ENH: A new experiment command

[ke-zhang-rd]

Expected Behavior

What will happen after execuate bsui

1. Welcome to XXX beamline. May I have your name:
   >>> 

   If you know your gup number, type in below:
   >>>

   If you know your saf number, type in below:
   >>>

2. Create session.json and store above data, if session.json exist, overwrite it.
3. init conda and ipython(other bsui task)

After entering bsui(ipython)

• new_experiment(gup=0, saf=0, name='Def'...) could overwrite session.json
• list_experiment_info will print session.json
• end_experiment will delete session.json
• RE() will check if session.json exist. If it's, load it as equivalent RE.md['saf'] = "123456"

Options bsui --develop will skip prompt to make development easier bsui --continuewon't overwrite existing session.json

Possible Solution

1. Three functions in packagenslsii. BMM has a example in 08-user.py

• new_experiment
• list_experiment_info
• end_experiment

2. Update bsui. A good source of prompt code might be cookiecutter. It's depend on click

3. A location and format of file session.json, maybe:

   {
   'name': "Abc",
   'date': "yyyy-mm-dd HH:mm:ss.SSS",
   'gup': "123456",
   'saf': "123456"
   }

[ke-zhang-rd]

@stuartcampbell @danielballan

This is a checklist of features of new command function

• [ ] 1. It's a python module
• [ ] 2. A API(function) take user's input
• [ ] 3. prefer use prompt.py from cookiecutter
• [ ] 4. It should have a schema to verify user's input's format and type
  • [ ] a. name: str
  • [ ] b. gup: int
  • [ ] c. saf: int
• [ ] 5. It should have one or several TODO functions for future authentication verify with PASS database, return True now
• [ ] 6. return a dict or update through dict reference in argument
• [ ] 7. tests and docs

[danielballan]

For (3) an IPython magic might be used instead, since it can occur in the same process where RE.md is defined. But I think that's an open question, and we may want both.

For (4) I am unclear on whether we want to collect the saf or just the proposal number. (Is that what "gup" means?") I think we should treat them all as strings, because we are not in control of the specification and have to assume that they may not always be integers.

[prjemian]

Use a string, makes it future proof. APS has at least three kinds of proposal. It is useful to include the kind of proposal as a prefix.

On Mon, Jun 8, 2020, 12:31 PM Dan Allan notifications@github.com wrote:

For (3) an IPython magic might be used instead, since it can occur in the same process where RE.md is defined. But I think that's an open question, and we may want both.

For (4) I am unclear on whether we want to collect the saf or just the proposal number. (Is that what "gup" means?") I think we should treat them all as strings, because we are not in control of the specification and have to assume that they may not always be integers.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/NSLS-II/nslsii/issues/85#issuecomment-640768467, or unsubscribe https://github.com/notifications/unsubscribe-auth/AARMUMHWYFOYBKMCZEC2ZQDRVUN6TANCNFSM4NIDJSKQ .

[prjemian]

NeXus went through this same decision tree and agreed, finally, string was the robust choice.

On Mon, Jun 8, 2020, 12:36 PM Pete Jemian prjemian@gmail.com wrote:

Use a string, makes it future proof. APS has at least three kinds of proposal. It is useful to include the kind of proposal as a prefix.

On Mon, Jun 8, 2020, 12:31 PM Dan Allan notifications@github.com wrote:

For (3) an IPython magic might be used instead, since it can occur in the same process where RE.md is defined. But I think that's an open question, and we may want both.

For (4) I am unclear on whether we want to collect the saf or just the proposal number. (Is that what "gup" means?") I think we should treat them all as strings, because we are not in control of the specification and have to assume that they may not always be integers.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/NSLS-II/nslsii/issues/85#issuecomment-640768467, or unsubscribe https://github.com/notifications/unsubscribe-auth/AARMUMHWYFOYBKMCZEC2ZQDRVUN6TANCNFSM4NIDJSKQ .