Add support for dates to Ptolemy

[cxbrooks]

Note: the issue was created automatically with bugzilla2github tool

Original bug ID: BZ#506 From: Elizabeth Latronico <beth@berkeley.edu> Reported version: 9.1.devel CC: ptango@chess.eecs.berkeley.edu

Blocker for: BZ#504

[cxbrooks]

• Allow picking a date range for UCBPowerServer demo. Get to work with new.openbms.org website (note: URLs in current UCBPowerServer demo only return 94 points max).

• Show different output based on time of request. Able to get time of request from client? From server?

• Type system: Add new DateToken class to Ptolemy – copy from Kepler? What conversions are allowed?

• Actors: Copy from Kepler, Mail from Christopher:

BTW - Kepler has a DateToken kepler/util/src/org/kepler/date/DateToken.java

And some DateActors:

./actors/src/org/kepler/date/CreateDate.java ./actors/src/org/kepler/date/DateDifference.java ./actors/src/org/kepler/date/DateToString.java ./actors/src/org/kepler/date/RandomDate.java

The DateToken uses a java.util.Date as its base object.

Using a Long might make more sense, but from an ontological perspective, having a separate DateToken would be more type safe.

If we had a separate DateToken, would it be losslessly convertable to and from a LongToken?

If we added a DateToken, then maybe we could remove the duplicate code in Kepler?

_Christopher

[cxbrooks]

• Look at open source SW for web page dates

On 6/16/12 4:17 AM, Edward A. Lee wrote:

Another one (this is BSD):

http://microba.sourceforge.net/

Edward

---

Edward A. Lee EECS, UC Berkeley eal@ eecs.berkeley.edu

On Jun 16, 2012, at 4:12 AM, "Edward A. Lee"<eal@ eecs.berkeley.edu> wrote:

This looks potentially useful for entering dates:

http://www.toedter.com/en/jcalendar/index.html

It is LGPL.

We could create a date style for the Query class and a DateParameter that sets its type to long.

Edward

[cxbrooks]

• Info on data point limit from new.openbms.org: Hi everyone,

I’ve experimentally determined that the /backend/ interface, e.g. URLs of the format http://new.openbms.org/backend/api/data/uuid/18866fbe-53ff-5511-86c7-95d1cea0a061?start=1337140800000&end=1336622400000

returns a max of 94 points regardless of larger start/end time pairs, and regardless of the a limit parameter instead of start/end parameters e.g. ?limit=100

I’m not sure what the sampling interval is either. For e.g. Birge Hall if I get the last two data points: http://new.openbms.org/backend/api/data/uuid/18866fbe-53ff-5511-86c7-95d1cea0a061?limit=2

I get the result: [{"uuid": "18866fbe-53ff-5511-86c7-95d1cea0a061", "Readings": [[1339357500000.0, 427.10000000000002], [1339358400000.0, 432.60000000000002]]}]

So the interval is 1339358400000.0 - 1339357500000.0, i.e. 900000 somethings 

I think we can work around this in different ways – e.g. the /plot/ interface appears to handle longer time frame, or we could break up a long request to /backend/ into multiple 94-point requests. But, I think it’s fine without a time interval for the demo tomorrow. It would be nice to have but this will take a bit of work.

Best, Beth