Flamyngo is a YAML-powered Flask frontend for MongoDB. The aim is to delegate most settings to a YAML configuration file, which then allows the underlying code to be reused for any conceivable collection. The official web page is at https://materialsvirtuallab.github.io/flamyngo/.
Complete change log: link.
Install via pip:
pip install flamyngo
Create your .flamyngo.yaml
.
flm --config <path/to/config.yaml>
If --config
is not provided, it defaults to ~/.flamyngo.yaml
.
A sample commented configuration yaml file is given below. You can start from the one below and customize it to suit your needs.
# Provide an optional title for your app.
title: "My Flamyngo App"
# Provide some optional help text (html format) for the query.
help: "Supported queries: last name (string)"
# Uncomment the parameter below to provide an optional template folder, which
# will be passed into the Flamyngo app. Path to the template folder should be
# specified relative to localtion where flm is run. If not provided, the
# default provided in the flamyngo.templates will be used. It is highly
# recommended that you start from the default provided and make only stylistic
# changes. It is imperative you do not change the variable names in the Jinja
# templates or Flamyngo will not work.
# template_folder: my_templates
# MongoDB settings
db:
host: mongo.host.com
port: 27017
username: user
password: password
database: mydb
# Alternatively, MongoDB settings can just be provided as a connection string.
# dnspython must be installed if you are using the connection string method.
# connection_string: "mongodb+srv://user:password@mydb.mongodb.net/"
# List of collection settings. Note that more than one collection is supported,
# though only one collection can be queried at any one time.
collections:
-
name: mycoll
# These set the special queries as an ordered list of [<key>, <regex string>, <type>].
# If the query string satisfies any of the regex, the Mongo query is set as
# {<key>: type(<search_string>)}. This allows for much more friendly setups for common
# queries that do not require a user to type in verbose Mongo criteria. Each
# regex should be uniquely identifying.
# Types can be any kind of callable function that takes in a string and return
# a value without other arguments. E.g., int, str, float, etc. You can support
# more powerful conversions by writing your own processing function, e.g.,
# mymodule.convert_degress_to_radians.
# If none of the regex works, the criteria is interpreted as a Mongo-like dict query.
query:
- [last_name, '^[A-Za-z]+$', str]
- [phone_number, '^[0-9]+$', int]
# A default list of projection key, processing function to display as a table.
# Again, processing function can be any callable, and you can define your own.
# You can also supply any Python formatting string (starts with %) as the processing
# function. For example, "%.1f" would format that quantity as a float with one
# decimal.
summary:
- [_id, str]
- [first_name, str]
- [last_name, str]
- [phone_number, str]
- [age, "%d"]
# Aliases for various fields. These are used to display short names in the summary
# table. You can also directly perform queries using the short names instead of
# using the long names.
aliases:
phone_number: number
# Initial sorting for summary. Use asc for ascending and desc for descending.
# Note that the aliased name (if any) should be used for sorting.
sort: [last_name, asc]
# The following defines unique identifiers for each doc. This allows each
# specific doc to be queried and displayed using this key. If this key is
# present in the default list of projections, a link will be created to each
# unique document.
unique_key: _id
unique_key_type: bson.objectid.ObjectId
# The following defines keys to exclude from the doc view.
# This is sometimes useful (or necessary) to reduce the size of the
# individual documents being viewed (for very large documents).
# This only affects the doc view.
doc_exclude:
- key_to_exclude
# Basic auth can be set up by specifying user and password below. If these are not
# set, then no authentication. Note that this is not the most secure. It is merely
# used for a basic setup. For high security, look into proper implementations.
AUTH_USER: Iam
AUTH_PASSWD: Pink
API_KEY: IamPink
Assuming that you are running on local host at port 5000, the initial landing page will be at http://localhost:5000.
Pages for individual docs following the format http://localhost:5000/[collection_name]/doc/[unique_id].
- Getting all unique ids for a collection: http://localhost:5000/[collection_name]/unique_ids
- Getting individual docs as a json response: http://localhost:5000/[collection_name]/doc/[unique_id]/json.
- Getting a field of an individual doc as a plain text response: http://localhost:5000/[collection_name]/doc/[unique_id]/.