nembery / aframe Goto Github PK

The idea would be to allow someone to create screens and distribute them to be installed along with an aframe installation. For example, a salt state may install aframe, copy templates, actions, forms, and screens into an 'import' directory and have those available in the app as soon as it starts

this allows searching on screens across multiple labels. I.e. show all screens with language=english, app=blah, author=nate, etc

screen themes are currently hard coded into the app. It should be possible to import themes on startup, or allow link for an operator to upload a new theme. This should probably be a simple JSON object that describes the theme. Aframe will de-serialize the JSON theme object and write into an imported CSS file by the screen page.

For example: creating a form with the following label: Ubicación ends up with the label Ubicación after being exported and re-imported.

blank variables from text area input form widget when sanitize is no

should move this work to here: https://github.com/JNPRAutomate via a fork

file aframe/endpoints/plugins/FileDiscovery.py

While attempting to load a CSV file, I was getting an error at this part of the code:

    def load_instance_config(self, config):
        filename = config["file_path"]["value"]
        try:

Python was complaining about the index type that it can't be string and it should be integer.

I did the following modification in order to fix the problem but I don't know if this is the best way to solve. Hence, I am creating this issue instead of doing a pull merge.

The change I applied is this:

    def load_instance_config(self, config):
    print "**** HERE ***"
    print(dir(config))
    print "**** HERE ***"
        filename = config[0]["value"]

You can ignore the prints as I used them for troubleshooting.

we need a widget in the screens widget list that allows the operator to create a 'menu' of selected other screens. This would allow the creation of 'apps' that are comprised of several different screens. Updating the widget on one screen would update the widget on all screens.

also import screens on first startup to allows for packaging templates, forms, and screens during an aframe install

forms that use preloaded values from automation or preloaded list from automation do not get embedded in a screen export correctly

orange and white is just ok, need something a bit better

We need a websocket widget in the screens section. The websocket could potentially use the django channels support to connect directly to some external source via a proxy set up in django/aframe server

My template is configure the static route:

routing-options {
  static {
    route {{ prefix }} next-hop {{ nexthop }};
  }
}

The text format regex can't match above template, and therefore fallback to default "set" style and make the apply config failed.

To fix it, suggest to match the "-" also:

re.search(r"^[a-z:-]*\s*\w+\s+{", conf_string, re.I)

Should replace CMD in dockerfile with a simple shell script the detects an ENV variable. If something like USE_SSL is present, then issue the ./manage.py runsslserver command instead of ./manage.py runserver

ref: https://github.com/teddziuba/django-sslserver

Click 'Add' on custom headers in basic rest action throws a js error: add_kv_list not defined.

inventory file should be populated by queue or selected host and the 'template' should populate the group vars file. action options should determine the local directory of the ansible playbook

first plugin should be the current yaml file, with another being the Vault API

https://hub.docker.com/_/vault/

useful in device configurations where they require an md5 hash of a password instead of plaintext

config should take a URL
output should just be a styled Iframe with option to open url in new window

exception when attempting to clone an automation that has multiple input_forms (like when multiple languages are used)

Clone logic looks for a single input form from a template_id. The assumption is that templates always have a single input_form, which isn't always true. Need to fix the clone logic to use the input_form id and pull a single template from there, not vice versa

currently there is no paging support for these pages, proposal would be to use the django framework paging support to add these features. Would need to work with the proposed git backend as well

currently, all templates, forms, and screens are kept in a local sqlite3 db. We should add support for a git repository backend so all those can be saved into a directory structure in git. Will also allow pointing new installations at existing repositories for examples that we may distribute

input form list, template list, endpoint list, creation pages for each should all be available as widgets in aframe. Low priority items, but we make these pages just different 'screens' that the user can navigate to