Support OTN attenuator and amplifier

[jjin62]

1. Add a new precision attribute flag for optical analog values (e.g., gain, tilt, power), instead of using double/float point type which would introduce NaN issue.
2. Add support for optical attenuator in experimental extensions, derived from openconfig-optical-attenuator.yang.
3. Add support for optical amplifier in experimental extensions, derived from openconfig-optical-amplifier.yang.

[jjin62]

1. Introducing two new SAI objects OTN-OA and OTN-ATTENUATOR using SAI experimental extension mechanism. OA and ATTENUATOR are optical modules which can form a simplest optical amplifier.
2. Gauged/Analog values (optical power, gain, attenuation etc..) are usually represented in a float point format (double or float). However, double/float data type is prohibited in SAI to avoid NaN issue (see meta/Makefile compile flag "-Wfloat-equal"). Therefore, a @precision tag is introduced to represent the number of digits after the decimal point. Any integer type with a @precision tag can be used to represent a decimal type.
   Ref SAI Support optical transport device - proposed by sonic-wg-otn #2216

[azure-pipelines]

Commenter does not have sufficient privileges for PR 2217 in repo opencomputeproject/SAI

[kcudnik]

{ to new line, and add empty libe before return same in line 434