Instruction.backward_df()

Returns a list of all previous instructions/arguments/variables of the current point in the data flow graph.

backward_df() → APISet[Point]

The backward_df() function is an intra-procedural analysis function. This means that the function does not operate recursively and instead returns instruction/argument/variable within the current function instruction set.

The function returns the derived classes from Point, such as ArgumentPoint, VarValue, Instruction, etc.

Query Example

from glider import *


def query():
  # Fetch an instruction
  instructions = Instructions().with_callee_name('verify').exec(1)
  
  for points in instructions[0].backward_df():
    print(points.source_code())

  # Return the list of previous instructions of the current instruction
  return instructions

Output Example

For the same contract, this query showcases that the return list elements are type-casted from Point:

from glider import *


def query():
  # Fetch an instruction
  instructions = Instructions().with_callee_name('verify').exec(1)
  
  for points in instructions[0].backward_df():
    if isinstance(points, ArgumentPoint):
      print(points.get_variable().name)
    print(points.source_code())

  # Return the list of previous instructions of the current instruction
  return instructions

The function returns APISet, instead of APIList, in case the result of the function is used as the return value of the query it must be casted to list()