comment engine
Switch branches/tags
Nothing to show
Clone or download

README.md

remark42 Build Status Go Report Card Coverage Status

Remark42 is a self-hosted, lightweight, and simple (yet functional) comment engine, which doesn't spy on users. It can be embedded into blogs, articles or any other place where readers add comments.

  • Social login via Google, Facebook, Github and Yandex
  • Multi-level nested comments with both tree and plain presentations
  • Import from disqus and wordpress
  • Markdown support
  • Moderator can remove comments and block users
  • Voting, pinning and verification system
  • Sortable comments
  • Extractor for recent comments, cross-post
  • RSS for all comments and each post
  • Export data to json with automatic backups
  • No external databases, everything embedded in a single data file
  • Fully dockerized and can be deployed in a single command
  • Self-contained executable can be deployed directly to Linux, Windows and MacOS
  • Clean, lightweight and fully customizable UI
  • Multi-site mode from a single instance
  • Integration with automatic ssl via nginx-le
  • Privacy focused

Install

Backend

With Docker

this is the recommended way to run remark42

  • copy provided docker-compose.yml and customize for your needs
  • make sure you don't keep DEV_PASSWD=something... for any non-development deployments
  • pull prepared images from the docker hub and start - docker-compose pull && docker-compose up -d
  • alternatively compile from the sources - docker-compose build && docker-compose up -d

Without docker

  • download archive for stable release or development version
  • unpack with gunzip (Linux, macOS) or with zip (Windows)
  • run as remark42.{os}-{arch} server {parameters...}, i.e. remark42.linux-amd64 server --secret=12345 --url=http://127.0.0.1:8080
  • alternatively compile from the sources - make OS=[linux|darwin|windows] ARCH=[amd64,386,arm64,arm32]

Parameters

Command line Environment Default Description
url REMARK_URL url to remark42 server, required
secret SECRET secret key, required
site SITE remark site name(s), multi
store.type STORE_TYPE bolt type of storage, bolt or mongo
store.bolt.path STORE_BOLT_PATH ./var path to data directory
store.bolt.timeout STORE_BOLT_TIMEOUT 30s boltdb access timeout
mongo.url MONGO_URL mongo url for all stores using mongodb
mongo.db MONGO_DB mongo database
admin.shared.id ADMIN_SHARED_ID admin names (list of user ids), multi
admin.shared.email ADMIN_SHARED_EMAIL admin@${REMARK_URL} admin email
backup BACKUP_PATH ./var/backup backups location
max-back MAX_BACKUP_FILES 10 max backup files to keep
cache.max.items CACHE_MAX_ITEMS 1000 max number of cached items, 0 - unlimited
cache.max.value CACHE_MAX_VALUE 65536 max size of cached value, 0 - unlimited
cache.max.size CACHE_MAX_SIZE 50000000 max size of all cached values, 0 - unlimited
avatar.type AVATAR_TYPE fs type of avatar storage, fs, 'bolt, or mongo`
avatar.fs.path AVATAR_FS_PATH ./var/avatars avatars location for fs store
avatar.bolt.file AVATAR_BOLT_FILE ./var/avatars.db file name for bolt store
avatar.rsz-lmt AVATAR_RSZ_LMT 0 max image size for resizing avatars on save
auth.ttl.jwt AUTH_TTL_JWT 5m jwt TTL
auth.ttl.cookie AUTH_TTL_COOKIE 200h cookie TTL
auth.google.cid AUTH_GOOGLE_CID Google OAuth client ID
auth.google.csec AUTH_GOOGLE_CSEC Google OAuth client secret
auth.facebook.cid AUTH_FACEBOOK_CID Facebook OAuth client ID
auth.facebook.csec AUTH_FACEBOOK_CSEC Facebook OAuth client secret
auth.github.cid AUTH_GITHUB_CID Github OAuth client ID
auth.github.csec AUTH_GITHUB_CSEC Github OAuth client secret
auth.yandex.cid AUTH_YANDEX_CID Yandex OAuth client ID
auth.yandex.csec AUTH_YANDEX_CSEC Yandex OAuth client secret
auth.dev AUTH_DEV false local oauth2 server, development mode only
notify.type NOTIFY_TYPE none type of notification (none or telegram)
notify.queue NOTIFY_QUEUE 100 size of notification queue
notify.telegram.token NOTIFY_TELEGRAM_TOKEN telegram token
notify.telegram.chan NOTIFY_TELEGRAM_CHAN telegram channel
notify.telegram.timeout NOTIFY_TELEGRAM_TIMEOUT telegram timeout
ssl.type SSL_TYPE none https support, none - http, static - https
ssl.port SSL_PORT 8443 port for https server
ssl.cert SSL_CERT path to cert.pem file
ssl.key SSL_KEY path to key.pem file
max-comment MAX_COMMENT_SIZE 2048 comment's size limit
low-score LOW_SCORE -5 low score threshold
critical-score CRITICAL_SCORE -10 critical score threshold
edit-time EDIT_TIME 5m edit window
img-proxy IMG_PROXY false enable http->https proxy for images
dbg DEBUG false debug mode
dev-passwd DEV_PASSWD password for dev user
  • command line parameters are long form --<key>=value, i.e. --site=https://demo.remark42.com
  • multi parameters separated by , in the environment or repeated with command line key, like --site=s1 --site=s2 ...
  • required parameters have to be presented in the environment or provided in command line
Required parameters

Most of the parameters have sane defaults and don't require customization. There are only a few parameters user has to define:

  1. SECRET - secret key, can be any long and hard-to-guess string.
  2. REMARK_URL - url pointing to your remark42 server, i.e. https://demo.reamark42.com
  3. At least one pair of AUTH_<PROVIDER>_CID and AUTH_<PROVIDER>_CSEC defining oauth2 provider(s)

The minimal docker-compose.yml has to include all required parameters:

version: '2'

services:
    remark42:
        image: umputun/remark42:latest
        restart: always
        container_name: "remark42"
        environment:
            - REMARK_URL=https://demo.remark42.com  # url pointing to your remark42 server
            - SITE=YOUR_SITE_ID                     # site ID, same as used for `site_id`, see "Setup on your website"
            - SECRET=abcd-123456-xyz-$%^&           # secret key
            - AUTH_GITHUB_CID=12345667890           # oauth2 client ID
            - AUTH_GITHUB_CSEC=abcdefg12345678      # oauth2 client secret
        volumes:
            - ./var:/srv/var                        # persistent volume to store all remark42 data 

Register oauth2 providers

Authentication handled by external providers. You should setup oauth2 for all (or some) of them to allow users to make comments. It is not mandatory to have all of them, but at least one should be correctly configured.

Google Auth Provider
  1. Create a new project: https://console.developers.google.com/project
  2. Choose the new project from the top right project dropdown (only if another project is selected)
  3. In the project Dashboard center pane, choose "API Manager"
  4. In the left Nav pane, choose "Credentials"
  5. In the center pane, choose "OAuth consent screen" tab. Fill in "Product name shown to users" and hit save.
  6. In the center pane, choose "Credentials" tab.
    • Open the "New credentials" drop down
    • Choose "OAuth client ID"
    • Choose "Web application"
    • Application name is freeform, choose something appropriate
    • Authorized origins is your domain ex: https://remark42.mysite.com
    • Authorized redirect URIs is the location of oauth2/callback constructed as domain + /auth/google/callback, ex: https://remark42.mysite.com/auth/google/callback
    • Choose "Create"
  7. Take note of the Client ID and Client Secret

instructions for google oauth2 setup borrowed from oauth2_proxy

GitHub Auth Provider
  1. Create a new "OAuth App": https://github.com/settings/developers
  2. Fill "Application Name" and "Homepage URL" for your site
  3. Under "Authorization callback URL" enter the correct url constructed as domain + /auth/github/callback. ie https://remark42.mysite.com/auth/github/callback
  4. Take note of the Client ID and Client Secret
Facebook Auth Provider
  1. From https://developers.facebook.com select "My Apps" / "Add a new App"
  2. Set "Display Name" and "Contact email"
  3. Choose "Facebook Login" and then "Web"
  4. Set "Site URL" to your domain, ex: https://remark42.mysite.com
  5. Under "Facebook login" / "Settings" fill "Valid OAuth redirect URIs" with your callback url constructed as domain + /auth/facebook/callback
  6. Select "App Review" and turn public flag on. This step may ask you to provide a link to your privacy policy.
Yandex Auth Provider
  1. Create a new "OAuth App": https://oauth.yandex.com/client/new
  2. Fill "App name" for your site
  3. Under Platforms select "Web services" and enter "Callback URI #1" constructed as domain + /auth/yandex/callback. ie https://remark42.mysite.com/auth/yandex/callback
  4. Select Permissions. You need following permissions only from the "Yandex.Passport API" section:
    • Access to user avatar
    • Access to username, first name and surname, gender
  5. Fill out the rest of fields if needed
  6. Take note of the ID and Password

For more details refer to Yandex OAuth and Yandex.Passport API documentation.

Initial import from Disqus

  1. Disqus provides an export of all comments on your site in a g-zipped file. This is found in your Moderation panel at Disqus Admin > Setup > Export. The export will be sent into a queue and then emailed to the address associated with your account once it's ready. Direct link to export will be something like https://<siteud>.disqus.com/admin/discussions/export/. See importing-exporting for more details.
  2. Move this file to your remark42 host within ./var and unzip, i.e. gunzip <disqus-export-name>.xml.gz.
  3. Run import command - docker exec -it remark42 import -p disqus -f {disqus-export-name}.xml -s {your site id}

Initial import from WordPress

  1. Install WordPress plugin to export comments and follow it instructions. The plugin should produce a xml-based file with site content including comments.
  2. Move this file to your remark42 host within ./var
  3. Run import command - docker exec -it remark42 import -p wordpress -f {wordpress-export-name}.xml -s {your site id}

Backup and restore

Automatic backups

Remark42 by default makes daily backup files under ${BACKUP_PATH} (default ./var/backup). Backups kept up to ${MAX_BACKUP_FILES} (default 10). Each backup file contains exported and gzipped content, i.e., all comments. At any point, the user can restore such backup and revert all comments to the desirable state. Note: restore procedure cleans the current data store and replaces all comments with comments from the backup file.

For safety and security reasons restore functionality not exposed outside of your server by default. The recommended way to restore from the backup is to use provided scripts/restore-backup.sh. It can run inside the container:

docker exec -it remark42 restore -f {backup-filename.gz} -s {your site id}

Manual backup

In addition to automatic backups user can make a backup manually. This command makes userbackup-{site id}-{timestamp}.gz by default.

docker exec -it remark42 backup -s {your site id}

Restore from backup

Restore will clean all comments first and then will processed with complete import from a given file.

docker exec -it remark42 restore -f {backup file name} -s {your site id}

Backup format

Backup file is a text file with all exported comments separated by EOL. Each backup record is a valid json with all key/value unmarshaled from Comment struct (see below).

Admin users

Admins/moderators should be defined in docker-compose.yml as a list of user IDs or passed in the command line.

    environment:
        - ADMIN=github_ef0f706a79cc24b17bbbb374cd234a691a034128,github_dae9983158e9e5e127ef2b87a411ef13c891e9e5

To get user id just login and click on your username or any other user you want to promote to admins. It will expand login info and show full user ID.

Setup on your website

Comments

It's a main widget which renders list of comments.

Add this snippet to the bottom of web page:

<script>
  var remark_config = {
    site_id: 'YOUR_SITE_ID',
    url: 'PAGE_URL', // optional param; if it isn't defined window.location.href will be used
    max_shown_comments: 10, // optional param; if it isn't defined default value (15) will be used 
  };

  (function() {
    var d = document, s = d.createElement('script');
    s.src = '/web/embed.js'; // prepends this address with domain where remark42 is placed
    (d.head || d.body).appendChild(s);
  })();
</script>

And then add this node in the place where you want to see Remark42 widget:

<div id="remark42"></div>

After that widget will be rendered inside this node.

Last comments

It's a widget which renders list of last comments from your site.

Add this snippet to the bottom of web page:

<script>
  var remark_config = {
    site_id: 'YOUR_SITE_ID', 
  };

  (function() {
    var d = document, s = d.createElement('script');
    s.src = '/web/last-comments.js'; // prepends this address with domain where remark42 is placed
    (d.head || d.body).appendChild(s);
  })();
</script>

And then add this node in the place where you want to see last comments widget:

<div class="remark42__last-comments" data-max="50"></div>

data-max sets the max amount of comments (default: 15).

Counter

It's a widget which renders a number of comments for the specified page.

Add this snippet to the bottom of web page:

<script>
  var remark_config = {
    site_id: 'YOUR_SITE_ID',
  };

  (function() {
    var d = document, s = d.createElement('script');
    s.src = '/web/counter.js'; // prepends this address with domain where remark42 is placed
    (d.head || d.body).appendChild(s);
  })();
</script>

And then add a node like this in the place where you want to see a number of comments:

<span class="remark42__counter" data-url="https://domain.com/path/to/article/"></span>

You can use as many nodes like this as you need to. The script will found all them by the class remark__counter, and it will use data-url attribute to define the page with comments.

Also script can uses url property from remark_config object, or window.location.href if nothing else is defined.

Build from the source

  • to build docker container - make docker. This command will produce container umputun/remark42.
  • to build a single binary for direct execution - make OS=<linux|windows|darwin> ARCH=<amd64|386>. This step will produce executable remark42 file with everything embedded.

Development

You can use fully functional local version to develop and test both frontend & backend.

To bring it up run:

# if you mainly work on backend
docker-compose -f compose-dev-backend.yml build
docker-compose -f compose-dev-backend.yml up
# if you mainly work on frontend
docker-compose -f compose-dev-frontend.yml build
docker-compose -f compose-dev-frontend.yml up

It starts Remark42 on 127.0.0.1:8080 and adds local OAuth2 provider “Dev”. To access UI demo page go to 127.0.0.1:8080/web. By default, you would be logged in as dev_user which defined as admin. You can tweak any of supported parameters in corresponded yml file.

Backend docker compose config by default skips running frontend related tests. Frontend docker compose config by default skips running backend related tests and sets NODE_ENV=development for frontend build.

Backend development

In order to run backend locally (development mode, without docker) you have to have latest stable go toolchain installed.

To run backend - go run backend/app/main.go --dbg --secret=12345 --dev-passwd=password --site=remark --url=http://127.0.0.1:8080 It stars backend service with embedded bolt store on port 8080 with basic auth, allowing to authenticate and run requests directly, like this: HTTP http://dev:password@127.0.0.1:8080/api/v1/find?site=remark&sort=-active&format=tree&url=http://127.0.0.1:8080

To run backend with mongodb store mongo container should be started first - docker run -d -p 27017:27017 -name=mongo mongo:3.6 --smallfiles and then go run backend/app/main.go --dbg --secret=12345 --dev-passwd=password --site=remark --url=http://127.0.0.1:8080 --store.type=mongo --store.mongo.url=localhost

Frontend development

Build

  • install Node.js 8 or higher;
  • install NPM 6.1.0;
  • run npm install inside ./web;
  • run npm run build there;
  • result files will be saved in ./web/public.

Note Running npm install will set up precommit hooks into your git repository. It used to reformat your frontend code using prettier and lint with eslint before every commit.

Devserver

For local development mode with Hot Reloading use npm start instead of npm run build. In this case webpack will serve files using webpack-dev-server on localhost:9000. By visiting 127.0.0.1:9000/web you will get a page with main comments widget communicating with demo server backend running on https://demo.remark42.com. But you will not be able to login with any oauth providers due to security reasons.

You can attach to locally running backend by providing REMARK_URL environment variable.

npx cross-env REMARK_URL=http://127.0.0.1:8080 npm start

Developer build running by webpack-dev-server supports devtools for React and Redux.

API

Authorization

  • GET /auth/{provider}/login?from=http://url&site=site_id&session=1 - perform "social" login with one of supported providers and redirect to url. Presence of session (any non-zero value) change the default cookie expiration and makes them session-only.
  • GET /auth/logout - logout
type User struct {
    Name    string `json:"name"`
    ID      string `json:"id"`
    Picture string `json:"picture"`
    Admin   bool   `json:"admin"`
    Blocked bool   `json:"block"`
    Verified bool  `json:"verified"`
}

currently supported providers are google, facebook, github and yandex

Commenting

  • POST /api/v1/comment - add a comment. auth required
type Comment struct {
    ID        string          `json:"id"`      // comment ID, read only
    ParentID  string          `json:"pid"`     // parent ID
    Text      string          `json:"text"`    // comment text, after md processing
    Orig      string          `json:"orig"`    // original comment text
    User      User            `json:"user"`    // user info, read only
    Locator   Locator         `json:"locator"` // post locator
    Score     int             `json:"score"`   // comment score, read only
    Votes     map[string]bool `json:"votes"`   // comment votes, read only
    Timestamp time.Time       `json:"time"`    // time stamp, read only
    Pin       bool            `json:"pin"`     // pinned status, read only
    Delete    bool            `json:"delete"`  // delete status, read only
}

type Locator struct {
    SiteID string `json:"site"`     // site id
    URL    string `json:"url"`      // post url
}
  • POST /api/v1/preview - preview comment in html. Body is Comment to render

  • GET /api/v1/find?site=site-id&url=post-url&sort=fld&format=tree|plain - find all comments for given post

This is the primary call used by UI to show comments for given post. It can return comments in two formats - plain and tree. In plain format result will be sorted list of Comment. In tree format this is going to be tree-like object with this structure:

type Tree struct {
    Nodes []Node `json:"comments"`
    Info  store.PostInfo `json:"info,omitempty"`
}

type Node struct {
    Comment store.Comment `json:"comment"`
    Replies []Node        `json:"replies,omitempty"`
}

Sort can be time, active or score. Supported sort order with prefix -/+, i.e. -time. For tree mode sort will be applied to top-level comments only and all replies always sorted by time.

  • PUT /api/v1/comment/{id}?site=site-id&url=post-url - edit comment, allowed once in EDIT_TIME minutes since creation. Body is EditRequest json
 	type EditRequest struct {
 		Text    string `json:"text"`    // updated text
 		Summary string `json:"summary"` // optional, summary of the edit
 		Delete  bool   `json:"delete"`  // delete flag 
 	}{}
  • GET /api/v1/last/{max}?site=site-id - get up to {max} last comments

  • GET /api/v1/id/{id}?site=site-id - get comment by comment id

  • GET /api/v1/comments?site=site-id&user=id&limit=N - get comment by user id, returns response object

    type response struct {
        Comments []store.Comment  `json:"comments"`
        Count    int              `json:"count"`
    }{}
  • GET /api/v1/count?site=site-id&url=post-url - get comment's count for {url}

  • POST /api/v1/count?site=siteID - get number of comments for posts from post body (list of post IDs)

  • GET /api/v1/list?site=site-id&limit=5&skip=2 - list commented posts, returns array or PostInfo, limit=0 will return all posts

    type PostInfo struct {
        URL   string      `json:"url"`
        Count int         `json:"count"`
        ReadOnly bool     `json:"read_only,omitempty"`
        FirstTS time.Time `json:"first_time,omitempty"`
        LastTS  time.Time `json:"last_time,omitempty"`
    }
  • GET /api/v1/user - get user info, auth required

  • PUT /api/v1/vote/{id}?site=site-id&url=post-url&vote=1 - vote for comment. vote=1 will increase score, -1 decrease. auth required

  • GET /api/v1/userdata?site=site-id - export all user data to gz stream auth required

  • POST /api/v1/deleteme?site=site-id - request deletion of user data. auth required

  • GET /api/v1/config?site=site-id - returns configuration (parameters) for given site

    type config struct {
        Version       string   `json:"version"`
        EditDuration  int      `json:"edit_duration"` // seconds
        Admins        []string `json:"admins"`
        Auth          []string `json:"auth_providers"`
        LowScore      int      `json:"low_score"`
        CriticalScore int      `json:"critical_score"`
    }
  • GET /api/v1/info?site=site-idd&url=post-ur - returns PostInfo for site and url

RSS feeds

  • GET /api/v1/rss/post?site=site-id&url=post-url - rss feed for a post
  • GET /api/v1/rss/site?site=site-id - rss feed for given site
  • GET /api/v1/rss/reply?site=site-id&user=user-id - rss feed for replies to user's comments

Admin

  • DELETE /api/v1/admin/comment/{id}?site=site-id&url=post-url - delete comment by id.
  • PUT /api/v1/admin/user/{userid}?site=site-id&block=1&ttl=7d - block or unblock user with optional ttl (default=permanent)
  • GET api/v1/admin/blocked&site=site-id - list of blocked user ids
    type BlockedUser struct {
        ID        string    `json:"id"`
        Name      string    `json:"name"`
        Until     time.Time `json:"time"`
    }
  • GET /api/v1/admin/export?site=side-id&mode=[stream|file] - export all comments to json stream or gz file.
  • POST /api/v1/admin/import?site=side-id - import comments from the backup.
  • PUT /api/v1/admin/pin/{id}?site=site-id&url=post-url&pin=1 - pin or unpin comment.
  • GET /api/v1/admin/user/{userid}?site=site-id - get user's info.
  • DELETE /api/v1/admin/user/{userid}?site=site-id - delete all user's comments.
  • PUT /api/v1/admin/readonly?site=site-id&url=post-url&ro=1 - set read-only status
  • PUT /api/v1/admin/verify/{userid}?site=site-id&verified=1 - set verified status
  • GET /api/v1/admin/deleteme?token=token - process deleteme user's request

all admin calls require auth and admin privilege

Privacy

  • Remark42 is trying to be very sensitive to any private or semi-private information.
  • Authentication requesting the minimal possible scope from authentication providers. All extra information returned by them dropped immediately and not stored in any form.
  • Generally, remark42 keeps user id, username and avatar link only. None of these fields exposed directly - id and name hashed, avatar proxied.
  • There is no tracking of any sort.
  • Login mechanic uses JWT stored in a cookie (httpOnly, secured). The second cookie (XSRF_TOKEN) is a random id preventing CSRF.
  • There is no cross-site login, i.e., user's behavior can't be analyzed across independent sites running remark42.
  • There are no third-party analytic services involved.
  • User can request all information remark42 knows about and export to gz file.
  • Supported complete cleanup of all information related to user's activity.
  • Cookie lifespan can be restricted to session-only.
  • All potentially sensitive data stored by remark42 hashed and encrypted.

Technical details

  • Data stored in boltdb (embedded key/value database) files under STORE_BOLT_PATH
  • Each site stored in a separate boltbd file.
  • In order to migrate/move remark42 to another host boltbd files as well as avatars directory AVATAR_FS_PATH should be transferred.
  • Automatic backup process runs every 24h and exports all content in json-like format to backup-remark-YYYYMMDD.gz.
  • Authentication implemented with jwt stored in a cookie. It uses HttpOnly, secure cookies.
  • All heavy REST calls cached internally in LRU cache limited by CACHE_MAX_ITEMS and CACHE_MAX_SIZE.
  • User's activity throttled globally (up to 1000 simultaneous requests) and limited locally (per user, usually up to 10 req/sec)
  • Request timeout set to 60sec
  • Development mode (--dev-password set) allows to test remark42 without social login and with admin privileges. Adds basic-auth for username: dev, password: ${DEV_PASSWD}. should not be used in production deployment
  • User can vote for the comment multiple times but only to change the vote. Double-voting not allowed.
  • User can edit comments in 5 mins (configurable) window after creation.
  • User ID hashed and prefixed by oauth provider name to avoid collisions and potential abuse.
  • All avatars resized and cached locally to prevent rate limiters from oauth providers.
  • Images can be proxied (IMG_PROXY=true) to prevent mixed http/https.
  • Docker build uses publicly available base images.