OpenMRS Planet

This week I installed the module and do the testing. The swagger resources generator required several changes to compatible with the latest release.

After completing the swagger generator changes, I tested available resources and fixed bugs. I’m at final stages of the testing and need to move quickly to the next phase.

Continue reading "GSoC 2017 – Week 8"

This week involved studying AngularJs, which will be needed to create the UI Fragment, for the clinician facing Patient Dashboard. I learnt everything that will be required which includes:- Modules, Directives, Controllers, Databinding, Service, Filter, Http, etc.

Apart from this, I also begin to setup the module along with a reference application distribution, so that i can view the ui changes.
This required me to do the following :-

Using the SDK
First create a ref app distribution on a server using

mvn openmrs-sdk:setup

then put the module on watch

mvn openmrs-sdk:watch -DserverId=myserver

finally cd into the created module
then run

mvn clean install mvn openmrs-sdk:run -DserverId=myserver

also put port:xxxx for remote debugging, in IDE

However I seemed to face some kind of liquibase error which I trying to debug.
I also added the required dependencies, in the module along with a JSON file which will help in creating an extension to the secondColumnFragment in the clinician facing patient dashboard.

Hi All,

This week I was quite busy with my college work so I couldn’t do much GSoC work. This week I just took very light tasks and discussed future tasks with mentor.

My this week tasks are:

• DINT-74: Improve the workflow for running single rule
• DINT-82: Create and expose API of data integrity module

DINT-74

Right now the flow of running single rule is not very intuitive and informative. In this task, I need to think of a way which is more engaged and intuitive.

DINT-84

This one is quite interesting and challenging task. Although I did have some background in API and relative stuff but I don’t have much experience with it in context of OpenMRS systems.

That’s all for this week.

Thanks!

Well the previous week was pretty hectic and kind of disrupted my time table a bit too. This was because of the sudden need of a new fix/task due to which I had to pause my existing work and work on.

College too has been very hectic over the past week.

Hopefully , I am able to complete the work ASAP.

Another thing to note is that the second evaluation is round the corner so I have to fill it in. Looking forward to the coming week with Gsoc @ OpenMRS!

This week was the most successful week. I was able to complete all the core functionalities of my project. Now the DHISReporting module has the capability to map the exported reports with the report module reports. Not only for sql query based report, but also for other Cohort reports.

Now any data element can be mapped with any existing indicator.

Mapping

Currently the elements mapped with other reporting indicators are displayed as an empty section. Now working on adding mapped report label to that section.

Mapped queries

I implemented the functionality to execute those reports when generating the report.

Generated report

Those are the achievements for this week. Next week I have to work on some UI changes which suggested by my mentor and continue writing test cases to the tickets I claimed.

That’s all for this week… :)

Background:

As I  said before, my project is to develop a web application where end users can browse and download OpenMRS packages hosted on Bintray using the Bintray REST API.So I had to make AJAX (XMLHttpRequest) requests in order to communicate with the Bintray REST API.

Problem:

AJAX (XMLHttpRequest) requests have traditionally been limited to accessing the same domain as the parent web page (as per the same-origin security policy). “Cross-domain” AJAX requests are forbidden by default because of their ability to perform advanced requests (POST, PUT, DELETE and other types of HTTP requests, along with specifying custom HTTP headers) that introduce many cross-site scripting security issues.

I have communicated with Bintray support only to know that CORS standard is not supported by Bintray yet and that they will be adding this feature by month end.

Solution:

The response from the remote server to which we are making AJAX calls should have headers like this

1. Access-Control-Allow-Headers:Origin, X-Requested-With, Content-Type, Accept
2. Access-Control-Allow-Origin:*

The solution is to develop a proxy server which will add the required headers to the responses received from Bintray.

So I have developed a proxy server using the Expressjs framework which exposes REST endpoints from where I called  api.bintray.com  and added the required header’s to the response received from Bintray.

Here’s the final setup,My web app will be running on a port(say localhost:8000) and my proxy server on port(say localhost:3000). I will call rest endpoint exposed by proxy server  (localhost:3000/packages )  from my script in webapp (localhost:8000). The endpoint at (localhost:3000/packages )  contains code to communicate with Bintray, When the proxy server receives response from Bintray it will add required headers ( have a look at http://enable-cors.org/server_expressjs.html) and send it to browser.

I would like to thank Saptarshi and Rafal for guiding me through this blocker.

For more info on CORS also read:

http://blog.javascripting.com/2015/01/17/dont-hassle-with-cors/

enable-cors.org

https://developer.mozilla.org/en-US/docs/Web/HTTP/Access_control_CORS

Week 8 [July 17, 2017 — July 24, 2017]

I spent this week with SysAdmin OWA. Because there are some problems encountered according to the module updates. When User trying to check the module updates, It shows some unexpected error,

My mentor asked me to investigate this problem and find some solutions to avoid this problem.

Here I have used modules.openmrs.org to retrieve the required module information. There is a REST endpoint which can be used to get the module details from the online server. This REST endpoint will give us these following details,

1. Module download URL
2. Module Name
3. Current/Latest version of the Module
4. Module Owner/Creator
5. Module description

The final REST response will be like this,

REST End point already developed with some different kind of metadata. They have used UID to indicated the module. But for some modules, those defined UID in the REST endpoint was different from actual UID in the system. So when I am trying to get details using actual UID, It’s response to the user as Error. This is the reason for this problem

OpenMRS | Add Ons

OpenMRS Community is going to retire the modules.openmrs.org platform. They are developing OpenMRS Add Ons Platform which can be used instead of modules.openmrs.org.

Adding user’s module to Add Ons for indexing is as easy as creating a pull request! Before they do that, they need to ensure that their module has been hosted on one of the sites that Add Ons currently supports indexing from:
1. Bintray
2. modules.openmrs.org

Once they have confirmed the above prerequisite, they can head over to the Publishing an Add on document on Github which contains a comprehensive list of steps.

Using OpenMRS Add Ons for Updates Purpose

After some discussions about the current UID problem with my mentor Daniel, There is another precious Idea mentioned by Darius in my talk thread. That is using OpenMRS Add Ons platform instead of module.openmrs.org platform.

OpenMRS Add Ons platform already contains some REST Endpoints which are based on module.openmrs.org. It can be used to get information about the modules using moduleID which is similar to the actual module’s package name.

https://addons.openmrs.org/api/v1/{moduleID}

But for some modules, I can’t get details for some modules using their package name

I can get details using package name(or UID) for some modules,

1. Atlas Module (package : org.openmrs.module.atlas ) — URL : https://addons.openmrs.org/api/v1/addon/org.openmrs.module.atlas
2. UI Commons Module (package : org.openmrs.module.uicommons ) — URL :https://addons.openmrs.org/api/v1/addon/org.openmrs.module.uicommons

But I can’t get modules details for this module using package name or UID

1. Legacy Module(package name : org.openmrs.module.legacyui ) :
   Original URL : https://addons.openmrs.org/api/v1/addon/org.openmrs.module.legacy-ui-module1
2. Data Exchange Module (package name : org.openmrs.module.dataexchange)
   Original URL : https://addons.openmrs.org/api/v1/addon/org.openmrs.module.data-exchange-module

Here the problem is, They have predefined those UID which is similar to the module’s package name but It differed to some modules. Predefined module data JSON file is given below,

openmrs/openmrs-contrib-addonindex

So after some discussions with Darius, He promised to provide another REST endpoint which can be used with actual module package names. The REST endpoint will be like this following,

https://addons.openmrs.org/api/v1/addon?type=OMOD&modulePackage={x}

So I am waiting to that REST endpoint to complete the Updates part in SysAdmin OWA.

Follow this talk thread to get more information and discussions about the module update using OpenMRS Add Ons.

GSoC 2017 - More Metadata Management in AdminUI

Time flies. This week is the last week of the second coding period, and I have done lots of things to complete the second project objective. As mentioned last week blog, This week I focus on the coding work, and I do works likes said following:

• I write a detailed proposal about the second objective and my technologies. As I am not familiar with the project, I did not write this in last evaluation, mentor Jon said that he wishes that I give him a detailed proposal in evaluation. So I do this, and complete some mockups.
• The most important is coding, I use AngularJS, HTML and CSS to complete the html view, and to interact with web back-end. I think it is nice to me because this is my first time to use AngularJS in a project.
• The third work is that I completed the functions like concepts and mappings differences viewer. In last month, Jon had said to me that he is looking forward this objective. It is wonderful to us to see the differences of concept and mapping graphically.

There are also many works in the second objective: diff viewer, like concepts and mappings selection, compare details. And I won’t put them here, as detail are listed in the proposal. The blog records my work and my sweat, it is enough.

In the end, Jon gives me a lot help, and guides. And our community is helpful to us to get solutions about some problems. Thanks to Jon, and all of people who are enthusiastic。

This week I completed the app demo. In my previous PR my mentor mentioned to use buttons for switching to next showcase view but the previous library I used[1] there was no option to do that so I switched to a new library[2]. It was way more customizable then the previous one. So I shifted to it and everything was working fine but to test it properly I had to clear the app data and then I was not able to login to app again for further testing.I was always getting the error of invalid credentials but strange thing is everything was working fine on the website.I tried three different devices on two different networks, I also tried with the app published on play store but no luck. So, I made a new thread on the talk forum[3] and the issue is still not resolved.

I also completed the forms creating on the server if they are not present from within the app only but I had not pushed the code to the GitHub because due to login problem I was not able to test it properly.

I also read about the provider module in the meantime so that it will further escalate progress afterwards.

References :

[1]https://github.com/deano2390/MaterialShowcaseView

[2]https://github.com/amlcurran/ShowcaseView

[3]https://talk.openmrs.org/t/login-error-on-qa-refapp/12531

This week I was primarily focus on multiple tasks. Basically my intention was to implement the navigation among the different report pages inside the OWA. I was able to complete in using the latest React router v4 dom library. So, Now the pages are available and users can smoothly navigate through them.

I have created a new thread to get feedback on how the reporting UIs is going to display.

Feedback on built-in reports for reference app wireframes

There was a minor style issue and fixed it. While working on these, I found out that 3 reports is not rendering. I observed that this is because those reports were accepting the location as a parameter. Even though I pass the location as just string parameter inside a json, it is not recognized. As my mentor advised me not to spend more than 3 hours on a problem, I raised this problem which I was facing.

How to pass Location object from OWA to reporting REST API

At the same time I found another issue where the dates from the Reporting REST API is not returned as formatted properly. So, I created another issue in JIRA. My mentor replied me to work on this issue first and resolve the date formatting issue first. So, Currently I’m working on this.

• PTM-82 - Load patients for the incremental matching
• PTM-83 - Generate and save reports in incremental patient matching process
• PTM-84 - Perform patient match with two datasources

Continue reading "GSoC 2017 – Week 7"

This week I did some changes to the code I committed regarding DRM-09 and DRM-28 tickets as Daniel suggested to do. After having a discussion with my mentor, I started to work on second phase of the project which is to map other report types other than SQLCohortDefinitions.

Currently I have implemented a method to execute and display the report results. Next thing I am going to work on is to display the executed report results in DHISReportingModule report.

I have to work on writing test cases for the previous tickets as well. That’s all for this week… :)

This week involved me continuing my work on the REST API. After another call with my mentor we were able to organize our efforts and finally come up with How the resource will work. This week also involved me cleaning up the restApi Code So that it could be finally merged. It is finally at the end and the code will be merged my today or tomorrow. With that I can start my work on the UI Fragments. I have already recognized the dependencies that will be needed.
The REST Resource will support searching by :-
1) objectType and objectUuid
2)objectType and tagName

This week I dedicated to test the module. After the version upgrade of the HAPI FHIR library, I encountered several issues during module loading. My mentor asked me to do comprehensive testing for the module. So I’m checking all the available resource and their operations.

This week was all about writing tests for my previously commited work. I also had to make some changes in the UI workflow as pointed out by my mentor Darius. I was initially really stuck at some places because I really couldn’t figure everything out from the documentation available online.

I am also a newbie when it comes to testing and hence had to take some time off to understand the meaning and usage of the different testing tools out there for java. My mentor asked me to stick to Spring based testing as far as possible and that was a greater challenge for me as it make finding relevant examples all the more tougher. None the less, although it took me some time, I can proudly say at the end of this week that I have been successful in creating tests for my use case. Technically speaking, I should’ve followed the TDD approach as pointed out by my mentor and will keep that in mind for future changes.

I also improved the UI . The select option is now disabled if there are no module versions which support the User given OpenMRS Core version.

All in all, this week can be summarised as productive but not really the best when it comes to efficiency( can be attributed to the fact that my college has reopened and hence it took me time to settle in). I hope to speed up the work the coming week. On to week 8 with Gsoc @ OpenMRS!

Week 7 [July 10, 2017 — July 17, 2017]

This week also I wanted to work with SysAdmin OWA module. Because my mentor Daniel wanted to make this better for Production level. At the beginning of the week, I received some feed backs from Darius and my Mentor Danieal.

Working with my mentor Daneial

He was very active to guide me about the module. The major problem encountered by him was wrong Alert Messages contents. I displayed an alert message when a user triggers Delete/Stop module buttons. These alerts messages contain some module names.

There are two type of module lists inside the Module Details JSON object, That are,

• Required Modules — A module can only access the API/service of another module if it is marked as required.
• Aware of Module — This means that if the other module is also installed, do some extra features in Atlas module

Initially, I just displayed those Required Modules list as the affected modules. But he encountered that was wrong. I can’t figure out that concept immediately, but after some discussions, I can understand the concept of dependent modules. Pseudo code for the concept is,

Get module package name A
Get all modules data
Loop all modules data {
   Get required modules data B
   check weather package name A is in the list B {
       Add modules name to display
   }
 }
}

Then I have implemented Dependent Module list inside the Module View details page as follows.

Working with Darius

Darius mentioned some GUI enhancements which will lead to SysAdmin better. He observed deeply and suggested some feed backs to improve my OWA. Some of the main objective from his view are,

• Remove the red color for Delete/Stop button. Because It may avoid the eye contact from other parts of the module.
• Be consistent about whether we’re going to have Icon and Text for Start, Stop, and Delete.
• All actions buttons should be inside the one column.

So I completed those changes and asked him again for the reviews.

New Features

One of the new feature is Dependent Module data and another one is Getting module information from Online.

In the module view details, there are a lot of modules will be displayed under Required Modules and Aware of Modules. Some of those modules might not be installed on the server at that time. So when user click those modules, OWA wanted to face some Exceptions. So to handle this problem, I have implemented Get Details from the Online option.

It will gather details from Online OpenMRS module’s repository and will show it to the user as follows,

Happiness is getting some greetings from my mentor like this :-p ,

Hello everyone!

This week I worked on DINT-80 which was relatively easy issue, I just needed to remove hardcoded strings and replace them with variable which fetches localised strings. Just after merging respective PR, my mentor released new version of Data Integrity Module . There are some major updated in Data Integrity Module — 4.4.1 which are:

• Now users can filter the data integrity results by patient name, program name , by rule name and even by free text search.
• Filtered/Unfiltered result can be exported to CSV, Excel and PDF for offline and other usage.
• Some small fixes like improving styling of data quality dashboard and localising the hard-coded strings.

This week I’ll working on DINT-75 which is again a new stuff for me but I hope with the help of OpenMRS community I’ll able to complete it.

Links to all relevant pull requests:

• DINT-80
• DINT-76
• DINT-77
• DINT-73

Thanks!

Useful Links

• Data Integrity Module-4.4.1
• Discussion thread

To help the new users know and utilise all the functionalities of the android app a proper app demo is necessary. So, this week I used MaterialShowcaseView library[1] and added the app demo. Below are the few screenshots of the same.

Also I have almost fixed the forms bug and will send the PR in some time after properly reviewing the code.

It is useful to record every week’s progress for me. And this is the newest blog to record what I did this week and what I will do next week. And I realize that writing blog with English is also good to be familiar with English for me, because my English is not good.

At first, I must say a good new about my family, it is that a new baby was born this week in my family, I can not express my happiness. So I went back home to visit the wonderful new mom, my old sister. I want to share the good new to everyone.

OK, let’s be back my work. I have to say the new baby consume some my work time, LOL. I have learned the work listed in last week blog, and I won’t write them again. Following are my some new work:

1. I searched some implement idea from Google and Baidu, and I have some ideas about the second task.
2. I am working on the second task’s mockup and proposal, because mentor Jon said that he hoped to get a specific proposal when submit one task.
3. Some work about coding, I have implement some basic function about the task, like get the concepts which will be compared.

Next week, I will work more, because there are many work to do. I will do

1. complete the project mockup
2. complete the project proposal
3. coding to implement the function

OK. That’s all, I have to say, I must to work much time to complete my second task, LOL, but I enjoy it. Work is better to waste my life. Thank you.

I was unable to work as much as I desired this week due to illness and all the mess I had to sort out after moving into my new hostel.

This week, I fixed the grant types issue in the UI frame. Now, we have a working client registration using UI and the next problem is the Authorization errors. For example, consider the following GET token request :

http://localhost:8080/openmrs/ws/oauth/token?grant_type=password&client_id=app10&username=admin&password=Admin123

and adding the client credentials in the header in an encoded form, I get a stack trace as JSON. The notable line in this stack trace is

      {

The post Week 7 Grant types issue fixed, authorization problem appeared first on Sanatt Abrol.

Last week I was trying to fetch data and show them in OWA which I have created now. We had the weekly meeting with my mentor, and showed him about the project structure. He modified the package.json file which needed to add some dependencies.

I was able to create the components for each of the reports which I have created in referenceMetaData module. I was glad that I could complete it. I updated the OpenMRS talk thread, and I saw that it has been updated since my last comment was 15 days back. I think I need more attention to my project work right now.

I have created a very basic structure for those components as shown like here. My mentor told that we don’t need to focus much on styling yet, but to fetch the data from server module.

I think, I was able to complete what we have targeted for this week. Now I am looking to implement the routing feature from navigation bar to display the relevant components when it has been clicked. Tomorrow we have our weekly meeting for this week and I hope to get feedback on this from my mentors.

6 weeks into Gsoc and I still learn new things each week! So this week was all about modifying my previous commits as per my mentor’s feedback on my Pull Requests. This week , I also learned as to how to place code where and as to how to write unit tests for Spring applications(I did know how to write them even before but I needed to know the difference between a unit test and an Integration test) . I also had a call with my mentor which was indeed very productive.

Next week I will be focusing on using the Bintray rest API to fetch details like download counts etc . Since we have finally migrated all OpenMRS modules to Bintray(Thanks to Rafal :)), we will sl0wly migrate from OpenMRS modulus to Bintray. Hopefully this process will be smooth.

I was also a bit busy as I had to pack my bags and move back to college for the next semester so basically that marked an end of my summer vacations :(

I look forward to the next week as we have a lot in store. So , on to week 7 with Gsoc@OpenMRS !

GSoC’17@OpenMRS — Week Six

Hi everyone!

This week I worked on DINT-76 which again involves work related to datatables. More specifically, I need to add capability to export the filtered list in different formats such as CSV, Excel and PDF.

I started looking for different options then my mentor pointed me to the export feature provided by button library of datatables.

Data exported in PDF format

With the help of datatables, it was just matter of minutes to add export capability but real work involved improving the look of the export buttons. By default they look very ugly and non-intuitive.

Export option given in Excel, CSV and PDF

Final look of the export buttons — attractive and intuitive.

OpenMRS doesn’t have icons for excel, csv and pdf inbuilt so I need to add them in the data integrity module. I dig little bit and asked around and came across sass and compass. Again, during GSoC I came a technology I wasn’t familiar with.

I used the fontello.com for generating required icons files such as svg, ttf and woff, etc. I also took me a while to learn about sass and compass. After that I just need to tweak default config to setup it according to our need — like adding filename, title and tooltip. Today, it got merged and will go in the productions. Next I will start working on DINT-80.

Useful Links

• Fontello.com
• DINT-76
• Import icons in OpenMRS
• DINT-80

In the Design call held on the 5th of July, a discussion was held on the way forward regarding the Generic-Tagging-Mechanism Project. It was decided that instead of working on Angular directives, the first step would be to add this mechanism to the clinician facing patient dashboard. The plan is to create a fragment in the tagging module and to provide an extension point to the secondColumnFragment in the clinician facing patient dashboard :-

Apart from that several changes were decided for the Rest Resource. I added searching capability by 3 parameters objectType, objectUuid, and tagName. Depending on which parameters are provided a List of tags will be returned. I also started cleaning up the Rest Api code according to the Code Review. The representation was also modified.

For full representation
GET .../v1/tag?tag=my-favorites?v=full

[

{
“uuid”: “uuid-of-tag-object”,
“tag”: “my-favorites”,
“objectType”: “patient”,
“objectUuid”: “…/patient/uuid-of-tagged-patient”,
“auditInfo”: “”
},
{
“uuid”: “uuid-of-tag-object”,
“tag”: “my-favorites-of-all”,
“objectType”: “concept”,
“objectUuid”: “…/patient/uuid-of-tagged-concept”,
“auditInfo”: “”
}
]

For Default Representation
GET .../v1/tag/my-favorites


{
“tag”: “my-favorites”,
“links”: [
{“rel”: “patient”, “uri”: “…/patient/uuid-of-tagged-patient”},
{“rel”: “concept”, “uri”: “…/patient/uuid-of-tagged-concept”}
//….
]
}

I still have a few doubts regarding the default representation, but other that that the Rest part is almost complete. I have also started looking into how I will add the UI fragment. For this I am using Allergies UI and the tagging mechanism of JIRA as reference.

Continue reading "GSoC 2017 – Week 6"

This week was very hectic ! I travelled a lot this week in fact right now also I am writing this blog post from train. I was at my home and due to some family issues I was not able to work properly but still I managed to get some of the work done.

I read the documentation on the provider module and the appointment module from the Openmrs wiki which will help in escalating the speed of my work when I will implement them.

Currently the vitals forms has to be added manually on the Openmrs server which is a bit tedious task and also the problem is if the JSON form is not present, then the app doesn’t display any form at all, so to resolve this issue it is decided that if the HTTP request gives a null value then the form should be generated on the client side i.e. in the app.

To implement this I started test on qa-refapp server through Postman but for some reasons I was getting a 403, I reported about this and Rafal told that he is not facing any problem with the server and everything is fine. So, I again tried with it but got the same error ! Rafal posted the request he was using and now I can debug the issue properly and finally get to know what I was doing wrong.

I will fix this and then implement the rendering of the forms on the app in a day or two and now I have to devote more time as it has put behind my schedule.

 @Override  
      public Patient savePatient(Patient patient) throws APIException {  
           //************work starts  
           //todo  
           //pass unknown att to core - is paasing along regCore as a separate para is ok?  
           if (unknown != null && unknown) {  
                PersonName name = new PersonName();  
                name.setFamilyName("UNKNOWN");  
                name.setGivenName("UNKNOWN");  
                patient.addName(name);  
                //patient.addAttribute(new PersonAttribute(emrApiProperties.getUnknownPatientPersonAttributeType(), "true"));  
                //work on addingAttribute - core can not depend on the relevant module. therefore rewrite that...  
                PersonAttributeType type = null;  
                type = this.personService.getPersonAttributeTypeByName("Unknown patient");  
                if(type == null) {  
                     throw new IllegalStateException("Configuration required: Unknown patient");  
                } else {  
                     return type;  
                }  
                //resolve errors  
           }  

As the last blog said that I need to cost 2 weeks to learn how to use AngularJS. I am glad that this I can be familiar with the AngularJS within two weeks. So it’s time to end the learning period, and start my second task.

This week I have learned some knowledges about the rest of AngularJS, like

• Service : A little confused
• HTML DOM
• Bootstrap
• event
• form

I believe that I can implement the second task with these basics I learned these two weeks. In fact, I have to say, javascript is powerful for web applications.

It is better to have a plan for the next week. I hope I can understand the whole requirement about the second task. And I have to an complete idea about it, implement a specific technical proposal. Maybe I have time to code few to implement a part of this task.

That’s all, I think all the technical detail will record in next week blog. Thank you for all the people who care about me.

Make SysAdmin better!

Week 6 [July 03, 2017 — July 10, 2017]

After the First evaluation, I have started again my works under my mentor Daniel Kayiwa. There are some tasks according to my timeline, but he advised me to complete the SysAdmin reviews and improvements before moving to the new stage of the development.

According to my mentor reviews, I got some import GUI bugs and latency problems such as,

1. Anyone can delete installed modules without any confirmation
2. Header fonts differ from Existing Open Web Apps
3. Breadcrumbs contains hyperlink for active pages also and There are two Home buttons displayed (One as Home image and other One as ‘Home’ text)
4. Initial Page loading latency and some ugly characters printed before data fetching
5. Show module names in the pages instead of module legacy-ID

So I investigated those mistakes and problems and find out the corrections. Then after I updated my modules with Daniel’s feedback. You can check the feedback and my updated information through this following OpenMRS Talk thread,

GSoC 2017 - More Metadata Management in AdminUI

Major Updates — SysAdmin

• Implemented user Confirmation before delete modules

Updated Confirmation Message

• Implemented user alerts when user try to STOP some modules

Now I have completed this following parts in SysOWA,

• Module Management -
  All Module Details — Start/Stop Module, Delete Module, Check Module updates, and Start All modules
  Selected Module View — All details of selected module, Start/Stop Module, Check for Module update
  Upload New Module
  Search and Install New modules
• System Information

I am working with Encounters and Observations along to this improvements. Completed parts in the OpenMRS Encounters,

1. Encounter Search/Add/Edit
2. Encounter Type Management
3. Encounter Roles Management

This week I planned to complete Encounters and Observations :-)

Best way to Manage your Logs using WSO2 Integration Cloud

This is my second week of Internship at WSo2 and I have implemented a simple dashboard which can retrieve and show the logs from AWS CloudWatchLog.

Dashboard to show Log Groups, Log Streams and Log Events with Filters

AWS CloudWatchLogs API

I have used AWS CloudWatchLogs API with Node.js and Angular.js to develop this dashboard. Amazon docs give better documentation to implement a Node.js application with AWS SDK. You can follow this documentation to get more information about AWS.CloudWatchLogs API.

Class: AWS.CloudWatchLogs - AWS SDK for JavaScript

Development Coverage

I have developed back-end with Node.js and It can retrieve the data from AWS CloudWatchlogs using REST APIs. There are several APIs which can play a major role in this development.

1. Cloudwatchlogs.describeLogGroups() — Used to retrieve Log groups from AWS CloudWatchLogs.
   Log groups — groups of log streams that share the same retention, monitoring, and access control settings. Each log stream has to belong to one log group (source).
   Limits : You will get only 50 log groups per a request with nextToken. So you can get next set of log groups using that nextToken.
2. cloudwatchlogs.describeLogStreams() — Used to retrieve Log Streams under given Log Group from AWS CloudWatchLogs.
   Log streams — A log stream is a sequence of log events that share the same source. More specifically, a log stream is generally intended to represent the sequence of events coming from the application instance or resource being monitored (source).
   Limits : You will get only 50 log streams per a request with nextToken. So you can get next set of log streams using that nextToken.
3. cloudwatchlogs.filterLogEvents() — Used to retrieve Log Events under given Log Group and set of Log Streams (max 100) from AWS CloudWatchLogs.
   Log Events — A log event is a record of some activity recorded by the application or resource being monitored. The log event record that CloudWatch Logs understands contains two properties: the timestamp of when the event occurred, and the raw event message (source).
   Limits : You will get only 10000 log events per a request with nextToken. So you can get next set of log events using that nextToken. You can give only 100 log streams per request as an array.

I have implemented all the REST calls for the necessary actions using Node.js(Express framework) and Stared the server.

Now I can get the Logs from AWS CloudWatchlogs to my client PC — As JSON Objects …!

JSON Object for the Log Events request

Now, I wanted to display those JSON Objects to the User, So I advised to implement using Jaggery, But for the demonstration part, I have started to build using Angular — because I already familiarized with Angular.

Dashboard Log Groups page

You can check my Repo to get more about the code :D

suthagar23/aws-cloudwatchlogs-dashboard

WSO2 Integration Cloud

One of the cloud service under WSO2 is Integration Cloud. WSo2 just implemented their Integration product in Cloud to make it easy for their customers.

You can just go to http://wso2.com/cloud/ to access WSO2 Cloud and you will be given 14 days trial period at the beginning. But unfortunately, If you have a trial license, then your deployed application will be turned off after 8 hours from that application’s deployed time.

You can just Sign up using your Email, No more requirements :-)

After the successful login/signup, you will be redirected to WSO2 Cloud Dashboard.

Cloud Dashboard of WSO2 Cloud

Just Click on Go under WSO2 Integration Cloud and then, you will be automatically redirected to WSO2 Integration Cloud dashboard. WSO2 provides great UI/UX for their customers with reduced complexity. You can see, there are only a few functionalities(buttons) are available for the customers!

Just Click on Create New Application. Then you can see there is some support applications list.

Currently, they are providing Integration Support for,

1. WSO2 ESP Composite Application — Allows you to deploy an ESB configuration that is supported by WSO2 Enterprise Service Bus
2. Ballerina — Allows you to deploy an integration scenario developed by Ballerina.

And they are providing Service support for,

1. Java Web Application — Allows you to create dynamic websites using Servlets and JSPs and deploy web services.
2. Java Micro Service — Allows you to create dynamic websites using Servlets and JSPs and deploy web services.
3. PHP Web Application —Allows you to create dynamic websites using Servlets and JSPs and deploy web services.
4. WSO2 Jaggery Application — Allows you to write all parts of web applications, services, and APIs in a completely JavaScript way.
5. WSO2 Data Service —Allows you to deploy a data service that is supported in WSO2 Data Services Server.
6. Node.js Application —A llows you to deploy a data service that is supported in WSO2 Data Services Server.

Deploy Amazon CloudWatchLogs Application into Integration Cloud

Just create Node.js Application to deploy this Amazon CloudWatchLogs Application.

One Minute! You want to include your Node.js route file/initial file as Index.js. Because when Integration cloud tries to start your application, It will try to start from Index.js.

When you moved to create New Node.js application, you will be required to Upload Zipped file or provide URL for the application.

Are You new to WSO2 Integration Cloud, then just click on Try out Samples ….

Then after, you will be required to provide relevant information such as Application Name, Version and Description for your deployment. Finally, WSO2 Integration cloud will create Docker images for your application and start to build your application in Cloud.

What is Docker? Just follow my blog post about the Dockers !

After the creating the application, You will be redirected to the application Dashboard which will give more information about your deployment. There will be an auto generated URL for your Node Application and you can just click on that to start your Application.

Cool… Now Amazon CloudWatchLogs Application is running on WSO2 Integration Cloud …!

Referred Source -

1. http://docs.aws.amazon.com/AWSJavaScriptSDK/latest/AWS/CloudWatchLogs.html
2. http://docs.aws.amazon.com/AmazonCloudWatch/latest/logs/CloudWatchLogsConcepts.html

This week I was trying to fetch the report data to the OWA through Reporting REST API. I was actually kind of stuck there as I couldn’t get the results which was supposed to be returned from the API. Last week we couldn’t have the weekly meeting with my mentor as he had some unexpected things situations. But he never forgot to send a message why he is late for the meeting.

He was so willing even to ask for a another time slot to continue this week meeting. So, we have met again on Saturday (08-July). At that day, both my mentors were present, I was so happy to see my backup mentors was also in our call. So, I explained them what I was stuck and we dig into the problem.

We debug the server and found out that some of my earlier development had a problem. Actually the exception was a NullPointerException. The reason when I request for a report through reporting REST, the defined definition library for some of the reports had actually returning null. Once I fixed that, I was able to see the results coming through the Reporting REST. I was so much happy that at the end of almost one an half weeks of struggling, I could get through that issue.

I’m still in the process of creating separate React components for each of those reports. Hope to finish that within few days. Tomorrow I have my weekly meeting. So, stay in touch for more info about my GSoC’17 experience :-)

Finally this week I’m able to fix all code incompatibility issues during STU3 migration and compile the module. It required lot of effort to understand the API changes and change the code. My mentor advised me to fully test all the features after the migration as it was subject to major changes. Now I’m going to do that during next week.

Week 6 : Client Registration fixed

After finishing migration last week, the next hurdle was to fix client registration API. The problem in the previous API was that client credentials were not properly generated and persisted in the database. After checking the local MySQL deployment, I found out that client_identifier and client_secret were both NULL. The problem was that after generation of credentials, the new state of the Client object was not saved.

So, I ran a few tests and realised the problem. Calling the method saveOrUpdate(client) again after credential generation and persistence solved the error.

Next Work : This week’s work means that now we have a migrated module with API fixed. The work in the coming week involves testing the functionality of the module, introducing the REST based Client Registration and updation mechanism made earlier, uploading everything to OpenMRS official GitHub and adding documentation of the module.

The post Week 6 Client Registration API fixed appeared first on Sanatt Abrol.

Last week I was struggling with the database issue I mentioned in last update. I was able to regain access to the database after resetting the database. After fixing the database, I started working on saving mapped report details to the database.

In this week, I was able to complete he main functionalities of my project and created a pull request for that. With the changes I did to the module, now it is possible to map any indicator of a particular Reporting module report with DHIS report data element. Mapped reports are displayed in the combo box. Currently I am waiting for the comments and feedback from my mentors. After their feedback, I’m planing to improve the solution.

In addition to the main task of my project, I worked on two tickets during this week. Both of them were regarding the DHISReport reporting module. Ticket DRM-28 was about Orgunit null pointer exception and this is the PR. Ticket DRM-9 was about handling missing parameter exception and this is the PR.

Next week I am planing to do the improvements to the project after the review of my mentors. Currently I am waiting for their feedback. That’s all for this week. :)

Build JavaScript App with Amazon Web Services

I just joined with WSO2 as an Intern for Cloud Team. After the first chat with my mentor, I was assigned to the Project about Amazon Cloud Watch Logs. Before working on that project, I prepared with AWS Javascript SDK. I have shared some useful information or the beginners here!

AWS provides JavaScript SDK for Node.js or Browsers which can allow to interact with Amazon Web services through Java Script developments. The JavaScript APIs which are available in the SDK allows developers to build libraries or applications that make use of AWS services.

Using the AWS SDK for Node.js developments

Node.js is the one of the major cross-platform run-time for running server-side JavaScript applications. So here you need to configure your environment for Node.js before working with AWS SDK. You can follow this tutorial to install Node.js for your environment.

How To Install Node.js on Ubuntu 16.04 | DigitalOcean

Now we are moving to create first JavaScript App which can interact with AWS!

First Development

• There is a sample project in the GitHub, So you can simply clone to your system. You must have Git installed to get this project from GitHub.

suthagar23/aws-nodejs-sample

If you clone the project, then You can see these followings line in the sample.js file.

// Load the SDK and UUID
var AWS = require(‘aws-sdk’);
var uuid = require(‘node-uuid’);

• It means, the JS files requires those dependencies for the development. So you need to run npm installcommand to install those required dependencies including AWS SDK.
• Now you want to configure your environment with AWS. To configure the AWS credentials, Create a credentials file at ~/.aws/credentialson Mac/Linux or C:\Users\USERNAME\.aws\credentials on Windows

In Linux, You want to go to your Home directory and press Ctrl+H to get the view of hidden folders. Then after create a directory as .aws and then create a credentials file inside the folder.

Create credentials file using this following lines,

[default]
aws_access_key_id = your_access_key
aws_secret_access_key = your_secret_key

How to get AWS Access Key and Secret Key?

If you haven’t Amazon Access Key and Secret Key already, then follow these ways to get it,

• Go to your AWA Console and then Search Identity and Access Management under Services and Click to enter IAM. In the IAM Dashboard, you can see this following menus,

• Click on the first menu(Delete your root access keys) and then click Manage Security Credentials to get your keys.

• Then after you will be redirected to Security Credentials management dashboard. At there, Click access Keys(Access Key ID and Secret Access Key menu). Now you can see Create New Access Key button.

• Click Create New Access Key button to get your Access Keys. If you click Show Access key there, then you will get your Access Key and Secret Access Keys.

Run your First Application

Now you are almost ready to run your first JavaScript application with AWS. So just type node sample.js command to execute the code file. If you configured and manged all the dependencies correctly, then most properly you will get a success message like this,

Finally, If you check your Amazon S3 account, then you can see there was a new bucket created and the hello_world.txt file was uploaded inside that bucket.

What happened? You just uploaded hello_world.txt file from your client machine to Amazon S3 cloud :-)

Image http://docs.aws.amazon.com/sdk-for-javascript/v2/developer-guide/s3-examples.html

Help from

1. http://docs.aws.amazon.com/sdk-for-javascript/v2/developer-guide/getting-started-nodejs.html
2. http://docs.aws.amazon.com/sdk-for-javascript/v2/developer-guide/installing-jssdk.html

This week was all about somehow completing AO 4. Thankfully, I have completed it and also created a PR for the same. This week was also amazing because I passed the first evaluation with my mentor being happy with me and me receiving my first proper stipend!

I will describe my work in more detail a few days later :)

For now, on to week 6!

The best part of this week, was to know that i had managed to pass the 1st evaluations. 1/3rd period of the time allotted was over, and i am glad that i have been able to adhere to the timeline. A lot of changes have been made to the original time-line. Adding Rest resources to expose the module and working on Angular directives, was something that was planned on later, since we felt that an AngularJs directive would be more useful than a UI fragment. There was just one teeny tiny problem , I had no experience with AngularJs.
This however made the project even more interesting. I for one love the learning curves involved in projects, and the challenges that come with it. With this in mind, I invested my time in understanding AngularJs with focus on directives.

What we aim on achieving is to have an angular directive that one can include on a form for an OpenmrsObject that allows a user to view/add/remove tags. They would take in the following arguments:

viewOnly (boolean): Specifies if the user can only view or can view and also add/remove tags
objectUuid (String): The uuid of the OpenmrsObject
objectType (String): The fully qualified java class name of the domain object

Some good sources for AngularJs:-
https://docs.angularjs.org/tutorial ( for basics)
https://docs.angularjs.org/guide (developers guide)

This week, I entirely worked on migrating the resources to STU3 compatible resources. It’s a very very complicated task as STU3 is subjected to several major changes when comparing to STU3. After the discussion with mentor, I created several JIRA tickets for the changes. My current changes can be seen at [1]. Created below JIRA tickets. There are several changes yet to be done.

• [FM-194] ResourceNotFoundException exception does not take Encounter class extend from Resource class as argument in STU3 migration - OpenMRS Issues
• [FM-195] Condition.Status unmapped - OpenMRS Issues
• [FM-196] No attachment in ImageStudy - OpenMRS Issues
• [FM-197] Obs with setEncounter removed STU3 - OpenMRS Issues
• [FM-198] Migrating to STU3 encounter.setIndication(indications); not found - OpenMRS Issues

[1] https://github.com/openmrs/openmrs-module-fhir/pull/132

Continue reading "GSoC 2017 – Week 5"

Week 5 [June 26, 2017 — July 03, 2017]

This is my first evaluation week for the Google Summer of Code program. The evaluation means My mentor and I want to complete a survey to pass the evaluation. The evaluation period staretd was active during June 26 — June 30, 2017.

The survey asks some questions about myself, Work done and feedback for the Organization and GSoC Program like these following,

First Evaluation Questions

• Is this your first year with your organization?
• If you’ve participated in GSoC before, which organizations did you work with?
• Did you contribute to this organization before GSoC?
• Is your project a continuation of a previous year’s project?
• When did you first communicate with your organization this year?
• How often do you interact with your mentor(s)?
• How do you communicate with your mentor(s)?
• Rate the quality of interactions with your mentor(s)
• How responsive are your mentor(s)?
• What is your favorite part of participating in GSoC?
• What is the most challenging part of participating in GSoC?
• Feedback for your organization and mentors
• Anything else you’d like to tell Google?

After the evaluation period, I received a mail about the Evaluation results. I have passed successfully :-) .
Yah , Now time to receive the stipend from Google through Payoneer.

Student Stipends

Stipends are paid in three parts, one after each successful evaluation.

• First Evaluation (paid early July): 30%
• Second Evaluation (paid early August): 30%
• Final Evaluation (paid mid September): 40%

Stipend amounts are calculated based on my location. According to my PPP, I am going to receive 720 USD at this time. (for Sri Lanka)

Development Coverage

I have advised to upload my OWA to https://modules-refapp.openmrs.org/openmrs/ for the mentor’s view. So I need to finalize the Module for the mentor’s view. I want to complete this following parts immediately at that time,

• Implemented OpenMRS Header for OWA
• Implemented OpenMRS Breadcrumbs for OWA

• Modified buttons with OpenMRS UI Icons

• Implemented Check Updates for Module feature.

Finally, I have uploaded my OWA to online reference application and received some feedback from my mentor. He advised me to start work on Encounters after some alterations on existing OWA module. So Currently I am working on existing OWA to complete the pending works.

GSoC’17@OpenMRS — Week Five

Hello Everyone,

This week I worked on issue DINT-76 in which I have to add capability to export a particular list.

I guess you’ve remembered previous posts in which I mentioned a list which generated by running a particular set of rule which check for data violations. In DINT-73, I worked on improving filtering of the list. In that issue, I have used filters provided by datatables library.

DINT-76 was also related issue. I need to add feature which enables users to export the filtered list in different formats which includes Excel, PDF and CSV. Due to plugins available in the datatables, I was able to implement it very quickly and it worked nicely.

Current UI — you can see export option on top left

One more things I need to take care of is that the UI shouldn’t very cluttered due to new features available to user. For time being — I kept it like shown in the above image. Later my mentor suggested — it would be nice to add small buttons with respective icons indication excel, pdf and csv.

I looked up on the internet and landed to fontawesome.io . It makes it very easy to add new icons — but that’s not my case. I need to find a way to integrate particular icons in the OpenMRS. I asked around and dig on community pages and found the OpenMRS is already using FontAwesome Library but it is only importing specific icons. So the whole thing boiled down to adding a bunch of new icons in the OpenMRS icon set apart from datatable thing. It the things I’m currently working on. I hope I will get it done with soon.

Thanks for reading!

Useful Links:

• FontAwesome.io
• Datatables Export Plugin

I am glad that I have passed the first evaluation, I got the approval from most people, especially the mentor and organization. I feel happy that I implement the first task and mentor Jon gave me some suggestions about specific implementation proposal for certain task. I can be better when I correct my shortcomings.

This month’s task is more challenging for me because I have not did something about front end development. I have not used JavaScript and library about it. And our project is developed by AngularJS. So this week I learn something basics about AngularJS. I have to say, front end is hard to me. But I am glad that I can get something about JS, because as a developer, we should own more technologies and broad view.

So I list something I learned this week and I prepare to learn it with two weeks, then I will write the implementation proposal about the second task. I have seen

• Expressions
• Modules
• Scope
• Controller
• Filter
• Directives

So I will learn the rest of the knowledge in next week. This two weeks are absorbing weeks. That’s all. Thank you.

This week I worked on adding charts to the patient dashboard of the android client.

Various vitals of a patient like respiration rate, pulse, body temperature etc are recorded during a visit. Currently, all the past data was displayed as a simple list to the user. A better representation was to use graphs so that proper variation in the patient’s vitals data can be easily visualised.

I intended to use William Charts[1] as the graph library but later on found that the documentation was not proper and then I did little more digging and found MPAndroidChart[2]. It is far better than William Chart because the graphs were more responsive and more features are included in it. Also the documentation was more user friendly so I used MPAndroidChart as the graph library and added graphs for the patients vitals.

Below are some of the screenshots after addition of the Graphs.

Patient dashboard home screen

Charts tab

Sample Chart

The main challenge I faced while working on the graphs was putting dates on the x-axis. The problem was that the entry object of the chart took only floats as input for x and y co-ordinates. So, I created a custom value formatter[3] and sent my date value array list of strings to the formatter and set x co-ordinate as the index in the adapter and then mapped them in the formatter.

Also, the devtest04 server was down and the JSON forms were not added on the qa-refapp server so I made a dummy patient on the qa-refapp.openmrs.org server and then downloaded the patient on the android client for testing.

According the discussion on the Openmrs talk thread[4] and issue no. AC-423[5], it was decided that the form should not be created on the manually on the server, rather it should be auto-created by the Android Client if missing. So, I also started working on it.

References :

[1] https://github.com/diogobernardino/WilliamChart

[2] https://github.com/PhilJay/MPAndroidChart/

[3] https://github.com/DefCon-007/openmrs-contrib-android-client/blob/fe58964b0ba48ad749d8ac8eca22dbae9ce642ec/openmrs-client/src/main/java/org/openmrs/mobile/utilities/DayAxisValueFormatter.java

[4] https://talk.openmrs.org/t/dynamic-form-ui-generation-from-json-schema/6707

[5]https://issues.openmrs.org/browse/AC-423?jql=project%20%3D%20AC

Last week I was able to add report definitions to the combo box and developed a triggering function for mapping buttons. In this week, I worked on connecting mapping results to the controller and modifying the dhisreport_datavalue_template table.

Instead of using the button for mapping, I added an anchor tag and connected the results to the MapReportsController. After successfully mapping the jsp with the controller, I defined two alternative columns to the dhisreport_datavalue_template table. Adding those columns was successful.

After modifying the database, I tried to modify the DataValueTemplate model class as the next step. But, still I couldn’t edit that class. Because of the issue I faced after modifying that class, my local instance got crashed. :(

Currently, I am working on restoring my instance. After that, I am planning to store the mapping data to the table within next two weeks.

That’s all for this week… :)

It’s soon after my 1st evaluation. I’m excited. I even received 30% of my stipend amount thanks to Google. This week actually I started to work on OWA. That’s another reason for me to excite as I could now learn about ReactJS view library which has developed by Facebook. first 2 days of this week I did some remaining work from previous week, which was to write unit tests. I completed it and made a PR to reference Metadata module.

I created a new repository for OWA and hosted it in my github account. This was communicated with my Mentor and he will be able to make reviews in there.

I thought to make a separate branch for each week of my work so that it will be clear enough to review my work. OpenMRS follows a standard naming convention for OWA naming. I followed, but sometimes my mentor would have a better name than what I thought of.

Currently I’m working on fetching a simple report data to my OWA via the Reporting REST API.

This was a pretty good week. I was able to complete the migration of the existing code to OpenMRS 2.x using Spring 4.1.4.RELEASE and Spring security 4.0.4.RELEASE.

The next step was fixing the API so that I can register clients and manage them. The aim was to fix the API and include the REST Client made before. The problem in the API was of the duplicate hibernate mapping. The Client class in the OAuth Module inherited the BaseOpenMRSData and overriding a few methods would mean re declaration of their instances which would cause duplicate hibernate mapping errors. I removed all the duplicate entries from the class.

Next problem? The tests won’t pass.

The test data didn’t specify UUID column and that was a must. UUID can not be kept null or undeclared. I added unique UUID values and that passed the tests.

The module ran successfull registration for the first time in a month and I was able to reach token endpoints as well.

New problem? Generation of client credentials. Client credentials were generated and encrypted for better security by the module. However, this generation and encryption failed on the Migrated module due to change in Java version. The next step involves fixing this client credentials generation issue.

The post Week 5 – Migration Complete! API fixed appeared first on Sanatt Abrol.

• Identifying what differences are made in the process when patient is an unknown.
  In RegistrationApp following configurations are made if the patient is  marked unknown.
  

  1 2 3 4 5 6

  if (unknown != null && unknown) { // TODO make "UNKNOWN" be configurable name.setFamilyName("UNKNOWN"); name.setGivenName("UNKNOWN"); patient.addAttribute(new PersonAttribute(emrApiProperties.getUnknownPatientPersonAttributeType(), "true")); }

  These are the only changes that are specific to an unknown patient in the current implementation. Other steps are common for both identified and unknown. 
• In order to move the unknown patient registration functionality to the core, these configurations should be done in the core.
• These should be configured inside the savePatient method of core.
  

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

  @Override public Patient savePatient(Patient patient) throws APIException { requireAppropriatePatientModificationPrivilege(patient); if (!patient.getVoided() && patient.getIdentifiers().size() == 1) { patient.getPatientIdentifier().setPreferred(true); } if (!patient.getVoided()) { checkPatientIdentifiers(patient); } setPreferredPatientIdentifier(patient); setPreferredPatientName(patient); setPreferredPatientAddress(patient); return dao.savePatient(patient); }

  
  
• Unknown attribute should be passed along with the patient object, and over writing name and etc should be done in core.
  
• I am planning to get the mentor's idea on this and implement the solution.

• Fetch patients considering the date created or date changed with the report date
• Fetch all the patients (except the patients that have been already fetched)

Tadaaa… I was able to make through the first evaluation of Google Summer of Code 2017. I was regularly working with my project and communicated with my mentor and others who are in the community. I am always excited to see congratulations emails from GSoC. :D

My mentor Rafal at OpenMRS is really helpful and very calm person. He is ver supportive. Without him, it would be really difficult for me to get through some of the tough times I had. So, big thanks goes out to my loyal mentor as well. :)