anonymous wrote on Saturday, May 08, 2010:
We plan to enhance and expand the number of patient reports that are available in OpenEMR. They will enable OpenEMR to meet the certification requirements of Patient Lists. Please review the proposed solution below:
tmccormi wrote on Saturday, May 08, 2010:
I like it.
-Tony
anonymous wrote on Tuesday, June 01, 2010:
Initial coding for Patient Lists has been completed. You can log into http://www.medbloom.com to see the reports. For data range, please start with 2008/01/01.
The codes can be reviewed at http://github.com/thomaswong2/openemr
bradymiller wrote on Saturday, June 12, 2010:
Thomas,
What is difference in this patch from the new one. I have both patches on my git repository and a diff between the two shows no changes:
http://github.com/bradymiller/openemr/tree/thomas_report_1
http://github.com/bradymiller/openemr/tree/thomas_report_2
thanks,
brady
bradymiller wrote on Saturday, June 12, 2010:
oops,
I got confused. Forget my above confusion. Have yanked my second branch, so now have your code here only:
http://github.com/bradymiller/openemr/tree/thomas_report_1
-brady
bradymiller wrote on Saturday, June 12, 2010:
Thomas,
Placed my code review in the tracker:
http://sourceforge.net/tracker/index.php?func=detail&aid=3011938&group_id=60081&atid=1245239#
thanks,
-brady
anonymous wrote on Thursday, June 24, 2010:
A new patch has been submitted to the Tracker.
bradymiller wrote on Tuesday, June 29, 2010:
Thomas,
An important aside is none of the search functioning is working; I could
get nothing to return for nay search (except that all pt get returned when
click a ‘diagnosis’ search).
To deal with other code issues, I’ve contributed quite a bit to your code,
and put the patch here. To see what changes I made from yours, can see my
commit in my github branch here:
http://github.com/bradymiller/openemr/commits/thomas_report_3
Specifically my commit is:
http://github.com/bradymiller/openemr/commit/9c0ea0d0b6594b71f956211f70bc98bd22649385
My contributions are:
—INCORPORATED NEW SECURITY METHODS ((set global reverse magic quotes
flag, set no fake globals flag, use sql binding/placemakers(stop
sql-injection), use htmlspecialchars(stop xss attacks)) Note no need for
formdata.inc.php functions anymore when use this stuff.
—FIXED incorrect demographics2.php links
—CHANGED FULL PHP tags, <?php instead of <?
—USED generate_display_field() for output of a list item.
—USED dropdown_facility function
Thread and links to examples to explain the new security implementations
can be found here:
http://sourceforge.net/projects/openemr/forums/forum/202506/topic/3530656
Stuff you still need to do:
-Incorporate this recent additions from
interface/reports/prescriptions_report.php:
http://github.com/openemr/openemr/commit/212bc5a8137c4ce8dc98e3a2e50123c762ca6b54
-Please explain real quick the need/use for the date range stuff (I’m
confused by it)
-HAVE THE SEARCHING ACTUALLY WORK
-brady
anonymous wrote on Wednesday, June 30, 2010:
Please make sure that you have data in all the tables. The search functions work on local computer as well as at www.medbloom.com. You can log in to see them.
We will review your patch and incorporate the suggested changes shortly.
bradymiller wrote on Wednesday, June 30, 2010:
Thomas,
My changes are extensive in order for your programmers to learn the new security implementation (you’ll be better off just starting from my patch because the extensive htmlspecialchars use basically makes back patching almost impossible). These security steps are discussed here:
http://sourceforge.net/projects/openemr/forums/forum/202506/topic/3530656
It actually makes the code much cleaner, because do not need to use the confusing formdata.inc.php functions anymore (sort of a lie, rarely will need to escape a variable if it’s a column label in a sql statement).
New pages should follow this method (changes to old pages can use old formdat.inc.php method unless page has been converted to the new method).
-brady
anonymous wrote on Saturday, July 10, 2010:
The changes have been made. Please review the codes.
1. Note that the following tables need to have data:
Diagnosis
- patient_data (Main)
- lists (Second)
- form_encounter (Date)
Procedure
- patient_data (Main)
- procedure_order (Second)(Date)
- procedure_report
- procedure_type
Prescription
- prescriptions (Main) (Date)
- drugs
- drug_sales
- drug_inventory
Medical History
- patient_data (Main)
- history_data (Date)
Lab Results
- patient_data (Main)
- procedure_result (Date)
Service Codes
- patient_data (Main)
- billing (Date)
The main data is from ‘Main’ table.
Tables marked with “Date” also need to have data. The date (From, To) is compared with the date field in those tables.
2. The date range is used for filtering and displaying the search results. It’s common for a lot of reports.
bradymiller wrote on Monday, July 12, 2010:
Thomas,
TESTING REVIEW:
–Need to ensure the functionality of the original prescription report (that you deleted) is still working. In your code, if I create a new user and make a prescription for the new user, it does not show up in the report. However, it does show up in old deleted report.
–No point in me testing the rest of the functionality until above works. I will say that I noted the age filter is not working (if I use it in the diagnosis function, it will include patients that should not be included (ie. do not fit within the filter). Please ensure comprehensive testing before next submission, since this is getting close to being able to commit to the codebase.
Also, base your patches on the most current cvs codebase. I think your now basing it off your first version off your reports; since I’m not sure what your basing it off, this is another source of potential issues with your code when we test it.
thanks,
-brady
bradymiller wrote on Monday, July 12, 2010:
Thomas,
Another testing issue is that the patient name links in the report output do not work (they always go to the current active patient).
-brady
anonymous wrote on Tuesday, July 13, 2010:
Sorry about the patch. I applied the latest file without first applying your patch. I have resubmitted the patch. There are only a few lines of changes.
Visolve will do the QA. The patient links work at www.medbloom.com. So we will wait to hear from Visolve.
bradymiller wrote on Tuesday, July 13, 2010:
Thomas,
Make your patches from the most recent cvs(either SF development tip or the github master branch), so anybody can test your stuff. I can easily compare across your branches from git.
Why are you sending code to QA that does not work in the development tip (having it work at www.medbloom.com isn’t useful for the codebase). You guys need to test and get your stuff working from the development tip. One cardinal rule of open source is to not break functionality; the prescription report obviously does not work like the previous report. Sending code that doesn’t even fulfill this minimal testing requirement to QA by visolve is just wasting resources. You guys need to get in the habit of testing from the tip.
thanks,
-brady
anonymous wrote on Tuesday, July 13, 2010:
Of course, the programmer is using patch from Github Master. It’s tested to work on his local machine and at www.medbloom.com so that everyone can see the changes and that everyone can see the development and testing are done.
We didn’t have access to the SF Tip or the Github Master before (remember)? Please provide instructions if there is another way to test.
anonymous wrote on Tuesday, July 13, 2010:
Also, starting today, the programmer(s) will work with GitHub directly under username “ossllc”. I will monitor their work from time to time. Please grant access to that account.
bradymiller wrote on Tuesday, July 13, 2010:
Thomas,
Thomas,
You guys aren’t using git in an ideal way. I suggest not ever touching your master or rel-320 branches on your repository; you only update these from the official github openemr repository. I’d suggest making another repository(even better for each programmer to have their own repository, since it’s very easy to grab branches from each other). I recommend initially following these instructions, which I created for new git users to get started “correctly” with OpenEMR:
http://www.openmedsoftware.org/wiki/Git_for_dummies
For testing, I place this script in the git directory (one above openemr), and run it whenever I want to do a native test (on mandriva, but can quickly script any OS):
#!/bin/bash
rm -fr /var/www/html/openemr/*
mysqladmin -f -h localhost -u root drop openemr
mysql -f -u root -h localhost -e "DELETE FROM mysql.user WHERE User = 'openemr';FLUSH PRIVILEGES;"
rsync --recursive --exclude .git openemr/* /var/www/html/openemr/
chmod 666 /var/www/html/openemr/library/sqlconf.php
chown -R apache:apache /var/www/html/openemr/documents
chown -R apache:apache /var/www/html/openemr/edi
chown -R apache:apache /var/www/html/openemr/era
chown -R apache:apache /var/www/html/openemr/library/freeb
chown -R apache:apache /var/www/html/openemr/custom/letter_templates
chown -R apache:apache /var/www/html/openemr/interface/main/calendar/modules/PostCalendar/pntemplates/compiled
chown -R apache:apache /var/www/html/openemr/interface/main/calendar/modules/PostCalendar/pntemplates/cache
chown -R apache:apache /var/www/html/openemr/gacl/admin/templates_c
service mysqld restart
So running above script will place newly created version of whatever git branch is checked out to be tested. Doing this kind of testing is vital (this is why most programmers, such as Rod, myself, Aron, Visolve and others can check in straightforward code without a QA cycle). If you then combine this testing with correct use of git/github, then the sky it the limit for your programmers.
-brady
atlanticcan wrote on Tuesday, July 13, 2010:
I don’t suppose anyone has created a script for Windows?
Thanks,
Marc
blankev wrote on Tuesday, July 13, 2010:
Medbloom. What happened with the User and Password?
I can’t remember mine it seems………………