This repository contains source code of Calendar module for OpenPaaS ESN.
The current module is activated by default in the OpenPaaS ESN repository. Doing a npm install from there will install and enable the calendar module in OpenPaaS.
1. Clone linagora.esn.calendar
git clone git+https://ci.linagora.com/linagora/lgs/openpaas/linagora.esn.calendar.git
cd linagora.esn.calendar
2. Install dependencies and link in OpenPaaS
Note: The following instructions assumes that you have already installed OpenPaaS ESN in the path referenced by $ESN below.
In order to develop, you will have to launch several commands from your favorite terminal:
- In the current repository, install dependencies with
npm install - In the current repository, create a npm link by using
npm link - In your OpenPaaS ESN repository, link the calendar module
cd $ESN
npm link linagora.esn.calendar
3. Run tests
In order to run tests, you have to create a symbolic link for the linagora.esn.resource module:
ln -s $PWD/node_modules/linagora.esn.resource node_modules/linagora-rse/node_modules/
Make sure you have gitlab-ci-multi-runner installed locally, following the installation instructions. Make sure you have Docker installed.
Run tests:
$ gitlab-ci-multi-runner exec docker test
This will pull all required images, and run the whole tests suite in a Docker container. If you only want to run linters and unit-frontend tests, you can run them on your machine directly using:
$ grunt linters test-frontend
Please have a look on the issue document.
I can only search events from my default calendar "Events", the events from other calendars are not reachable!
This may be due the fact that Elasticsearch is not will indexed. To verify please do the following in your terminal (or maybe in Sense)
GET /events.idx/_mapping
and search for the calendarId in the events mapping type. You may have the following output:
"calendarId": { "type": "string" }
In such case, the calendarId is analyzed string. For more information, have a look here. Normally, this attribute should be not_analyzed. To solve it, you should reindex Elasticsearch.
To do so you may delete the events.idx and start again from scratch as explained in the Elasticsearch Index section in the OpenPaaS documentation. Or you can simply update the events.idx as explained in the next section
Once the elasticsearch is reindexed, the calendarId in the events mapping type should have the following output:
"calendarId": { "type": "string", "index": "not_analyzed" }
I want to reindex events.idx with out restarting from scratch
this can be easily done using Scripting provided by elasticsearch.
-
Before starting you should enable scripting in your elasticsearch.yml
locate elasticsearch.yml // to locate the elasticsearch configuration file script.inline: true // this line is to be added to the elasticsearch.yml -
Once done, you should restart elasticsearch
service elasticsearch restart -
Copy the content of
events.idxinto a temporary index (events-tmp.idxfor example). For here I will use Sense. For sure you can the same using curl ;-). In the following command, we will try to update the_idfield so it has the following format:userId--calendarIdPOST /_reindex { "source": { "index": "events.idx" }, "dest": { "index": "events-tmp.idx" }, "script": { "inline": "ctx._id= ctx._source['userId'] + '--' + ctx._source['calendarId']" } } -
Please verify that
events-tmp.idxis created before going to the next step. Both commands must give the same outputGET /events-tmp.idx/_search GET /events.idx/_search -
Now we have a backup of our
events.idx, we can delete it:DELETE /events.idx -
Let us create a new
events.idxwith the good configuration. Either you do it using esn-elasticsearch-configuration, or do it manually. It is worth noting that the configuration we are going to use is the same used by esn-elasticsearch-configuration as can be seen herePOST /events.idx { "settings": { "analysis": { "filter": { "nGram_filter": { "type": "nGram", "min_gram": 1, "max_gram": 36, "token_chars": [ "letter", "digit", "punctuation", "symbol" ] }, "truncate_filter": { "type": "truncate", "length": 36 } }, "analyzer": { "nGram_analyzer": { "type": "custom", "tokenizer": "whitespace", "filter": [ "truncate_filter", "lowercase", "asciifolding", "nGram_filter" ] }, "whitespace_analyzer": { "type": "custom", "tokenizer": "whitespace", "filter": [ "truncate_filter", "lowercase", "asciifolding" ] } } } }, "mappings": { "events": { "_all": { "analyzer": "nGram_analyzer", "search_analyzer": "whitespace_analyzer" }, "properties": { "uid": { "type": "string", "index" : "not_analyzed" }, "calendarId": { "type": "string", "index" : "not_analyzed" }, "userId": { "type": "string", "index" : "not_analyzed" }, "sequence": { "type": "string", "analyzer": "nGram_analyzer", "search_analyzer": "whitespace_analyzer" }, "summary" : { "type": "string", "analyzer": "nGram_analyzer", "search_analyzer": "whitespace_analyzer" }, "description" : { "type": "string", "analyzer": "nGram_analyzer", "search_analyzer": "whitespace_analyzer" }, "allDay": { "type": "boolean" }, "durationInDays": { "type": "short" }, "start" : { "type": "date" }, "end" : { "type": "date" }, "dtstamp": { "type": "date" }, "organizer" : { "properties": { "email": { "type": "string", "index": "not_analyzed" }, "cn": { "type": "string", "analyzer": "nGram_analyzer", "search_analyzer": "whitespace_analyzer" } } }, "attendees" : { "properties": { "email": { "type": "string", "analyzer": "nGram_analyzer", "search_analyzer": "whitespace_analyzer" }, "cn": { "type": "string", "analyzer": "nGram_analyzer", "search_analyzer": "whitespace_analyzer" } } } } } } } -
Now we have the
events.idxcreated, you can verify this doing the following command. This should display theevents.idxwith the configuration we have passed in the previous stepGET /events.idx/_mapping -
Now we need to copy back the data from our backup
events-tmp.idxPOST /_reindex { "source": { "index": "events-tmp.idx" }, "dest": { "index": "events.idx" } } -
That it. Now you can verify that you have all the events indexed back in
events.idxGET /events.idx/_search -
This step is optional. You can keep the
events-tmp.idxas a backup if you want, or simply delete it if you are sure that you do not need it any more:DELETE /events-tmp.idx
I want to reproduce locally the UseCases tests that I perform on the production platform
The thing to know is that all the ics sent from Exchange/Gmail pass in JAMES. Then JAMES sends them to OP through RabbitMQ using the james:events exchange (please do not confuse between exchange of RabbitMQ and Exchange ;-) )
Before starting, ask an OP member for the RabbitMQ URL, name and password. (I cannot write it here, you know it is on Github... Yolo!!)
To get the message from the james:events exchange, we can create a DURABLE queue (let us call it SuperQueue) and then bind it to the james:events exchange. Once you do so you will get benefit from the fact that the james:events exchange has the type fanout. That is, it will forward all its message to all bound queue. So the SuperQueue will receive a copy of all messages the original queue receives.
Now go to your SuperQueue and press on Get messages and so you can get all messages from it. Do not forget to purge this queue manually cause it is durable. So you do not have to store useless messages.
Once you get the message from your SuperQueue, you can go to your local RabbitMQ. Then go to the james:events exchange and publish the messages you got from the production here. The corresponding events are then displayed in OP.
If you want to restart the debugging with the same ics, you can publish the same ics again. By doing so the status of the corresponding events go back to NEED_ACTION. So you can restart your debugging.