Re: Understand pvget and pvxget

[Michael Davidsaver via Tech-talk <tech-talk at aps.anl.gov>]

Hi Tong,

Firstly, the pvx* CLI tools from PVXS are so far not recommended
for end users since they are not as "friendly" as would be expected.
There is a moderately sized TODO project here.

https://urldefense.us/v3/__https://github.com/epics-base/pvxs/issues/66__;!!G_uCfscf7eWS!eOwRR_1vh8ILXAdmGDb9OP4pToo7RjtWPg2f4cMmaVYpAQ2kCJ9GzsCvmxxP0cIrMWUDJjZLkyf2-d1MUEy4Y-RV0Q$ 

> Where I tried applying `-r value` option to the latter, and hoping it
> would print only the value field ...
The handling of pvRequest does not in fact work like dereference as
you (and most users) first expect.  Rather, pvRequest acts as a bit
mask to select a set of field names.

So "-r value" or "-r value,alarm" are equally valid.  In either case
the network operation will yield a top level PVA Structure containing
a member field 'value' which is marked as "changed".

Another way of phrasing this.  The pvRequest bit mask is applied
to all PVA delta values.  So unselected fields values will never
be sent, and will never appear as "changed" to the client.

> ... some issues are with the tools.
Yes!

The pvtools from pvAccessCPP are not consistent about rendering
partial Structures in "nt" format.  In part this is a difficult
problem since PVA is so flexible.  There are a very large number of
possible permutations for even the "simple" NTScalar schema, and it
would be difficult to define how each should be rendered.


On 1/7/25 11:54, Zhang, Tong via Tech-talk wrote:
> Happy new year, community!
>
> Probably a dummy question here: how to correctly use the "-r <pv
> request>" option of `pvxget` or `pvget`?
>
> First of all, the tool versions I'm tested:
> $ pvxget -V:
> PVXS 1.3.2 (1.3.1-41-g87abad5387c22045cb5e-dirty)
> EPICS 7.0.8.1
> libevent 2.2.1-alpha-dev
>
> $ pvget -V:
> pvAccess 7.1.7
> pvData 8.0.6
> Base 7.0.8.1
>
> And the record testing against:
> record(calc, "pva:test:counter") {
>      field(INPA, "pva:test:counter")
>      field(CALC, "A+1")
>      field(SCAN, "1 second")
> }
>
> The default behavior of the two commands are slightly different:
> $ pvget pva:test:counter  # just outputs the value, which is I expected
> pva:test:counter 2025-01-07 14:39:44.950  2663
> $ pvxget pva:test:counter  # outputs the full string, is not I
> expected, but    some options may help
> pva:test:counter
>      struct "epics:nt/NTScalar:1.0"
>      value double = 2708
>      alarm.severity int32_t = 0
>      alarm.status int32_t = 0
>      alarm.message string = "NO_ALARM"
>      timeStamp.secondsPastEpoch int64_t = 1736278829
>      timeStamp.nanoseconds int32_t = 950403289
>      timeStamp.userTag int32_t = 0
>      display.limitLow double = 0
>      display.limitHigh double = 0
>      display.description string = ""
>      display.units string = ""
>      display.precision int32_t = 0
>      display.form.index int32_t = 0
>      display.form.choices string[] = {7}["Default", "String", "Binary",
> "Decimal", "Hex", "Exponential", "Engineering"]
>      control.limitLow double = 0
>      control.limitHigh double = 0
>      control.minStep double = 0
>      valueAlarm.active bool = false
>      valueAlarm.lowAlarmLimit double = nan
>      valueAlarm.lowWarningLimit double = nan
>      valueAlarm.highWarningLimit double = nan
>      valueAlarm.highAlarmLimit double = nan
>      valueAlarm.lowAlarmSeverity int32_t = 0
>      valueAlarm.lowWarningSeverity int32_t = 0
>      valueAlarm.highWarningSeverity int32_t = 0
>      valueAlarm.highAlarmSeverity int32_t = 0
>      valueAlarm.hysteresis int8_t = 0
>
> Where I tried applying `-r value` option to the latter, and hoping it
> would print only the value field, but it still acts the same. So that
> brought up the question at the very beginning.
>
> As per the documentation at
> https://urldefense.us/v3/__https://epics-base.github.io/pvxs/overview.html*sub-field-lookup__;Iw!!G_uCfscf7eWS!ZTNZy0ZfewN7Pbf068ihPM9Im5m94ieAKdmc_jbwxEAIt7IA2x4hfw5YGU5OrolGrnUIMQTseB29vj8Q0W6-yqti$ , I
> assume the "value" is the one of the *fields* that should be defined in
> `-r` option, but I cannot find the precision definition.
>
> Intuitively, I tried `pvget -r "timeStamp" pva:test:counter` hoping to
> get only the timestamp info, but it outputs the same as without the `-
> r` option.
>
> So either is my understanding to this *-r option* is wrong, or I passed
> the wrong field, that is my understanding to *field* is wrong, or some
> issues are with the tools.
>
> Very much appreciate to any hints.
>
> Cheers,
> Tong