hoardr's Introduction

Hoardr - Save your stuff from the web!

Introduction:

Hoardr is a self hosted open source (basic) alternative to pocket

Live preview:

Hoardr.in

Built with:

Testing:

Dependencies:

Social logins supported:

Install dependencies (some might need to use `sudo` for various reasons):

brew/apt-get install nodejs
brew/apt-get install redis
brew/apt-get install mongodb
npm install -g mocha
npm install -g gulp
npm install

Create an application on both Facebook, Google+ and point the callbacks to (respectively):

http://localhost:3000/auth/facebook/callback
http://localhost:3000/auth/google/callback

Email verification via Mandrill:

Create an application on mandrill

Environment variables:

PORT - Port exposed by this component.
example: 3000
SERVICE_NAME - The name of your fantastic service/platform!
example: Express Passport Application
DEBUG - Debug output (* for all) (optional)
example: *
NODE_ENV - Environment ('development', 'staging', 'production')
example: development
DOMAIN - The domain the service is running on
example: http://localhost:3000
SECURE_DOMAIN - HTTPS endpoint of the domain (only for production)
example: https://yourdomain.com
CLIENT_API_PATH - Path to the client REST api (relative)
example: /api
MONGO_URL - MongoDB url (including authentication)
example: mongodb://user:pass@localhost:27017/mydatabase
REDIS_URL - Redis url (including authentication)
example: redis://user:pass@localhost:6379
REDIS_DB - Redis database number (0-15)
example: 1 - defaults to 0
REDIS_SESSION_PREFIX - Prefix for redis session entries (optional)
example: sess:
APPSECRET - Application session secret
example: sOmeCrAzYhAsH894372
SESSION_KEY - Application session secret (optional)
example: express.sid (defaults to connect.sid)
FACEBOOK_CLIENT_ID - Facebook application client id
example: abcdefghijklmnopqrstuvxyz1234567890
FACEBOOK_CLIENT_SECRET - Facebook application client secret
example: abcdefghijklmnopqrstuvxyz1234567890
FACEBOOK_CALLBACK_URL - Facebook application callback url
example: http://yourdomain.com/auth/facebook/callback
GOOGLE_CLIENT_ID - Google application client id
example: abcdefghijklmnopqrstuvxyz1234567890
GOOGLE_CLIENT_SECRET - Google application client secret
example: abcdefghijklmnopqrstuvxyz1234567890
GOOGLE_CALLBACK_URL - Google application callback url
example: http://yourdomain.com/auth/google/callback
MANDRILL_API_KEY - Mandrill API key
example: abcdefghijklmnopqrstuvxyz1234567890
MANDRILL_SENDER - Mandrill email sender address
example: John Doe <[email protected]>
EMAIL_VERIFICATION_ROUTE - Email verification route
example: http://yourdomain.com/auth/local/verify

Run tests:

npm test

Run the application:

set environment variables
gulp
alternatively create a shellscript for the above (for development)
navigate your browser to http://localhost:PORT

Development shellscript example:

#!/bin/sh

export PORT=3000 \
# override port if argument is present
if [ $1 ]; then
   export PORT=$1
fi

export SERVICE_NAME="Hoardr Web Service" \
export DEBUG="*,-engine:polling" \
export NODE_ENV="development" \
export DOMAIN="http://localhost:3000" \
export SECURE_DOMAIN="https://localhost:3000" \
export CLIENT_API_PATH="/api" \
export MONGO_URL="mongodb://localhost/hoardr" \
export REDIS_URL="redis://localhost:6379" \
export REDIS_DB=0 \
export REDIS_SESSION_PREFIX="sess:" \
export APPSECRET="somecrazyhash" \
export SESSION_KEY="express.sid" \
export FACEBOOK_CLIENT_ID="000000000000000" \
export FACEBOOK_CLIENT_SECRET="000000000000000000000000000000000000000000000" \
export FACEBOOK_CALLBACK_URL="http://localhost:3000/auth/facebook/callback" \
export GOOGLE_CLIENT_ID="000000000000000000000000000000.apps.googleusercontent.com" \
export GOOGLE_CLIENT_SECRET="000000000000000000000000000000" \
export GOOGLE_CALLBACK_URL="http://localhost:3000/auth/google/callback" \
export MANDRILL_API_KEY="000000000000000000000000000000" \
export MANDRILL_SENDER=" <[email protected]>" \
export EMAIL_VERIFICATION_ROUTE="http://localhost:3000/auth/local/verify" \

gulp test && gulp

TODO

UI design needs work
tagging of articles
better content aggregation (fork node-read and adjust for specific needs)
lots more..

hoardr's People

Contributors

Stargazers

Watchers

hoardr's Issues

pagination / infinite scroll

Use: http://mongoosejs.com/docs/api.html#query_Query-slice
and the create pagination for infinite scroll
(this will return only a subset of the array each time)

Logo

make a nice logo :-)

add error message for when pages cannot be saved

for various reasons pages cannot be saved (https policies, etc), and the user should get a message about it.

use cookie to store user id and access token

store user id and access token in cookie, and use that to decide if a user is logged in

Uncaught HierarchyRequestError: Failed to execute 'appendChild' on 'Node': This node type does not support this method.

For this url (and others): http://www.nytimes.com/2014/07/16/business/a-push-to-give-steadier-shifts-to-part-timers.html?hp&action=click&pgtype=Homepage&version=LargeMediaHeadlineSum&module=photo-spot-region&region=photo-spot&WT.nav=photo-spot&_r=0

add source to meta-data for articles

preserve <pre> tag formatting

During the process, the 'pre' tag formatting is discarded and flattened. This should not be the case (makes code examples hard to read).

fetch a remote script from the bookmarklet for execution to enable versioning and seemless updating

Bookmarklet:

(function (e, a) {
    c = a.createElement('script');
    c.type = 'application/javascript';
    c.src = 'http://{{domain}}/bookmarklet.js/{{user.id}}/{{user.accessToken}}';
    var nodes = a.documentElement.childNodes;
    var wantedNode;
    for (var i = 0; i < nodes.length; i++) {
        if (nodes[i].nodeType == 1) {
            wantedNode = nodes[i];
            break;
        }
    }
    wantedNode.appendChild(c);
})(window, document);

Remote script:

(function ($, userId, accessToken, domain) {
    $.ajax({
        type: 'POST',
        url: 'http://' + domain + '/api/article',
        data: {
            id: userId,
            accessToken: accessToken,
            url: document.URL
        },
        success: function (response) {
            console.log(response);
            alert('Hoardr.in: page successfully saved!');
        },
        error: function (xhr, text, err) {
            console.log(err);
            alert('could not save page - error: ' + err);
        }
    });
})($, userId, accessToken, domain);

create chrome extension

This way jQuery or other libraries can be bundled without issues with cross-site request for loading a script

create bookmarklet guide for android

handle erronous article urls

show 404 for articles requested that does not exist, or has a bogus url

if an article has no body when saved, refer the the original page only

default all url's to http

To avoid issues with sites with https (f.ex youtube), all urls should be defaulted to http, as the request will be redirected automatically to https if necessary

refactor article model and saving to create unique entries for each user (no more global articles)

rewrite the article model:

{
    url: {
        type: String,
        unique: true,
        required: true
    }
    owner: {
        type: mongoose.Schematypes.ObjectId,
        ref: 'User'
    }
}

Then make a compound index of owner and `url``

Then change the saving/upserting to:

Article.upsert({ owner: user.id, url: body.url }, .... );

article route does not work for some articles

original:
http://www.iflscience.com/brain/brains-genes-and-chemical-imbalances---how-explanations-mental-illness-affect-stigma

url-encoded:
http://www.hoardr.in/article/http%3A%2F%2Fwww.iflscience.com%2Fbrain%2Fbrains-genes-and-chemical-imbalances-%25E2%2580%2593-how-explanations-mental-illness-affect-stigma

url-decoded (should be the same as original, but is not)
http://www.iflscience.com/brain/brains-genes-and-chemical-imbalances-%E2%80%93-how-explanations-mental-illness-affect-stigma

<-- find out why this does not work

eiriklv / hoardr Goto Github PK