|
'next_url': next_url,
|
2017-11-30 16:30:10 +00:00
|
|
|
}, extra_template, (
|
|
|
|
'table-{}-{}.html'.format(to_css_class(name), to_css_class(table)),
|
|
|
|
'table.html'
|
|
|
|
)
|
2017-11-18 03:09:32 +00:00
|
|
|
|
2017-11-17 15:39:36 +00:00
|
|
|
|
2017-11-23 21:51:16 +00:00
|
|
|
class RowView(RowTableShared):
|
2017-11-05 02:21:44 +00:00
|
|
|
async def data(self, request, name, hash, table, pk_path):
|
2017-10-27 07:16:18 +00:00
|
|
|
table = urllib.parse.unquote_plus(table)
|
2017-10-24 05:54:58 +00:00
|
|
|
pk_values = compound_pks_from_path(pk_path)
|
2017-11-05 02:21:44 +00:00
|
|
|
pks = await self.pks_for_table(name, table)
|
2017-11-09 14:39:50 +00:00
|
|
|
use_rowid = not pks
|
|
|
|
select = '*'
|
|
|
|
if use_rowid:
|
|
|
|
select = 'rowid, *'
|
|
|
|
pks = ['rowid']
|
2017-10-24 05:54:58 +00:00
|
|
|
wheres = [
|
2017-11-06 02:38:06 +00:00
|
|
|
'"{}"=:p{}'.format(pk, i)
|
|
|
|
for i, pk in enumerate(pks)
|
2017-10-24 05:54:58 +00:00
|
|
|
]
|
2017-11-09 14:39:50 +00:00
|
|
|
sql = 'select {} from "{}" where {}'.format(
|
|
|
|
select, table, ' AND '.join(wheres)
|
2017-10-24 05:54:58 +00:00
|
|
|
)
|
2017-11-06 02:38:06 +00:00
|
|
|
params = {}
|
|
|
|
for i, pk_value in enumerate(pk_values):
|
|
|
|
params['p{}'.format(i)] = pk_value
|
2017-11-14 00:44:08 +00:00
|
|
|
# rows, truncated, description = await self.execute(name, sql, params, truncate=True)
|
|
|
|
rows, truncated, description = await self.execute(name, sql, params, truncate=True)
|
|
|
|
columns = [r[0] for r in description]
|
2017-10-24 05:54:58 +00:00
|
|
|
rows = list(rows)
|
|
|
|
if not rows:
|
|
|
|
raise NotFound('Record not found: {}'.format(pk_values))
|
2017-11-18 03:09:32 +00:00
|
|
|
|
|
|
|
async def template_data():
|
|
|
|
return {
|
|
|
|
'database_hash': hash,
|
|
|
|
'foreign_key_tables': await self.foreign_key_tables(name, table, pk_values),
|
2017-11-23 21:51:16 +00:00
|
|
|
'display_columns': columns,
|
|
|
|
'display_rows': await self.make_display_rows(name, hash, table, rows, columns, pks, False, use_rowid),
|
2017-11-18 03:09:32 +00:00
|
|
|
}
|
|
|
|
|
2017-11-23 21:09:45 +00:00
|
|
|
data = {
|
2017-10-24 05:54:58 +00:00
|
|
|
'database': name,
|
|
|
|
'table': table,
|
|
|
|
'rows': rows,
|
|
|
|
'columns': columns,
|
2017-10-24 14:10:58 +00:00
|
|
|
'primary_keys': pks,
|
2017-11-11 20:36:20 +00:00
|
|
|
'primary_key_values': pk_values,
|
2017-11-23 21:09:45 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
if 'foreign_key_tables' in (request.raw_args.get('_extras') or '').split(','):
|
|
|
|
data['foreign_key_tables'] = await self.foreign_key_tables(name, table, pk_values)
|
|
|
|
|
2017-11-30 16:30:10 +00:00
|
|
|
return data, template_data, (
|
|
|
|
'row-{}-{}.html'.format(to_css_class(name), to_css_class(table)),
|
|
|
|
'row.html'
|
|
|
|
)
|
2017-10-24 05:54:58 +00:00
|
|
|
|
2017-11-17 18:15:44 +00:00
|
|
|
async def foreign_key_tables(self, name, table, pk_values):
|
|
|
|
if len(pk_values) != 1:
|
|
|
|
return []
|
|
|
|
table_info = self.ds.inspect()[name]['tables'].get(table)
|
|
|
|
if not table:
|
|
|
|
return []
|
|
|
|
foreign_keys = table_info['foreign_keys']['incoming']
|
|
|
|
sql = 'select ' + ', '.join([
|
2017-11-23 21:09:45 +00:00
|
|
|
'(select count(*) from {table} where "{column}"=:id)'.format(
|
2017-11-17 18:15:44 +00:00
|
|
|
table=escape_sqlite_table_name(fk['other_table']),
|
|
|
|
column=fk['other_column'],
|
|
|
|
)
|
|
|
|
for fk in foreign_keys
|
|
|
|
])
|
|
|
|
try:
|
|
|
|
rows = list(await self.execute(name, sql, {'id': pk_values[0]}))
|
|
|
|
except sqlite3.OperationalError:
|
|
|
|
# Almost certainly hit the timeout
|
|
|
|
return []
|
2017-11-23 21:09:45 +00:00
|
|
|
foreign_table_counts = dict(
|
|
|
|
zip(
|
|
|
|
[(fk['other_table'], fk['other_column']) for fk in foreign_keys],
|
|
|
|
list(rows[0]),
|
|
|
|
)
|
|
|
|
)
|
2017-11-17 18:15:44 +00:00
|
|
|
foreign_key_tables = []
|
|
|
|
for fk in foreign_keys:
|
2017-11-23 21:09:45 +00:00
|
|
|
count = foreign_table_counts.get((fk['other_table'], fk['other_column'])) or 0
|
|
|
|
foreign_key_tables.append({**fk, **{'count': count}})
|
2017-11-17 18:15:44 +00:00
|
|
|
return foreign_key_tables
|
|
|
|
|
2017-10-24 05:54:58 +00:00
|
|
|
|
2017-11-10 19:05:57 +00:00
|
|
|
class Datasette:
|
2017-11-13 19:33:01 +00:00
|
|
|
def __init__(
|
|
|
|
self, files, num_threads=3, cache_headers=True, page_size=100,
|
2017-11-13 21:58:34 +00:00
|
|
|
max_returned_rows=1000, sql_time_limit_ms=1000, cors=False,
|
2017-12-03 16:33:36 +00:00
|
|
|
inspect_data=None, metadata=None, sqlite_extensions=None,
|
|
|
|
template_dir=None, static_mounts=None):
|
2017-11-10 19:05:57 +00:00
|
|
|
self.files = files
|
|
|
|
self.num_threads = num_threads
|
|
|
|
self.executor = futures.ThreadPoolExecutor(
|
|
|
|
max_workers=num_threads
|
|
|
|
)
|
2017-11-10 20:26:37 +00:00
|
|
|
self.cache_headers = cache_headers
|
2017-11-10 20:41:14 +00:00
|
|
|
self.page_size = page_size
|
2017-11-13 19:33:01 +00:00
|
|
|
self.max_returned_rows = max_returned_rows
|
2017-11-13 21:58:34 +00:00
|
|
|
self.sql_time_limit_ms = sql_time_limit_ms
|
2017-11-13 18:17:42 +00:00
|
|
|
self.cors = cors
|
2017-11-13 15:20:02 +00:00
|
|
|
self._inspect = inspect_data
|
2017-11-13 18:39:25 +00:00
|
|
|
self.metadata = metadata or {}
|
2017-11-15 02:41:03 +00:00
|
|
|
self.sqlite_functions = []
|
2017-11-16 16:46:04 +00:00
|
|
|
self.sqlite_extensions = sqlite_extensions or []
|
2017-11-30 16:05:01 +00:00
|
|
|
self.template_dir = template_dir
|
2017-12-03 16:33:36 +00:00
|
|
|
self.static_mounts = static_mounts or []
|
2017-11-13 15:20:02 +00:00
|
|
|
|
2017-11-29 02:38:15 +00:00
|
|
|
def asset_urls(self, key):
|
|
|
|
for url_or_dict in (self.metadata.get(key) or []):
|
|
|
|
if isinstance(url_or_dict, dict):
|
|
|
|
yield {
|
|
|
|
'url': url_or_dict['url'],
|
|
|
|
'sri': url_or_dict.get('sri'),
|
|
|
|
}
|
|
|
|
else:
|
|
|
|
yield {
|
|
|
|
'url': url_or_dict,
|
|
|
|
}
|
|
|
|
|
|
|
|
def extra_css_urls(self):
|
|
|
|
return self.asset_urls('extra_css_urls')
|
|
|
|
|
|
|
|
def extra_js_urls(self):
|
|
|
|
return self.asset_urls('extra_js_urls')
|
|
|
|
|
2017-11-26 22:51:42 +00:00
|
|
|
def prepare_connection(self, conn):
|
|
|
|
conn.row_factory = sqlite3.Row
|
|
|
|
conn.text_factory = lambda x: str(x, 'utf-8', 'replace')
|
|
|
|
for name, num_args, func in self.sqlite_functions:
|
|
|
|
conn.create_function(name, num_args, func)
|
|
|
|
if self.sqlite_extensions:
|
|
|
|
conn.enable_load_extension(True)
|
|
|
|
for extension in self.sqlite_extensions:
|
|
|
|
conn.execute("SELECT load_extension('{}')".format(extension))
|
|
|
|
|
2017-11-13 15:20:02 +00:00
|
|
|
def inspect(self):
|
|
|
|
if not self._inspect:
|
|
|
|
self._inspect = {}
|
|
|
|
for filename in self.files:
|
|
|
|
path = Path(filename)
|
|
|
|
name = path.stem
|
|
|
|
if name in self._inspect:
|
|
|
|
raise Exception('Multiple files with same stem %s' % name)
|
|
|
|
# Calculate hash, efficiently
|
|
|
|
m = hashlib.sha256()
|
|
|
|
with path.open('rb') as fp:
|
|
|
|
while True:
|
|
|
|
data = fp.read(HASH_BLOCK_SIZE)
|
|
|
|
if not data:
|
|
|
|
break
|
|
|
|
m.update(data)
|
|
|
|
# List tables and their row counts
|
|
|
|
tables = {}
|
2017-11-22 20:12:15 +00:00
|
|
|
views = []
|
2017-11-13 15:20:02 +00:00
|
|
|
with sqlite3.connect('file:{}?immutable=1'.format(path), uri=True) as conn:
|
2017-11-26 22:51:42 +00:00
|
|
|
self.prepare_connection(conn)
|
2017-11-13 15:20:02 +00:00
|
|
|
table_names = [
|
|
|
|
r['name']
|
|
|
|
for r in conn.execute('select * from sqlite_master where type="table"')
|
|
|
|
]
|
2017-11-22 20:12:15 +00:00
|
|
|
views = [v[0] for v in conn.execute('select name from sqlite_master where type = "view"')]
|
2017-11-13 15:20:02 +00:00
|
|
|
for table in table_names:
|
2017-11-18 03:09:32 +00:00
|
|
|
count = conn.execute(
|
|
|
|
'select count(*) from {}'.format(escape_sqlite_table_name(table))
|
|
|
|
).fetchone()[0]
|
|
|
|
label_column = None
|
|
|
|
# If table has two columns, one of which is ID, then label_column is the other one
|
|
|
|
column_names = [r[1] for r in conn.execute(
|
|
|
|
'PRAGMA table_info({});'.format(escape_sqlite_table_name(table))
|
|
|
|
).fetchall()]
|
|
|
|
if column_names and len(column_names) == 2 and 'id' in column_names:
|
|
|
|
label_column = [c for c in column_names if c != 'id'][0]
|
2017-11-16 01:34:32 +00:00
|
|
|
tables[table] = {
|
2017-11-20 16:57:41 +00:00
|
|
|
'name': table,
|
|
|
|
'columns': column_names,
|
2017-11-18 03:09:32 +00:00
|
|
|
'count': count,
|
|
|
|
'label_column': label_column,
|
2017-11-20 16:57:41 +00:00
|
|
|
'hidden': False,
|
2017-11-16 01:34:32 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
foreign_keys = get_all_foreign_keys(conn)
|
|
|
|
for table, info in foreign_keys.items():
|
|
|
|
tables[table]['foreign_keys'] = info
|
2017-11-13 15:20:02 +00:00
|
|
|
|
2017-11-20 16:57:41 +00:00
|
|
|
# Mark tables 'hidden' if they relate to FTS virtual tables
|
|
|
|
fts_tables = [
|
|
|
|
r['name']
|
|
|
|
for r in conn.execute(
|
|
|
|
'''
|
|
|
|
select name from sqlite_master
|
|
|
|
where rootpage = 0
|
|
|
|
and sql like '%VIRTUAL TABLE%USING FTS%'
|
|
|
|
'''
|
|
|
|
)
|
|
|
|
]
|
|
|
|
for t in tables.keys():
|
|
|
|
for fts_table in fts_tables:
|
|
|
|
if t == fts_table or t.startswith(fts_table):
|
|
|
|
tables[t]['hidden'] = True
|
|
|
|
continue
|
|
|
|
|
2017-11-13 15:20:02 +00:00
|
|
|
self._inspect[name] = {
|
|
|
|
'hash': m.hexdigest(),
|
|
|
|
'file': str(path),
|
|
|
|
'tables': tables,
|
2017-11-22 20:12:15 +00:00
|
|
|
'views': views,
|
|
|
|
|
2017-11-13 15:20:02 +00:00
|
|
|
}
|
|
|
|
return self._inspect
|
2017-11-10 19:05:57 +00:00
|
|
|
|
|
|
|
def app(self):
|
|
|
|
app = Sanic(__name__)
|
2017-11-30 16:30:10 +00:00
|
|
|
default_templates = str(app_root / 'datasette' / 'templates')
|
2017-11-30 16:05:01 +00:00
|
|
|
if self.template_dir:
|
2017-11-30 16:30:10 +00:00
|
|
|
template_loader = ChoiceLoader([
|
|
|
|
FileSystemLoader([self.template_dir, default_templates]),
|
|
|
|
# Support {% extends "default:table.html" %}:
|
|
|
|
PrefixLoader({
|
|
|
|
'default': FileSystemLoader(default_templates),
|
|
|
|
}, delimiter=':')
|
|
|
|
])
|
|
|
|
else:
|
|
|
|
template_loader = FileSystemLoader(default_templates)
|
2017-11-30 15:51:40 +00:00
|
|
|
self.jinja_env = Environment(
|
2017-11-30 16:30:10 +00:00
|
|
|
loader=template_loader,
|
2017-11-16 01:59:42 +00:00
|
|
|
autoescape=True,
|
2017-11-10 19:05:57 +00:00
|
|
|
)
|
2017-11-30 15:51:40 +00:00
|
|
|
self.jinja_env.filters['escape_css_string'] = escape_css_string
|
|
|
|
self.jinja_env.filters['quote_plus'] = lambda u: urllib.parse.quote_plus(u)
|
|
|
|
self.jinja_env.filters['escape_table_name'] = escape_sqlite_table_name
|
|
|
|
self.jinja_env.filters['to_css_class'] = to_css_class
|
2017-11-11 17:49:47 +00:00
|
|
|
app.add_route(IndexView.as_view(self), '/<as_json:(.jsono?)?$>')
|
2017-11-10 19:05:57 +00:00
|
|
|
# TODO: /favicon.ico and /-/static/ deserve far-future cache expires
|
|
|
|
app.add_route(favicon, '/favicon.ico')
|
|
|
|
app.static('/-/static/', str(app_root / 'datasette' / 'static'))
|
2017-12-03 16:33:36 +00:00
|
|
|
for path, dirname in self.static_mounts:
|
|
|
|
app.static(path, dirname)
|
2017-11-10 19:05:57 +00:00
|
|
|
app.add_route(
|
|
|
|
DatabaseView.as_view(self),
|
|
|
|
'/<db_name:[^/\.]+?><as_json:(.jsono?)?$>'
|
|
|
|
)
|
|
|
|
app.add_route(
|
|
|
|
DatabaseDownload.as_view(self),
|
|
|
|
'/<db_name:[^/]+?><as_db:(\.db)$>'
|
|
|
|
)
|
|
|
|
app.add_route(
|
|
|
|
TableView.as_view(self),
|
|
|
|
'/<db_name:[^/]+>/<table:[^/]+?><as_json:(.jsono?)?$>'
|
|
|
|
)
|
|
|
|
app.add_route(
|
|
|
|
RowView.as_view(self),
|
|
|
|
'/<db_name:[^/]+>/<table:[^/]+?>/<pk_path:[^/]+?><as_json:(.jsono?)?$>'
|
|
|
|
)
|
|
|
|
return app
|