Struct move_stackless_bytecode::loop_analysis::LoopAnalysisProcessor

pub struct LoopAnalysisProcessor {}

Implementations

impl LoopAnalysisProcessor

pub fn new() -> Box<Self>

Trait Implementations

impl FunctionTargetProcessor for LoopAnalysisProcessor

fn process(
    &self,
    _targets: &mut FunctionTargetsHolder,
    func_env: &FunctionEnv<'_>,
    data: FunctionData
) -> FunctionData

Processes a function variant. Takes as parameter a target holder which can be mutated, the env of the function being processed, and the target data. During the time the processor is called, the target data is removed from the holder, and added back once transformation has finished. This allows the processor to take ownership on the target data.

fn name(&self) -> String

Returns a name for this processor. This should be suitable as a file suffix.

fn process_and_maybe_remove(
    &self,
    targets: &mut FunctionTargetsHolder,
    func_env: &FunctionEnv<'_>,
    data: FunctionData
) -> Option<FunctionData>

Same as process but can return None to indicate that the function variant is removed. By default, this maps to Some(self.process(..)). One needs to implement either this function or process.

fn initialize(&self, _env: &GlobalEnv, _targets: &mut FunctionTargetsHolder)

A function which is called once before any process call is issued.

fn finalize(&self, _env: &GlobalEnv, _targets: &mut FunctionTargetsHolder)

A function which is called once after the last process call.

fn is_single_run(&self) -> bool

A function which can be implemented to indicate that instead of a sequence of initialize, process, and finalize, this processor has a single run function for the analysis of the whole set of functions.

fn run(&self, _env: &GlobalEnv, _targets: &mut FunctionTargetsHolder)

To be implemented if is_single_run() is true.

fn dump_result(
    &self,
    _f: &mut Formatter<'_>,
    _env: &GlobalEnv,
    _targets: &FunctionTargetsHolder
) -> Result

A function which creates a dump of the processors results, for debugging.