Formulosity is a self-hosted app for building and deploying the surveys using code instead of traditional survey builders.
This approach offers a number of advantages, including:
- Version control: Survey files can be stored in a Github repository, which makes it easy to track changes and collaborate with others.
- Reproducibility: Survey files can be easily shared and reproduced, making it easy to reuse surveys or create variations of existing surveys.
- Automation: Survey files can be used to automate the creation and deployment of surveys, which can save time and effort.
Formulosity uses human-readable declarative language YAML.
- Management API
- Survey UI: for end users (respondents)
- Console UI: manage surveys
- YAML survey configuration
- Basic question types
- Default theme
- Custom themes support
- Personalized options: intro, outro, etc.
- Cookie/IP deduplication
- Admin user authentication
- Different database options: SQLite and Postgres
- Continue where you left off
- Advanced validation rules
- Detect survey changes in real time
Console UI - simple - custom_theme - short - many_options
Note: use user
/ pass
to login into the Console UI.
Each directory in SURVEYS_DIR
is a survey. You can configure the source of your surveys by setting different SURVEYS_DIR
env var.
surveys/
├── survey1/
│ ├── metadata.yaml
│ ├── questions.yaml
│ ├── security.yaml
│ ├── variables.yaml
│ └── ...
└── survey2/
├── metadata.yaml
├── questions.yaml
└── ...
To get started, check out the ./surveys
folder with multiple examples.
This file is required! The file consists of a YAML object with specific properties describing the survey.
- title: This is the main title displayed to users at the beginning of the survey.
- theme: This specifies the visual theme applied to the survey. Currently supported themes are: default.
- intro: This text appears as an introduction before the first question.
- outro: This text appears as a conclusion after the last question.
title: Survey Title
theme: default # or custom
intro: |
This is the introduction to the survey.
It can be multiple lines long.
outro: |
Thank you for taking the survey.
Your feedback is important to us.
This file is required! The file consists of a list of questions, each defined as a YAML object with specific properties.
- type: This specifies the type of question being asked.
- id: This provides a unique identifier for the question. It is useful for referencing specific questions in branching logic or data analysis. IDs must be unique across all questions in the survey.
- label: This is the text displayed to the user as the question itself.
- description: This provides additional information about the question for the user, such as clarification or instructions.
- options: This list defines the available answer choices for question types like
single-choice
,multiple-choice
andranking
. - optionsFromVariable: This property references a variable defined in a separate variables.yaml file. The variable should contain a list of options to be used for the question. This allows for reusability and centralized management of option lists.
- validation: This property is used to define validation rules for specific question types.
questions:
- type: single-choice
id: question1 # optional ID, must be unique across all questions
label: What is the capital of Germany?
description: You can select multiple options
optionsFromVariable: german-city-options # defined in variables.yaml
options:
- Berlin
- Munich
- Paris
- London
- Hamburg
- Cologne
validation:
min: 1
max: 3
This file is optional. The file consists of a YAML object with specific properties for survey security settings.
- duplicateProtection: This property defines how the platform handles duplicate responses from the same user.
duplicateProtection: cookie # cookie | ip
This file is optional. The file consists of a list of variables, each defined as a YAML object with specific properties.
- id: This unique identifier references the variable within questions. IDs must be unique across all variables defined in the file.
- type: This specifies the type of data stored in the variable. Currently supported types are: list.
variables:
- id: german-city-options # must be unique
type: list
options:
- Berlin
- Munich
- Hamburg
- Cologne
Prompts users for a brief written answer.
- type: short-text
label: What is the capital of Germany?
Prompts users for a detailed written answer.
- type: long-text
label: What is the capital of Germany?
Presents a question with only one correct answer from a list of options.
- type: single-choice
label: What is the capital of Germany?
options:
- Berlin
- Munich
- Paris
- London
- Hamburg
- Cologne
Presents a question where users can select multiple answers (with limitations). You can customize the minimum and maximum allowed selections in the validation section.
- type: multiple-choice
label: Which of the following are cities in Germany?
description: You can select multiple options
validation:
min: 1
max: 3
options:
- Berlin
- Munich
- Paris
- London
- Hamburg
- Cologne
Asks users to enter a specific date.
- type: date
label: When was the Berlin Wall built?
Presents a scale for users to rate something on a predefined range.
- type: rating
label: How much do you like Berlin?
min: 1
max: 5
Asks users to rank options based on a given criteria.
- type: ranking
label: Rank the following cities by population
optionsFromVariable: german-city-options
Presents a question where users can only answer "yes" or "no".
- type: yes-no
label: Is Berlin the capital of Germany?
docker-compose up -d --build
And you should be able to access the UI on http://localhost:3000 (default basic auth: user:pass
).
You can deploy individual services to any cloud provider:
- Go backend. It's packaged as a Docker container and can be deployed to any cloud provider.
- Next.js frontend. It's also packaged as a Docker container, but also can be deployed to Vercel or Netlify.
- [Optional] Postgres database. You can use managed Postgres services or deploy it yourself.
The demo service (links above) is deployed to Fly.io (Go, SQLite) and Vercel (Next.js) and are under the free tiers.
There is also a combined version of Go and Next.js in the same Docker container. You can run it with:
docker-compose -f compose-combined.yml up -d --build
API:
DATABASE_TYPE
-sqlite
orpostgres
DATABASE_URL
- Postgres or SQLite connection stringLOG_LEVEL
- Log level, e.g.info
SURVEYS_DIR
- Directory with surveys, e.g./root/surveys
. It's suggested to use mounted volume for this directory.
UI:
CONSOLE_API_ADDR_INTERNAL
- Internal address of the Go backend, e.g.http://api:8080
(could be the same asCONSOLE_API_ADDR
if UI amd API are not on the same network).CONSOLE_API_ADDR
- Public address of the Go backend.IRON_SESSION_SECRET
- Secret for session encryptionHTTP_BASIC_AUTH
- Format:user:pass
for basic auth (optional)
- Backend: Go, Postgres, SQLite
- UI: Next.js, Tailwind CSS
Make sure to install go-migrate first.
cd api
migrate create -dir migrations/postgres -ext sql -seq name
migrate create -dir migrations/sqlite -ext sql -seq name
cd api
make test
Pull requests, bug reports, and all other forms of contribution are welcomed and highly encouraged!