Add sha and time stamp keys to JSON output

[mienkoja]

These commits implement several changes to provide 2 new keys to the output of get_metric_list(): calc_date, and calc_sha.

• calc_date will provide a timestamp (via lubridate::today()) indicating the time that a measurement object was initialized (e.g. via build_all_metrics()).
• calc_sha will provide any text within /app/R/install-packages.R which matches the following REGEX expression: (?<=oliveR@).*(?=\\"). In principle, this should only be the SHA listed in the install-packages.R script within the oliver-metrics-api image. If this script does not exist in a given environment (e.g. due to running oliveR locally), the key should return 'SHA information not available'.

[mkornblum]

I don't have a lot of context for this, why do we need these pieces of information? Are we planning to display them in the UI somehow?

[mkornblum]

if we just return from inside the if/else here we could eliminate the assignment to sha_value as well as line 8

[mkornblum]

Confirmed calc_date and calc_sha are present in responses to the get_metric_list endpoint:

% curl -i -k -X POST 'http://localhost:443/ocpu/library/oliveR/R/get_metric_list/json' -H "Content-Type: application/json" -d '{"group_id":"7"}'         :( 1 17-07-12 - 12:13:07
HTTP/1.1 200 OK
Date: Wed, 12 Jul 2017 19:21:16 GMT
Server: libwww-perl-daemon/6.01
Content-Type: text/plain
Content-Length: 1377
Last-Modified: Wed, 12 Jul 2017 19:21:16 GMT

[{"id":"7","acceptance_to_schedule":{"characteristic":"Days Until Visit is Scheduled","characteristic_label":"Expected Days to Schedule","measurement_missing":true,"label":"Expected Days to Schedule","threshold":false,"template":"default","calc_date":"2017-07-12","calc_sha":"bfbc2c690012d934bd12267780cb7e2e3f96b5c7"},"acceptance_to_first_visit":{"characteristic":"Days Until First Visit, as Planned","characteristic_label":"Expected Days to First Planned Visit","measurement_missing":true,"label":"Expected Days to First Planned Visit","threshold":false,"template":"default","calc_date":"2017-07-12","calc_sha":"bfbc2c690012d934bd12267780cb7e2e3f96b5c7"},"child_count_value":{"characteristic":"Children per Referral","characteristic_label":"Children per Referral","characteristic_summary_value":1.7,"measurement_missing":false,"value":1.7,"label":"Children per Referral","threshold":false,"template":"default","calc_date":"2017-07-12","calc_sha":"bfbc2c690012d934bd12267780cb7e2e3f96b5c7"},"attendance_per_scheduled_visit":{"characteristic":"Percentage of Scheduled Visits, which Were Attended","characteristic_label":"Visit Attendance Rate","characteristic_summary_value":"23%","measurement_missing":false,"value":"23%","label":"Visit Attendance Rate","threshold":false,"template":"default","calc_date":"2017-07-12","calc_sha":"bfbc2c690012d934bd12267780cb7e2e3f96b5c7"}}]

So functionality is verified and I am fine to 🚢 but still curious what the intent of the change is

[mkornblum]

Another idea is that we could include these at a higher level of the response, like up where id is, since they are metadata that is the same for the whole response, instead of having to include them on every calculation object.

[mienkoja]

Matt can chime in here with other uses, but the idea as I understand it is to provide us some contextual information that can help us confirm that PPM has indeed been updated on a given day, and the build of oliveR from which it was updated. I would be fine moving both of these pieces of information up so that the metrics are all organizationally nested within metadata. I could envision future metric layouts that would have variable update schedules, but this will certainly work for PPM as designed. I'll work on that tonight unless I hear otherwise.

…

On Wed, Jul 12, 2017, 12:21 PM Mark Kornblum ***@***.***> wrote: Confirmed calc_date and calc_sha are present in responses to the get_metric_list endpoint: % curl -i -k -X POST 'http://localhost:443/ocpu/library/oliveR/R/get_metric_list/json' -H "Content-Type: application/json" -d '{"group_id":"7"}' :( 1 17-07-12 - 12:13:07 HTTP/1.1 200 OK Date: Wed, 12 Jul 2017 19:21:16 GMT Server: libwww-perl-daemon/6.01 Content-Type: text/plain Content-Length: 1377 Last-Modified: Wed, 12 Jul 2017 19:21:16 GMT [{"id":"7","acceptance_to_schedule":{"characteristic":"Days Until Visit is Scheduled","characteristic_label":"Expected Days to Schedule","measurement_missing":true,"label":"Expected Days to Schedule","threshold":false,"template":"default","calc_date":"2017-07-12","calc_sha":"bfbc2c690012d934bd12267780cb7e2e3f96b5c7"},"acceptance_to_first_visit":{"characteristic":"Days Until First Visit, as Planned","characteristic_label":"Expected Days to First Planned Visit","measurement_missing":true,"label":"Expected Days to First Planned Visit","threshold":false,"template":"default","calc_date":"2017-07-12","calc_sha":"bfbc2c690012d934bd12267780cb7e2e3f96b5c7"},"child_count_value":{"characteristic":"Children per Referral","characteristic_label":"Children per Referral","characteristic_summary_value":1.7,"measurement_missing":false,"value":1.7,"label":"Children per Referral","threshold":false,"template":"default","calc_date":"2017-07-12","calc_sha":"bfbc2c690012d934bd12267780cb7e2e3f96b5c7"},"attendance_per_scheduled_visit":{"characteristic":"Percentage of Scheduled Visits, which Were Attended","characteristic_label":"Visit Attendance Rate","characteristic_summary_value":"23%","measurement_missing":false,"value":"23%","label":"Visit Attendance Rate","threshold":false,"template":"default","calc_date":"2017-07-12","calc_sha":"bfbc2c690012d934bd12267780cb7e2e3f96b5c7"}}] So functionality is verified and I am fine to 🚢 but still curious what the intent of the change is — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#24 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AF0tRc-sNDwuf7Q4BOJoEG5nCL9otwDAks5sNRy6gaJpZM4OVw-G> .