Rob's Musings

OpenMRS Worldwide Summit 2015 Wrapup

2015-12-28T18:41:00.002-05:00

I have been around the OpenMRS Community in one form or another since 2008 when Burke Mamlin and Paul Biondich came by the Java IRC channel on freenode to recruit for Google Summer of Code and I was drawn by the fact that I could write code that saved lives. I went on to do the first two of my three Google Summer of Code stints with OpenMRS. My first Google Summer of Code was under Burke Mamlin and Ben Wolfe; my second was under Mike Seaton and Darius Jazayeri. Recently, I am currently Manager of Identity Systems, and also am a member of the OpenMRS Infrastructure Team and am having a blast.

Three weeks ago (8-14 Dec) I got to meet all of my Google Summer of Code mentors but Ben Wolfe in person for the first time at the first annual OpenMRS Worldwide Summit in Singapore. This wasn't even the highlight. OpenMRS is a special community, one which I have yet to encounter anything similar. I was lucky enough and eternally grateful for the OpenMRS Inc. Travel Grant Program, which covered my hotel costs partially so that I could attend this amazing event.

I arrived in Singapore, quite tired a little bit after 10:00 PM SGT on 6 December, and after getting through immigration and customs and grabbed my checked bags; I grabbed the shuttle for $9 SGD, which is reasonably cheap.

I spent most of the day following sleeping off jet lag, and then ran to the mall to troubleshoot what wound up being a case of PEBKAC with my simcard; and then met up with Michael Downey and Kaweesi and grabbed some food and then headed for a preview of Wonder Full, an amazing water and light show which is amazing to see, the group of us from the summit as a whole went to see it on Friday, 11 December. One of the things that made me chuckle was the crosswalks telling me to wait for the green man...I saw that and snapped a photo as I was headed back to meet up with Michael Downey:

I volunteered and helped out with registration and general event logistics such as set-up and breakdown each day, as well as helping out with registration. On Tuesday morning, I helped bring things to the event venue (Singapore Management University). Tuesday-Wednesday (8 Dec to 9 Dec) was tutorials, either day-long or half-day. Thursday-Saturday ( 8-12 Dec) was main Conference, with both formal talks and unconference sessions(more round-table discussion), my favorite of which was a cause which I feel strongly about is the Women in Technology session, which was an amazing session. Notes from the talk can be found in the link above. Hearing about the strategic goals of OpenMRS, as well as the the talks about the OpenMRS Governance, which was amazing.

Photo by James Arbaugh, via OpenMRS Talk

The lightning talks were amazing. I originally was not going to give a talk since I did not prepare slides, but last second I typed up some notes and gave a talk trying to recruit more people to mentor for Google Code-In, and I feel extremely passionate about this program. This year and last were by far my favorite years! It's exhilarating to look at the work of these students and just be amazed. We are fostering young open source developers, some as young as 13! I got so many compliments from people saying they loved my lightning talk. I just wish more people gave talks. The ones who did, hit it out of the park.

Photo Credit: Jan Flowers, via Twitter

The after-hours activities were a chance to bond with the awesome community. There was an Informal Group Dinner at Maxwell Food Centre, The Pre-Summit BBQ at the Labrador Nature Reserve, Powerpoint Karaoke at Thoughtworks(I went with the "this is going to go bad" theme and it worked...), Chinatown Food Crawl. We also took a trip to the Marina Bay Sands for Wonder Full and a light show by Gardens By The Bay, which you can't get enough of. I had a way better seat for it this time! The after-hours activities were closed out by a fun trip to Little India, followed by a group dinner for those who went. Afterward, some of us headed to the Mustafa Centre and then met up with a group that were enjoying the beautiful Clark Quay night-life; we wound up closing out the night at Life of Liquor (LoL). The after-hours events gave the community a chance to bond.

Little India group shot (Photo from Mayank Sharma)

Photo Credit Hong Phuc Dang

On Wednesday, 9 December, Burke, Paul, Mayank and I took a walk to the Funan Tech Mall to get the prize for the scavenger hunt and grabbed lunch there. It was amazing to spend time with such amazing people.

Mayank, Paul, Burke, and myself (from left to right) (from Mayank Sharma)

One of my favorite moments was when we waiting for the light show at Gardens By the Bay, Burke spotted a kid with a Rubik's cube, pulled his out and well this was the end result, it was about as awkward and funny as it looks. Photo Credit to Mayank Sharma though, this picture needs to be included somewhere official, it's an epic battle, the kid forfeited, so I guess Burke wins by default? It's one of those you had to be there to appreciate it, but hopefully this still works somewhat.

Photo Credit to Mayank Sharma

It is also amazing that Mayank Sharma, Michael Downey, and myself (whom form 3/4 of the OpenMRS Infrastructure Team) were able to be together. Ryan Yates was missing sadly. Would have been pretty awesome to have all four of us.

Photo from Mayank Sharma

I would like to thank OpenMRS, Inc for the Travel Grant to attend. I would also like to thank Michael Downey for the Korean BBQ on Sunday (12 Dec). Though one thing I'm sad about: I never got to try the infamous Durian, the only thing which doesn't seem to a carry a fine, nobody seems to know what actually happens...This sign was found in the MRT station.

How to checkout Github Pull Requests locally

2015-03-23T19:06:00.003-04:00

Today's post centers around a need I had not too long ago. While checking student's work during Google Code-In (GCI) I needed to check their branches out locally to run it. Turns out to be easier than you think:

The first is allows me to do: gco <branch>

The second file resides in .git/config -- in my case I set it up so that origin is my local working branch and upstream is the branch I forked from. If I do gco pr/1, then it will pull from upstream and checkout the working branch.

Pretty nifty eh? Don't forget to switch back to master, or whatever your main working branch is before merging or making changes!

If you prefer a more linear history, you can cherrypick commits, but still switch to master first.

Google Summer of Code Information Session at held at CUNY Hunter College

2013-03-16T01:46:00.000-04:00

On Wednesday 13 March 2013 Sumana Harihareswara of The Wikimedia Foundation, Daniel Packer (2011 alumnus), and yours truly (three time alumnus and mentor for the past 2 years) ran an information session organized by yours truly to provide prospective students with information about Google Summer of Code.

Students received a fun and informative presentation on Google Summer of Code basics including mentoring organizations, application and program timeline, project structure, and all the awesome benefits of Google Summer of Code.

As promised students received swag provided by the Google Open Source Participation Office (including cool shirts, stickers, pens and notebooks). Fresh pizza and cold, refreshing soda was enjoyed by all.

Sumana, Daniel and I fielded questions from the approximately 30 students in attendance (all of whom showed strong interest in the program). The interests of those in attendance was broad and ranged from bioinformatics to computer vision and just about anything in between.

A trail of links was followed from the Google Summer of Code site to the mentoring organization site, and then to mentors themselves, with a discussion on how to approach particular mentors and projects. Students also enjoyed sending a greeting to the #gsoc IRC channel and receiving replies and cheers from others in the channel. All in all it was a fantastic meeting which promised to result in some excited Google Summer of Code applicants.

All in all, I was very pleased with the outcome. Everybody seemed to have a great time.

I would like to thank my professor Stewart Weiss (for his assistance in getting us a room at Hunter College) and the Computer Science department for the pizza and soda! Specific thanks go to Joseph Driscoll, assistant to the department chair whom without his help this would not have happened.

Daniel Packer (2011 Google Summer of Code alumnus) contributed in part to this write-up.

Google NYC Office tour finally happens!

2011-06-08T00:35:00.006-04:00

After three years of attempting to organize a Google NYC office visit -- we finally had that happen! The above picture is the Google NYC Reception area.

I must say that the Google NYC office is amazing! There are many places to eat and take breaks scattered throughout their office spaces on many floors -- one of which is Lego themed! They have the usual conference rooms, a computer museum -- with the usual vintage systems placed around. Did I mention the view from one of their kitchens is absolutely breath-taking? Well, it is -- just look for yourself!

I enjoyed the tour very much -- and I am sure that my fellow Google Summer of Code students did as well! For me, Google Summer of Code has been largely social -- meetups are a big component of this program in my opinion -- sure we have to produce code -- but why not have some fun mingling with students who were fortunate enough to get selected in this very selective program? This was a very successful meetup indeed.

Google Summer of Code 2010: Code merged!

2010-09-03T09:11:00.004-04:00

Been a bit since an update, most of my updates were sent to the mailing list.

My code has been since merged into sahana eden trunk.

The code in its current state is very buggy and has issues, but most issues were fixed prior to being pushed to trunk. Those issues are mentioned in the user guidelines. It needs to be updated to reflect current issues, those that have been fixed and those that are still outstanding.

Development will continue in my branch and once I feel the feature is complete and stable, it will be merged to trunk.

Weekly GSoC Survey Tool meeting summary

2010-05-31T23:32:00.005-04:00

The weekly meeting was held in the #sahana-eden IRC channel on irc.freenode.org. Present were: Mark, Gavin, Michael, and of course whatever spectators:

What was discussed:

We removed some of the question types from the functional specification to simplify things
Discussed how michael did the ADPC _next stuff and how it can be applied to this project
Gavin suggested localization -- a suggested implementation is specified -- this is a "Nice to Have" as it's outside the scope of this project for GSOC.
Will discuss on the mailing list how best to generically support "wizard" like uis within S3 elegantly.

What's on the agenda for the next couple weeks (this week will be scarce):

Noodle over ways to generically support "wizards" in S3 elegantly
Implement the "Template" stage (page 1 of the wizard) [first milestone as per my project plan timeline]

Resources relevant to this meeting for the purpose of this summary:

The functional spec
The project plan timeline
Meeting log [meeting ends at 0109 as per the timestamp in the logs]

Cheers!

Accepted to Google Summer of Code 2010: HAT TRICK!!

2010-04-26T16:54:00.008-04:00

I have been accepted to Google Summer of Code 2010 for the third year running! This summer I will be working with Sahana Eden.

The Sahana Free and Open Source Disaster Management System was conceived during the 2004 Sri Lanka tsunami. The system was developed to help manage the disaster and was deployed by the Sri Lankan government's Center of National Operations (CNO), which included the Center of Humanitarian Agencies (CHA). A second round of funding was provided by the Swedish International Development Agency (SIDA). The project has now grown to become globally recognized, with deployments in many other disasters such as the Asian Quake in Pakistan (2005), Southern Leyte Mudslide Disaster in Philippines (2006) and the Jogjarkata Earthquake in Indonesia (2006).

The project is now being ported to Python as ~~an experimental fork~~ a replacement for Sahana Agasti so that the software can be extended. Sahana Agasti is written in PHP.

I will be working with Sahana Eden to create a tool to create, enter, and manage surveys. This should be a lot of fun!

Facility Data Module v1.0 released!

2010-03-28T08:32:00.002-04:00

It is with great pleasure that I announce the release of version 1.0 of the Facility Data Module.

It is available for downloaded via the OpenMRS module repository.

Feedback is appreciated!

Google Summer of Code 2010 is a GO!

2010-01-27T20:22:00.003-05:00

Great news! Google Summer of Code 2010 is a GO!! Read the initial e-mail and the follow-up e-mail for more information!

Be sure to subscribe to the Google Summer of Code Discussion list to keep yourself informed about the program and to receive updates as they become available.

Cheers!

Facility Data updates

2010-01-18T22:47:00.005-05:00

It's been a while since I last updated on the progress of the Facility Data Module. So without farther adieu here we go:

The relation between the FacilityDataFormSchema and FacilityDataFormSection is now Many to Many meaning that sections can now be re-used between several form schemas (or reports).
It now runs without errors on OpenMRS 1.5.x

The first item was a big problem that was a sore spot for this project since it wasn't too intuitive, ideally a user should be able to re-use a section. The second item was a freebie since the errors were likely all due to the incorrect modeling.

What's next on my list is the following:

Allow for sections to be re-ordered within the schema (this is currently not possible and is a show-stopper.
Fix up the calendar management pages -- those are horribly ugly, but this is after all a first pass, so that is to be expected right?
Currently, it does not validate inputs for Numeric data types -- It need to.
There still needs to be the ability to analyze the data that exists in the system to submit to funding sources; those pages need to be implemented but that is phase two of the project.
There are likely other sore spots that exist that I do not see at this point and time that will be fixed at a later date.

That is all for now.

Google Voice tips & tricks: Send free international text messages!

2009-12-30T20:46:00.008-05:00

Just throwing this out there in case anybody wants it.

You can SMS an international number such as one from the UK to their UK number (+447xxx) using Google Voice, then once they reply you get a special Google Voice number(406-###-####) and once you get this you can SMS them as you would any number and it will relay through google voice at no cost to you!

Caveat: Calling this 406 number *WILL* cost you at the Google Voice rate for calling the UK (or whatever country you are SMS'ing to).

Happy texting!

Google Summer of Code 2009 wrap-up

2009-08-19T05:56:00.006-04:00

Google Summer of Code 2009 officially ended on August 17, 2009 at 19:00 UTC (12:00 PDT). I had an amazing time this summer. While stressful at times, it was well worth it.

I would like to thank the following people for my success this summer:

My mentor Mike Seaton for helping me throughout various misunderstandings surrounding the project's requirements. I probably made him work more than he had to. I attribute the success this summer to that one factor. This man is quite brilliant and knows his stuff.

My backup mentor Darius Jazayeri, for stepping in and helping at times and providing guidance when necessary. He is truly a brilliant man.

Ben Wolfe for answering any questions I had. He didn't have to help, but he did. I am very grateful.

Paul Biondich and Burke Mamlin for believing in me and giving me a second chance this summer. Did I mention these guys are doctors? Burke, an internist and Paul a pediatrician. Sorry for clumping you guys together.

Finally, without Leslie Hawthorn, Cat Allman, and Ellen Ko from Google, this program would not exist. They deserve a huge thank you. These three women are amazing. Thank you so much!!

Facility Data Module User Guide

2009-08-17T06:36:00.006-04:00

The Facility Data Module is a tool to collect non-patient centric aggregate data for reporting to external funding sources.

To download it use the svn version: do a checkout:
svn co http://svn.openmrs.org/openmrs-modules/facilitydata

To create a report:

1) Create all your questions

2) Create all of your form sections

3) Finally add all the sections to your form schema

I'll go into detail how to do each:

Creating/Managing questions

From the admin screen: select "Manage Questions."

Now, once on the question management page you will see a list of existing questions. under the "Action" column is a garbage can image -- this deletes the question from the database and is irreversible.

Once created, a question's data type cannot be changed, this is due to the design.

Now, once on the "Add New Question" page there are several properties. Most of which are required.

1) name - required
2) question data type - required
3) aggregation method - required
4) description - optional

Question Data Types are as follows:

NumericQuestion - a question that has a numeric answer (optionally has a min/max value and whether or not to allow decimal values.) -- these fields show only if you select "NumericQuestion."

BooleanCodedQuestion - a coded question with 3 answers: "t","f","not applicable"

StockQuestion - a question that tracks stock of vaccines, supplies, etc. This is a special coded question with the following answers: "not_stocked_out", "stocked_out", "expired", "not_applicable" and the comments field used to track the numbers of days stocked out and reason.

To add more questions after you save a question -- click "Add New Question" link above the form box.

Creating/Managing Form Sections

Once you have created all of your questions -- it's time to organize them into sections. You can easily navigate to the "Manage Form Sections" page from the "Add New Question" page.

Now, once on the section management page -- you'll see a list of all sections saved and the associated schema. Next, you'll want to click "Add New Form Section"

Now, once on the add questions page: you'll enter:

1) Display Name - required
2) Description - optional
3) Form Schema -optional (can be deferred)

Mentioned seperately is questions you're adding to the section (these are referred to as form questions by the system).
1) Name
2) Question Number
3) Question
4) Description
To add a question: click "Add New Question" to remove a question, click the "Remove" button next to the button (see image after the one below).

The section below is a section that was defined programmatically in a mock schema that was designed and used throughout the summer.
As you can see, there is a "Remove" button next to each Form Question in the section.

Creating/Managing Form Schemas

From the Section creation page: click "Manage Form Schemas"

After clicking that link you will see the page below. It allows you to delete a schema, and information relevant to te form schema.

Now, after clicking the "Add New Form Schema" link you will see the page below. The following information is needed:
1) Display Name - required
2) Data Entry Frequency -required
3) Validity Period -optional
4) Description - optional
5) Sections specified in this form schema (and the ordering) -- just drag them into the left-side of the pane.

Now, for the shortcomings:

Ordering of questions cannot be changed once saved
Sections cannot be associated with two schemas at once.

Viewing and Entering Data

Will be completed at a later date.

Project Update

2009-08-04T06:51:00.004-04:00

So we're in the home stretch folks! 13 days to go until the "Pencils Down" date! Hope everybody's project is going well. Here's how mine is going:

Reports can be saved and reloaded (with all relevant data filled out (fields etc -- reports can be viewed in editable or "view only" mode).
Each component of the report (Questions, Form Questions(allows a question to exist on multiple reports), and Sections can be created in isolation
A page to view the status of a report (complete, partially complete or incomplete) is done for reports that are to be entered daily in a monthly calendar view; monthly still needs to be completed.

Overall, i feel the project is progressing well -- I expect to finish either on time, or early. (Let's hope for early!)

Cheers!

Midterm Progress Update

2009-07-16T01:31:00.001-04:00

Progress has been slow and steady. I'm on track to finish however! What's done:

Reports can be entered, saved, and the values entered will be reloaded into the proper form fields. If edited and if the values have changed -- the old values are voided and new values are saved (retaining the old values in the database for auditing purposes).

What's left:

Management pages See this and this for what it will ultimately look like.
Still need to get the "View Only" pages working; in theory it should work. The "View Only" will be the answers but no form fields present.
The UI to create the report, questions, sections, etc.

Ultimately, I think this shouldn't be too bad.

The overdue progress report

2009-06-21T00:00:00.009-04:00

Apologies for not blogging as much as I should. I've focused on getting what needed to be done, done.

Tasks that have been completed thus far with a target milestone of using a mock Form, enable persistence of Questions and Values to the database:

Mock out a report form using the domain classes
Using the mocked up schema, generate a simple report form
Design the SQL Schema (for just FacilityDataValue and FacilityDataQuestion)
Write Hibernate Mapping files (for just FacilityDataValue and FacilityDataQuestion)
Write Data Access layer (for FacilityDataValue and FacilityDataQuestion)
Write Service layer (for FacilityDataValue and FacilityDataQuestion)
Refactor the rendering logic to use the JSP and write EL function(s) to check types using instanceof.
Allow a simple the mocked form from Week 1 to save the question answers.

Tasks that are are in progress, soon to be finished with a target milestone of removing the code used to mock up everything from the first few weeks; ability to use the saved schemas for rendering the report form:

Design SQL Schema for the rest of the domain classes
Write mappings for the rest of the domain classes
Support loading the previously saved values for a form/startdate/enddate/location into a page for viewing or editing
Write methods to save the rest of the domain classes in the data access layer
Write methods to save the rest of the domain in the service layer

Now that I have summarized work completed and in progress, let's explain the overall design:

FacilityDataFormSchema serves as the overall representation of the report form in the system.
FacilityDataFormSection is simply that, sections on the form, e.g., monitoring equipment status, stock status of vaccinations, number of people vaccinated, etc.
FacilityDataFormQuestion holds metadata regarding a question.
FacilityDataQuestion is the question itself; it specifies the datatype; it is subclassed for each question datatype; if not subclassed, then the question is considered to be "freetext" -- in other words: just simply a text-based question.

CodedQuestion is a question that has a coded answer. This too is subclassed for each coded question datatype.

StockQuestion is exactly as the name says, to track stock of items such as vaccinations. The coded answers are: "not_stocked_out","stocked_out","expired","not_applicable"
BooleanCodedQuestion is a simple "yes","no","not applicable" type of thing; e.g., "Was there mobile clinic today?"

NumericQuestion is a question which has a numeric answer, e.g., "Number of Adults Vaccinated."

FacilityDataValue is what holds the values entered in the report forms for each question.
FacilityDataReportFormData is a non-persisted class used for retrieving the answers for a specific report for a specified period.

Hopefully this makes up for my lack of updates.

The long overdue progress report

2009-06-12T14:39:00.006-04:00

This is gonna be short, sweet and to the point:

Week 1:

I wrote a mock report form with 10 questions (screenshots in a later post)
I then wrote up some code to render it approriately for each question type.

Week 2 (still ongoing):

Designed the SQL schema, and wrote the service/database layer classes for 2 of the classes.
Wrote in the functionality to save the report data to the database.

Will explain the design at a later date.

Ciao!

Open Everything NYC Recap

2009-04-21T00:11:00.010-04:00

I attended Open Everything NYC on Saturday. For those who do not know, Open Everything is an "un-conference" which is a conference where the participants run the show. There are several open sessions, and a few keynotes.

Some of the open session topics were on education, Open Source Software licensing, and even relationships! The Open Relationships talk was really interesting. I find it amusing how many people wind up creating alter egos to prevent employers from finding out things about them.

(Some of) the keynote speakers were amazing:

Schuyler Erle, a guy who wrote some code for UNICEF, utilized text messaging as a means to track distribution of Bed Nets to prevent malaria via a web application, among other things. If you want to research this it's all on the wiki.

Robert Steele, a former spy (yes a spy!) whose talk in my opinion was pretty much utterly useless! I saw diagrams which were well.. horrid. His talk had ZERO flow planning, it seemed to be a bunch of conspiracies, one after the other. Let alone the fact the guy is abrasive. After his talk, I went up to him to let him know that his site (which he linked in his talk), was down; his response was condescending and RUDE. I turned to some people behind me as I was leaving, and said that there was no way I could have summarized anything that guy said (I was writing notes for the wiki); they agreed. The consensus on Robert Steele was pretty universal: Conspiracy nutjob.

The final speaker, Leslie Hawthorne, a woman I have a lot of respect for gave a talk that was short, sweet, and to the point. She disussed how open source software development can be used in a common sense way to solve complex problems we face. OpenMRS, solving the problem of the HIV/AIDS crisis in the developing world; Sahana for the disaster relief management in areas devastated by floods and such. Amazing talk! Best part, she did not use any slides!

John Britton, the conference organizer gave the closing talk, and he took notes as to what should change next time. Longer open sessions etc came up. Additionally, we found out: it's possible to do this ongoing, I'd REALLY enjoy that. I met some freaking awesome people.

We were to screen RIP: "A REMIX MANIFESTO" at McFaddens Bar and Pub by the UN, but the owner didn't allow us to do it; which was stupid. We wound up sitting around half price drinks/appetizers and had some good conversations and fun.

Overall, I had fun and would do it again. I went into this not knowing what to expect.

Accepted into Google Summer of Code 2009

2009-04-20T17:05:00.012-04:00

For the second year in a row, I was accepted into Google Summer of Code!!! I will be working with OpenMRS on the Facility Data Module.

My project will focus on developing tools for collecting and ultimately generating reports for aggregate data from external sources.

Groovy Forms Module is NOT dead

2009-02-08T17:09:00.004-05:00

Hey folks!

I'm still alive and not dead. The Groovy Forms Module is not a dead project, and WILL be finished. I plan on getting work started back up on it soonish. School is back in session for me. I'm sorry for the long delay.

That is all.

GSoC wrap-up

2008-09-10T10:02:00.004-04:00

The program officially ended on August 18^th. I'm awaiting the arrival of my t-shirt!!

Now, I didn't quite have time to finish my project prior to that date since my uncle suddenly passed away on August 10^th.

The project itself is about 80 percent completed. All that is left is to get the rendering working and finish up the management page to edit the form metadata as it appears in the system.

My time is now limited since school started up again for me, so as much time as I can I'm going to devote to finishing up this project.

I'd like to thank the following:

Burke Mamlin for helping me and answering all the questions I had and guiding me when I was lost.

Ben Wolfe for putting up with me and helping me when Burke was MIA.

OpenMRS and its wonderful worldwide network of developers for making this project into what it has become.

Paul Biondich for encouraging me (along with Burke) to apply for summer of code.

Last, but certainly not least, Leslie Hawthorn for managing this program and every problem that popped up. She must have super human powers or something.

The list can only be so long, I feel like I'm accepting an academy award here. It was a great experience, and I will definately continue to maintain my project in whatever free time I can find.

GSoC 2008: We're in the home stretch

2008-08-10T21:20:00.003-04:00

Well folks, I can't believe it's almost over. What a journey. I never imagined I'd be able to do this. I have an (almost) working project. I have a few more things that need to be done.

Let's highlight what IS done:

GroovyForm and its related metadata is stored
Model class is interrogated for its properties and those are stored in a container class.
From the data collected in the container class, I generate markup checking for a predefined set of data types.
The view and its related controller is generated -- sneaking in some groovy magic of course!
GroovyForms are successfully saved into the system, along with their respective, model, view and controller.
Forms are persisted when the module is shutdown and reloaded when it is restarted

What's left:

Editing metadata related to the forms currently in the system
Writing the servlet which handles submissions
Create some sample forms
Documentation

I did plan to add a lot of other stuff such as the ability to create custom form widgets, however there were a few problems which caused me to redirect my focus. These will come in after I finish with Summer of Code.

GSOC 2008: Week 5

2008-07-02T02:47:00.017-04:00

So, it's time for an update. This week got off to a sketchy start, but it's gained momentum. Let me enumerate what I have done thus far with the project. First, I've added AJAX using jquery. I figured I would use jquery mainly because it provided painless AJAX goodness. Prior to even thinking of using jquery, I wrote my own AJAX code using a tutorial that I found while googling. I had to tweak it a bit, but for the most part, it seemed very standard. However, after I wrote the AJAX equivalent code, it didn't feel too elegant. So, first I'm going to show the AJAX code I wrote; then I'll show you the jquery version; finally, I'm going to show you the servlet which handles the AJAX on the server-side.

First, the AJAX I wrote:


function AjaxValidation(url, callback) {
    var req = init()
    req.onreadystatechange = processRequest

    function init() {
        if (window.XMLHttpRequest) {
            return new XMLHttpRequest()
        } else if (window.ActiveXObject) {
            return new ActiveXObject("Mircosoft.XMLHTTP")
        }
    }

    function processRequest() {
        if (req.readyState == 4) {
            if (req.status == 200) {
                if (callback) {
                    chkSyntaxCallBack(req.responseXML)
                }
            }
        }
    }

    this.doGet = function() {
        req.open("GET", url, true)
        req.send(null)
    }
}
function chkSyntaxCallBack(responseXML) {
    var res = responseXML.getElementsByTagName("result")[0].firstChild.nodeValue
    document.getElementById("out").innerHTML = res
}

function checkSyntax() {
    var target = document.getElementById("groovyModel")
    var url = "${pageContext.request.contextPath}/moduleServlet/groovyforms/createGroovyForm?groovyModel=" + escape(target.value)
    var ajax = new AjaxValidation(url, chkSyntaxCallBack)
    ajax.doGet()
}

I warned you that it wasn't too elegant. Understanding this isn't too hard. Here is the call sequence: init() -> doGet() -> processRequest() -> chkSyntaxCallBack(). Not that bad. Now let's see the jquery version.


$(window).ready(function () {
    $("#groovyModel").bind("blur", function () {
        $.ajax({
            type: 'POST',
            data: { groovyModel: $("#groovyModel").val() } ,
            url: "${pageContext.request.contextPath}/moduleServlet/groovyforms/createGroovyForm" ,
            cache: false ,
            success: function(data) {
                var res = $(data).find("result").text()
                $("#out").html(res)

            }
        })
    });
})

Okay, that's much better. A few things are still happening here. When the window is finished loading, I bind my textarea element which has the CSS id of "groovyModel" to the blur event (lost focus). Then you see the AJAX. Now this is very straight forward. We're using the POST method, we're sending whatever value is inside of the textarea at the time the event is fired, we're posting to a servlet, not going to cache, and when we're done, it's printed to the screen. Very straight forward.

Now, like I said, this is all backed on the server-side by a servlet, which is written in Groovy. So here we go:


/**
 * The contents of this file are subject to the OpenMRS Public License
 * Version 1.0 (the "License") you may not use this file except in
 * compliance with the License. You may obtain a copy of the License at
 * http://license.openmrs.org
 *
 * Software distributed under the License is distributed on an "AS IS"
 * basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See the
 * License for the specific language governing rights and limitations
 * under the License.
 *
 * Copyright (C) OpenMRS, LLC.  All Rights Reserved.
 */
package org.openmrs.module.groovyforms.web

import javax.servlet.ServletException
import javax.servlet.http.HttpServlet
import javax.servlet.http.HttpServletRequest
import javax.servlet.http.HttpServletResponse
import org.apache.commons.logging.LogFactory
import org.codehaus.groovy.control.CompilationFailedException

class CreateGroovyFormServlet extends HttpServlet {
  def classLoader
  static final def log = LogFactory.getLog(CreateGroovyFormServlet.class)
  private static final long serialVersionUID = 066373513262051L

  @Override
  protected void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
    def generateTemplate = request.getParameter("template")
    def generateController = request.getParameter("controller")
    def finalMarkup = request.getParameter("markup")
    def clazz = URLDecoder.decode(request.getParameter("groovyModel"))
    def name = request.getParameter("formName")
    def version = request.getParameter("version")
    def res = this.checkSyntax(clazz)
    if (clazz) {
      if (checkSyntax(clazz)) {
        response.contentType = "text/xml"
        response.setHeader "Cache-Conrol", "no-cache"
        response.writer.write "\n\t$res\n"
      } else {
        response.contentType = "text/xml"
        response.setHeader "Cache-Control", "no-cache"
        response.writer.write "true"
      }
    } else {
      response.contentType = "text/xml"
      response.setHeader "Cache-Control", "no-cache"
      response.writer.write "\n\tPlease fill in the Form Model\n"
    }
  }

  @Override
  protected void doPost(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
    doGet(request, response)
  }

  @Override
  void init() throws ServletException {
    if (log.infoEnabled)
      log.info("Initializing...")
    classLoader = getClassLoader()
  }

  def getClassLoader() {
    def gcl = new GroovyClassLoader(this.getClass().getClassLoader())
    gcl
  }

/**
 * This method is used to relay errors to the user
 * @param clazz the class
 * @return the exception message or null if it was successful
 */
  def checkSyntax(clazz) {
    def sb = new StringBuilder()
    sb << "import org.openmrs.*\n\n\n"
    sb << clazz
    def res = null
    try {
      getClassLoader().parseClass(sb.toString())

    } catch (CompilationFailedException e) {
      res = "Exception: ${e.message}"
    }
    res

  }

/**
 * Check if it is result groovy code.
 * @param clazz the class
 * @return whether or not it is result groovy code
 */
  def isValidGroovy(clazz) {
    def sb = new StringBuilder()
    sb << "import org.openmrs.*\n\n\n"
    sb << clazz
    try {
      getClassLoader().parseClass(sb.toString())
    } catch (CompilationFailedException e) {
      return false
    }
    return true
  }
}

This servlet contains a lot of utility methods. One compiles, one initializes/returns the GroovyClassLoader, and of course doGet(), doPost() and init(). doGet() and doPost() both do the same thing, with doPost() simply delegating to doGet(). The code should be reasonably easy to understand. checkSyntax() returns null if it was parsed cleanly, otherwise it returns the exception message, the stack track wouldn't be useful in my case. It returns an XML tag <result> with "true" if it was successful, the exception message if it was not, and a message stating that the field must be filled in if it's empty or just not passed in. I implicitly import org.openmrs.* to allow for easier access to the OpenMRS domain model classes. The parent classloader for the GroovyClassLoader is set the servlet container's classloader. This gives me access to the classpath when loading Groovy classes.

I still have a bit to do, templating needs to be written in, for the most part it's done. Just have a few problems I'm facing, but I'll get through it. Too many people are relying on me succeeding. I feel that this project could help a lot of people, so i feel pressure to succeed. I still need to write in the "Edit" functionality of the "Manage Groovy Forms" page.

I'm definately making progress. More updates to come, that's for sure.

GSoC 2008: Groovy Forms getting some AJAX!

2008-06-21T18:48:00.010-04:00

No, not the cleaning product! The "Web 2.0" kind of AJAX! Now where will this goodness be placed? The pages for creating the forms and managing the forms.

First off, I used Direct Web Remoting (DWR) to provide a link between my backend java codebase and the front-end javascript! It's quite amazing. You can choose to publish/unpublish the forms with the tick of a checkbox and DWR will magically set the published attribute on the form to the proper state. I then decided to go one step furthur, and provide a mechanism to backup the forms via xstream to XML. Now, with the XML files, two files are maintained, one individual XML file for each form, and one global one, which makes up the model for the entire system. The rationale for having the XML file for each form is that you can just zip up the form folder, and unzip it into another OpenMRS install and you're good to go. Adding the ability to re-generate the XML files via the management interface is especially handle so that if you make a quick change on one system, then want to install it onto another you can.

Now, DWR will be used for creating the forms as well. This will be done mostly to easily relay errors, such as syntax errors in the Controller and the Model of the groovy form. Don't fear, I'm going to go overboard with AJAX! It'll only be used up until the point of where templates are generated. After that point, a Servlet will take over the bulk of the work (I am undecided on the Servlet bit.).

Now what's next? Several things need to happen.

First, creating the form:

We need to validate that form fields are not empty; also the input must be checked for validity.
The model's syntax needs to be compiled and syntax errors need to be relayed to the user.
We need to get the fields of the class and store them.
We need to generate templates like: <%= textArea(...); %> -- They will not be run it through the template engine immediately until the user has confirmed what they want,
After the view has been tweaked -- it is run through the template engine; at this point we will also run the controller through and generate that and present it at the same time the final view is presented.
The user will be presented with the HTML of the view and Groovy controller and given a final opportunity to tweak it before it is saved to the system.
The form is added to the system -- by default the form is not published.
Finally, to publish the form so that it can be used to enter data, they go to the management screen and tick the checkbox.
The form is now ready to be used for data entry.

Pictures are worth a thousand words so here we go: One and Two.

First one contains one form which is published, the rest are not. The second, no forms are published.

The user interface is intuitive and easy to use. It's a great design, of course I'm biased since It was was me who designed it.

GSoC 2008: Updates and such

2008-06-14T23:55:00.007-04:00

Okay, let's see what was accomplished in the past few weeks:

I tweaked the domain class interrogation code a bit. Basically, now there is a 1:1 relationship between a field and the domain model metadata storage container class instance.
I've started the templating work -- most of it right now is just methods which will be called like: <%= textField(...) %> -- that will generate a text field html form widget.

I know I haven't accomplished much, but there's a reason, I swear!

June 11th I went to Rockville, MD to the Hypoparathyroidism Association's 3rd Annual Hypoparathyroidism Conference. I just got home tonight.

Let me stray off a bit to explain what hypoparathyroidism is. Basically, the parathyroid is a gland in your neck, underneath the thyroid. Most people get it via surgical error, where the surgeon cuts out, or damages the parathyroid. When this occurs, the PTH (Parathyroid Hormone) is no longer produced. Now, this is responsible for regulating among other things, Calcium. Without this present, the calcium levels in the blood drop, and patients start to experience Tetnay, where their muscles involuntarily contract. It's not fun. If the levels drop low enough, you can go into convulsions, and ultimately seizures. Death is also a possibility, which is why it's important for patients with hypoparathyroidism to take their meds (Calcium Supplements, sometimes Magnesium, Vitamin D (usually D2 or D3), and a medication which is the active form of Vitamin D called Calcitriol.) Some patients also take other meds as well. These must be continued for the rest of their lives. I got Hypoparathyroidism idiopathically -- which means the origin nor the cause is known.

Back to the point: I was at this conference, and it's such a rare disease that I don't know many people. So I wound up talking with the people there. As a result, not much work was done. Monday, I plan to get the templating working, then probably do some Front end UI work.