a couchdb user space (FUSE) file system plus a cli for treating couchdb databases as a file system drives
- Free software: MIT license
- couchdb
- FUSE
- a simple list command,
lsstyle, for your attachments in couchdb - an upload command,
cp -r dir couchdb_docstyle, for storing directory hierarchy as attachments in couchdb - a download command,
cp -r couchdb_doc dirstyle, for dumping attachments from couchdb - fuse file system based on couchdb
- setup your environmental variable with the right credentials
export COUCHDB_URI='couchdb://username:password@host:port/database'
pip install git+https://github.com/thanos/couchfs.git
- prime your database
% couchfs init
checking the db
creating or updating the db _design/couchfs_views
db is now setup
Now you can use the simpe command line tools or the api.
couchfs ls let's you list the attachements in a couchdb database.
Just running ls without and path will give you the atatchments in every document in your database.
export CouchDBClient_URI=couchdb://xxx:[email protected]:5984/myrepo
% couchfs ls
TAKIS/takis/takis/asgi.py 387 Bytes
TAKIS/takis/takis/settings.py 3.8 kB
TAKIS_1/takis/takis/urls.py 470 BytesWhere TAKIS, TAKIS_1, TAKIS_2 are the ids of different documents.
You can spec a path starting with the document id.
% couchfs ls TAKIS_2
TAKIS_2/takis/takis/wsgi.py 470 BytesThis allows to upload and attach files documents to couchdb document. It follows closely the sematics of GNU cp -R.
So when doing couchfs upload source destination, source will match what you want to copy, going recursively, and desitnation is a path starting with the document id.
% couchfs upload ~/takis/all.json /TAKIS_1/
/Users/thanos/all.json http://127.0.0.1:5984/test/TAKIS_1/TAKIS_1/all.json 201:Created
There is a --dry_run flag that can save your ass!
% couchfs upload ~/takis/takis /TAKIS_1/
/Users/thanos/takis/takis/media/dl131.jpg /TAKIS_1/takis/media/dl131.jpg DRY RUN:
/Users/thanos/takis/takis/media/vs233.jpg /TAKIS_1/takis/media/vs233.jpg DRY RUN:
/Users/thanos/takis/takis/media/tavsi098.jpg /TAKIS_1/takis/media/tavsi098.jpg DRY RUN:
/Users/thanos/takis/takis/media/mw042_on6H5f6.jpg /TAKIS_1/takis/media/mw042_on6H5f6.jpg DRY RUN:
/Users/thanos/takis/takis/media/t126.jpg /TAKIS_1/takis/media/t126.jpg DRY RUN:
/Users/thanos/takis/takis/media/tad040.jpg /TAKIS_1/takis/media/tad040.jpg DRY RUN:This allows to download files that are attached to couchdb documents. It follows closely the sematics of GNU cp -R.
So when doing couchfs down source destination, source will docuemnt id followed by a synthetic
path, and desitnation is a local path and its directory tree will be created as necassery. Again there is a useful
--dry_run flag.
% couchfs upload ~/takis/takis /TAKIS_1/
/Users/thanos/takis/takis/media/dl131.jpg /TAKIS_1/takis/media/dl131.jpg DRY RUN:
/Users/thanos/takis/takis/media/vs233.jpg /TAKIS_1/takis/media/vs233.jpg DRY RUN:
/Users/thanos/takis/takis/media/tavsi098.jpg /TAKIS_1/takis/media/tavsi098.jpg DRY RUN:
/Users/thanos/takis/takis/media/01-MOUSEIO_KATO_202002191257-00_02_01.png /TAKIS_1/takis/media/01-MOUSEIO_KATO_202002191257-00_02_01.png DRY RUN:
/Users/thanos/takis/takis/media/d023.jpg /TAKIS_1/takis/media/d023.jpg DRY RUN:
/Users/thanos/takis/takis/media/mw042_on6H5f6.jpg /TAKIS_1/takis/media/mw042_on6H5f6.jpg DRY RUN:
/Users/thanos/takis/takis/media/t126.jpg /TAKIS_1/takis/media/t126.jpg DRY RUN:
/Users/thanos/takis/takis/media/tad040.jpg /TAKIS_1/takis/media/tad040.jpg DRY RUN:%couchfs download TAKIS/takis/takis/*.py dump
hhttp://127.0.0.1:5984/test/TAKIS/takis/takis/__init__.py dump/__init__.py 200:OK
http://127.0.0.1:5984/test/TAKIS/takis/takis/asgi.py dump/asgi.py 200:OK
http://127.0.0.1:5984/test/TAKIS/takis/takis/settings.py dump/settings.py 200:OK
http://127.0.0.1:5984/test/TAKIS/takis/takis/urls.py dump/urls.py 200:OK
http://127.0.0.1:5984/test/TAKIS/takis/takis/wsgi.py dump/wsgi.py 200:OKOption 1 set up an environment variable is the form
export COUCHDB_URI='couchdb://username:password@host:port/database'Or pass the connection string to CouchDBClient
from couchfs.api import CouchDBClient
connection_str = 'couchdb://username:password@host:port/database'
CouchDBClient(connection_str)if you whant to use another environment key you can always subclass CouchDBClient
from couchfs.api import CouchDBClient
class MyClient(CouchDBClient):
URI_ENVIRON_KEY='MY_COUCHDB_URI'You can use CouchDBClient().get_attachment to get a file handle on couchdb attachment. Attachments can be very big
so they are streamed down into a temporary file that is released when you exit the context.
from couchfs.api import CouchDBClient
for url, size in CouchDBClient().list_attachments():
print(url, size)
for url, size in CouchDBClient().list_attachments('TV', ):
print(url, size)
for url, size in CouchDBClient().list_attachments('TV', 'TAKIS'):
print(url, size)
for url, size in CouchDBClient().list_attachments('TV', 'TAKIS/iamges/*.jpeg'):
print(url, size)You can use CouchDBClient().get_attachment to get a file handle on couchdb attachment. Attachments can be very big
so they are streamed down into a temporary file that is released when you exit the context.
import csv
from requests.exceptions import RequestException
from couchfs.api import CouchDBClient, CouchDBClientException
doc_id = 'TAKIS'
doc_name = 'takis/takis/catelog.csv'
couch_uri ='couchdb://user:****@127.0.0.1:5984/test'
try:
with CouchDBClient(couch_uri).download_file(f'{doc_id}/{doc_name}') as csvfile:
reader = csv.reader(csvfile)
for row in reader:
print(row)
except (CouchDBClientException, RequestException) as error:
print(error)For those who like to use their memory...
import json
from couchfs.api import CouchDBClient
doc_id = 'TAKIS'
doc_name = 'takis/takis/catelog.json'
with CouchDBClient().download_file(f'{doc_id}/{doc_name}', in_memory=True) as somejson:
expenses = json.loads(somejson)For those who like to use their memory...
import json
from couchfs.api import CouchDBClient
doc_id = 'TAKIS'
doc_name = 'takis/takis/catelog.json'
with CouchDBClient().download_file(f'{doc_id}/{doc_name}', in_memory=True) as somejson:
expenses = json.loads(somejson)import json
from couchfs.api import CouchDBClient
do_something = lambda x: x
doc_id = 'TAKIS'
for url in CouchDBClient().attachments(f'{doc_id}/*.json'):
with CouchDBClient().download_file(url, in_memory=True) as somejson:
do_something(somejson)import json
from couchfs.api import CouchDBClient
doc_id = 'TAKIS'
attachments_path='/takis/takis'
destination = '/var/data'
CouchDBClient().download(f'{doc_id}/{attachments_path}', destination)from couchfs.api import CouchDBClient
doc_id = 'TAKIS'
attachment_path='/takis/takis'
some_file = 'takis/takis/catelog.json'
with open(some_file, 'rb') as fp:
CouchDBClient().upload_file(fp, f'{doc_id}/{attachment_path}')import json
from couchfs.api import CouchDBClient
my_data = {
'name': 'thanos'
}
doc_id = 'TAKIS'
attachment_file_name = 'takis/takis/catelog.json'
CouchDBClient().upload_bytes_file(bytes(json.dumps(my_data), 'utf-8'), f'{doc_id}/{attachment_file_name}')import json
from couchfs.api import CouchDBClient
source='/User/thanos/takis'
doc_id = 'TAKIS'
CouchDBClient().upload(source, doc_id)
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent