registers
This module contains all the functionality and operations for using and manipulating registers.
These often have a role in performing context-sensitive parsing tasks, where a Turing-powerful
system is required. While flatMap
is capable of such parsing, it is much less efficient
than the use of registers, though slightly more flexible. In particular, the persist
combinator
enabled by RegisterMethods
can serve as a drop-in replacement for flatMap
in many scenarios.
Attributes
- Since:
2.2.0
- Source:
- registers.scala
- Graph
- Supertypes
- Self type
- registers.type
Members list
Registers
The Reg
type is used to describe pieces of state that are threaded through a parser.
The creation and basic combinators of registers are found within Reg
and its companion
object.
This class is used to index registers within the mutable state.
This class is used to index registers within the mutable state.
Attributes
- Since:
2.2.0
- Note:
it is undefined behaviour to use a register in multiple different independent parsers. You should be careful to parameterise the registers in shared parsers and allocate fresh ones for each "top-level" parser you will run.
- Companion:
- object
- Source:
- registers.scala
- Graph
- Supertypes
This object allows for the construction of a register via its make
function.
This object allows for the construction of a register via its make
function.
Attributes
- Companion:
- class
- Source:
- registers.scala
- Graph
- Supertypes
- Self type
- Reg.type
Register-Based Combinators
Some combinators are made much more efficient in the presence of registers and they can be found here.
This combinator allows for the repeated execution of a parser in a stateful loop.
This combinator allows for the repeated execution of a parser in a stateful loop.
forP(init, cond, step)(body)
behaves much like a traditional for loop using init
, cond
, step
and body
as parsers
which control the loop itself. First, a register r
is created and initialised with init
. Then cond
is parsed, producing
the function pred
. If r.gets(pred)
returns true, then body
is parsed, then r
is modified with the result of parsing step
.
This repeats until r.gets(pred)
returns false. This is useful for performing certain context sensitive tasks.
Attributes
- body
the body of the loop performed each iteration.
- cond
the condition by which the loop terminates.
- init
the initial value of the induction variable.
- step
the change in induction variable on each iteration.
- Returns:
a parser that initialises some state with
init
and then parses body untilcond
is true, modifying the state each iteration withstep
.- See also:
forYieldP
for a version that returns the results of eachbody
parse.- Example:
the classic context sensitive grammar of
anbncn
can be matched usingforP
:val r = Reg.make[Int] r.put(0) *> many('a' *> r.modify(_+1)) *> forP[Int](r.get, pure(_ != 0), pure(_ - 1)){'b'} *> forP[Int](r.get, pure(_ != 0), pure(_ - 1)){'c'}
- Source:
- registers.scala
This combinator allows for the repeated execution of a parser body
in a stateful loop, body
will have access to the current value of the state.
This combinator allows for the repeated execution of a parser body
in a stateful loop, body
will have access to the current value of the state.
forP_(init, cond, step)(body)
behaves much like a traditional for loop using init
, cond
, step
and body
as parsers
which control the loop itself. First, a register r
is created and initialised with init
. Then cond
is parsed, producing
the function pred
. If r.gets(pred)
returns true, then body
is parsed, then r
is modified with the result of parsing step
.
This repeats until r.gets(pred)
returns false. This is useful for performing certain context sensitive tasks.
Attributes
- body
the body of the loop performed each iteration, which has access to the current value of the state.
- cond
the condition by which the loop terminates.
- init
the initial value of the induction variable.
- step
the change in induction variable on each iteration.
- Returns:
a parser that initialises some state with
init
and then parses body untilcond
is true, modifying the state each iteration withstep
.- See also:
forYieldP_
for a version that returns the results of eachbody
parse.- Example:
the classic context sensitive grammar of
anbncn
can be matched usingforP_
:val r = Reg.make[Int] r.put(0) *> many('a' *> r.modify(_+1)) *> forP_[Int](r.get, pure(_ != 0), pure(_ - 1)){_ => 'b'} *> forP_[Int](r.get, pure(_ != 0), pure(_ - 1)){_ => 'c'}
- Source:
- registers.scala
This combinator allows for the repeated execution of a parser in a stateful loop.
This combinator allows for the repeated execution of a parser in a stateful loop.
forYieldP(init, cond, step)(body)
behaves much like a traditional for comprehension using init
, cond
, step
and body
as parsers
which control the loop itself. First, a register r
is created and initialised with init
. Then cond
is parsed, producing
the function pred
. If r.gets(pred)
returns true, then body
is parsed, then r
is modified with the result of parsing step
.
This repeats until r.gets(pred)
returns false. This is useful for performing certain context sensitive tasks. Unlike forP
the
results of the body invokations are returned in a list.
Attributes
- body
the body of the loop performed each iteration.
- cond
the condition by which the loop terminates.
- init
the initial value of the induction variable.
- step
the change in induction variable on each iteration.
- Returns:
a parser that initialises some state with
init
and then parses body untilcond
is true, modifying the state each iteration withstep
. The results of the iterations are returned in a list.- See also:
forP
for a version that ignores the results.- Example:
the classic context sensitive grammar of
anbncn
can be matched usingforP
:val r = Reg.make[Int] r.put(0) *> many('a' *> r.modify(_+1)) *> forYieldP[Int](r.get, pure(_ != 0), pure(_ - 1)){'b'} *> forYieldP[Int](r.get, pure(_ != 0), pure(_ - 1)){'c'}
This will return a list
n
'c'
characters.- Source:
- registers.scala
This combinator allows for the repeated execution of a parser body
in a stateful loop, body
will have access to the current value of the state.
This combinator allows for the repeated execution of a parser body
in a stateful loop, body
will have access to the current value of the state.
forP_(init, cond, step)(body)
behaves much like a traditional for comprehension using init
, cond
, step
and body
as parsers
which control the loop itself. First, a register r
is created and initialised with init
. Then cond
is parsed, producing
the function pred
. If r.gets(pred)
returns true, then body
is parsed, then r
is modified with the result of parsing step
.
This repeats until r.gets(pred)
returns false. This is useful for performing certain context sensitive tasks. Unlike forP_
the
results of the body invokations are returned in a list.
Attributes
- body
the body of the loop performed each iteration, which has access to the current value of the state.
- cond
the condition by which the loop terminates.
- init
the initial value of the induction variable.
- step
the change in induction variable on each iteration.
- Returns:
a parser that initialises some state with
init
and then parses body untilcond
is true, modifying the state each iteration withstep
.- See also:
forP_
for a version that ignores the results of the body.- Example:
the classic context sensitive grammar of
anbncn
can be matched usingforP_
:val r = Reg.make[Int] r.put(0) *> many('a' *> r.modify(_+1)) *> forYieldP_[Int](r.get, pure(_ != 0), pure(_ - 1)){_ => 'b'} *> forYieldP_[Int](r.get, pure(_ != 0), pure(_ - 1)){_ => 'c'}
This will return a list
n
'c'
characters.- Source:
- registers.scala
Register Extension Combinators
These are implicit classes that, when in scope, enable additional combinators on parsers that interact with the register system in some way.
This class, when in scope, enables a method to create and fill a register with a given value.
This class, when in scope, enables a method to create and fill a register with a given value.
Attributes
- x
the value to initialise a register with.
- Constructor:
This constructor should not be called manually, it is designed to be used via Scala's implicit resolution.
- Source:
- registers.scala
- Graph
- Supertypes
This class, when in scope, enables a method to create and fill a register with a given value.
This class, when in scope, enables a method to create and fill a register with a given value.
Attributes
- x
the value to initialise a register with.
- Constructor:
This constructor should not be called manually, it is designed to be used via Scala's implicit resolution.
- Source:
- registers.scala
This class, when in scope, enables the use of combinators directly on parsers that interact with the register system to store and persist results so they can be used multiple times.
This class, when in scope, enables the use of combinators directly on parsers that interact with the register system to store and persist results so they can be used multiple times.
Attributes
- con
a conversion that allows values convertible to parsers to be used.
- p
the value that this class is enabling methods on.
- Constructor:
This constructor should not be called manually, it is designed to be used via Scala's implicit resolution.
- Source:
- registers.scala
- Graph
- Supertypes
This class, when in scope, enables the use of combinators directly on parsers that interact with the register system to store and persist results so they can be used multiple times.
This class, when in scope, enables the use of combinators directly on parsers that interact with the register system to store and persist results so they can be used multiple times.
Attributes
- con
a conversion that allows values convertible to parsers to be used.
- p
the value that this class is enabling methods on.
- Constructor:
This constructor should not be called manually, it is designed to be used via Scala's implicit resolution.
- Source:
- registers.scala