monasca-analytics/monasca_analytics/banana
gecong1973 c9ea9be579 Add __ne__ built-in function
In Python 3 __ne__ by default delegates to __eq__ and inverts the
result, but in Python 2 they urge you to define __ne__ when you
define __eq__ for it to work properly [1].There are no implied
relationships among the comparison operators. The truth of x==y
does not imply that x!=y is false. Accordingly, when defining __eq__(),
one should also define __ne__() so that the operators will behave as
expected.
[1]https://docs.python.org/2/reference/datamodel.html#object.__ne_

Change-Id: I52633a8a4b19c5e0c0d7d786f21770496e128c7a
2016-11-24 10:00:01 +08:00
..
bytecode Add the test back and include BytecodeAssembler in tree. 2016-07-05 11:11:34 +01:00
cli Fix a typo in documentation 2016-10-04 10:54:08 +07:00
deadpathck Fix crash when there's no connection in the banana file. 2016-09-21 15:22:26 +01:00
eval This commit introduces the first version of Banana configuration language. 2016-08-22 14:29:26 +01:00
grammar Add __ne__ built-in function 2016-11-24 10:00:01 +08:00
typeck Add __ne__ built-in function 2016-11-24 10:00:01 +08:00
README.md This commit introduces the first version of Banana configuration language. 2016-08-22 14:29:26 +01:00
__init__.py Add the test back and include BytecodeAssembler in tree. 2016-07-05 11:11:34 +01:00
emitter.py Improve error message and span calculation. 2016-09-21 15:55:24 +01:00
pass_manager.py Add Banana specific APIs to typecheck and get list of components. 2016-11-04 11:26:17 +00:00

README.md

Banana configuration language

This module contains everything related to Banana. In each sub-module (sub-folder) you will find a README.md file that describes:

  • Purpose of the module.
  • The current status of the implementation.
  • How testing is done.

The compiler is split in passes. Each pass performs some transformations and / or generates more data. Only the last step has side-effects on the Monanas instance.

Each sub-module roughly maps to one pass run by the compiler.

Passes

The Banana compiler runs the following passes:

  • parse, parse the input and build an AST.
  • typeck, type check the input.
  • deadpathck, remove dead path in the connections.
  • eval, evaluate the AST generated.

Each pass makes some assumptions about the state of the data, and in particular that the previous passes have run successfully. While this is made obvious by the arguments required to run some passes, it is less so for others.

Generally, things to remember:

  • Changing the ordering of passes is more likely to break things.
  • New passes are free to modify the AST / TypeTable.
  • New passes should not break invariants.

For more information on passes, have a look in their specific README.md file.