1 PLAI Scheme
1.1 Defining Types: define-type
define-type
1.2 Deconstructing Data Structures: type-case
type-case
1.3 Testing Infrastructure
test
test/ pred
error
test/ exn
test/ regexp
1.3.1 Test Flags
abridged-test-output
plai-catch-test-exn
halt-on-errors
print-only-errors
test-inexact-epsilon
plai-ignore-exn-strings
plai-all-test-results
2 GC Collector Scheme
2.1 Garbage Collector Interface
heap-size
location?
root?
heap-value?
heap-set!
heap-ref
get-root-set
read-root
set-root!
procedure-roots
with-heap
2.2 Garbage Collector Exports
init-allocator
gc: deref
gc: alloc-flat
gc: cons
gc: first
gc: rest
gc: set-first!
gc: set-rest!
gc: cons?
gc: flat?
3 GC Mutator Scheme
3.1 Building Mutators
allocator-setup
3.2 Mutator API
set-first!
set-rest!
import-primitives
3.3 Testing Mutators
test/ location=?
test/ value=?
printf
4 Web Application Scheme
start
5 Generating Random Mutators
save-random-mutator
find-heap-values

Programming Languages: Application and Interpretation

This is the documentation for the software accompanying the textbook Programming Languages: Application and Interpretation (PLAI). The full book can be found on the Web at:

http://www.cs.brown.edu/~sk/Publications/Books/ProgLangs/

In DrScheme, under the Language menu, select Choose Language.... Under the section Programming Languages: Application and Interpretation, you will find the following languages:

1 PLAI Scheme

#lang planet plai/plai:1:18

PLAI Scheme is derived from the scheme langauge. In addition, it includes the define-type and type-case forms and testing support.

1.1 Defining Types: define-type

(define-type type-id variant ...)
 
variant = (variant-id (field-id contract-expr) ...)
Defines the datatype type-id. A constructor variant-id is defined for each variant. Each constructor takes an argument for each field of its variant.

The value of each field is checked by its associated contract-expr. A contract-expr may be an arbitrary predicate or a contract.

In addition to the contructors, a define-type expression also defines:

1.2 Deconstructing Data Structures: type-case

(type-case datatype-id expr
   branch ...)
 
branch = (variant-id (field-id ...) result-expr ...)
  | (else result-expr ...)
Branches on the datatype instance produced by expr, which must be an instance of datatype-id (previously defined with define-type) Each branch extracts the values of the fields, and binds them to field-id ....

If a branch is not specified for each variant, you may use an else branch to create a catch-all branch. An else branch must be the last branch in the sequence of branches. type-case signals a compile-time error if all variants are not covered and the else branch is missing. Similarly, type-case signals a compile-time error if an else branch is unreachable because a branch exists for all variants.

1.3 Testing Infrastructure

PLAI Scheme provides the following syntactic forms for testing.

(test result-expr expected-expr)
If result-expr and expected-expr evaluate to the same value, result-value, the test prints

(good result-expr result-value expected-value location).

If they do not evaluate to the same value, the test prints

(bad result-expr result-value expected-value location).

If evaluating result-expr signals an error, the test prints

(exception result-expr exception-message <no-expected-value> location)

If evaluating expected-expr signals an error, the test prints

(pred-exception result-expr exception-message <no-expected-value> location)

(test/pred result-expr pred?)
Similar to test, but instead of supplying an expected value, the predicate pred? is applied to result-expr.

If evaluating pred? signals an error, the test prints

(pred-exception result-expr exception-message <no-expected-value> location)

The syntax of pred? is considered expected-value for the purposes of test reporting.

Like scheme’s error, but generates exceptions that are caught by test/exn.

(test/exn result-expr error-message)
This test succeeds if the expression evaluates to a call to error. Moreover, the error message contained in the exception must contain the string error-message. Note that test/exn only suceeds if the exception was explicitly raised by the user.

For example, the following test succeeds:

  (test/exn (error "/: division by zero") "by zero")

The error message is "/: division by zero", and "by zero" is a substring of the error message. However, the following test fails:

  (test/exn (/ 25 0) "by zero")

Although the expression raises an exception and the error string contains "by zero", since the error was not explicitly raised by user-written code, the test fails.

The evaluation of error-message is considered expected-value for the purposes of test reporting.

(test/regexp result-expr error-message-regexp)
This test is similar to test/exn,but the error message is matched against a regular expression instead.

The evaluation of error-message-regexp is considered expected-value for the purposes of test reporting.

1.3.1 Test Flags

(abridged-test-output [abridge?])  void?
  abridge? : boolean? = false
When this flag is set to true, the test forms never prints result-expr or location.

(plai-catch-test-exn [catch?])  void?
  catch? : boolean? = true
When this flag is set to true, exceptions from tests will be caught. By default, exceptions are caught.

(halt-on-errors [halt?])  void?
  halt? : boolean? = true
This flag determines whether the program immediately halts when a test fails. By default, programs do not halt on failures.

(print-only-errors [print?])  void?
  print? : boolean? = true
When this flag is set to true, only tests that fail will be printed. By default, the results of all tests are printed.

(test-inexact-epsilon epsilon)  void?
  epsilon : number?
When testing inexact values for equality, test permits them to differ by epsilon. The default value of epsilon is 0.01.

(plai-ignore-exn-strings ignore?)  void?
  ignore? : boolean?
If this flag is set to true, when testing for exceptions with test/exn and test/regexp, the message of the exception is ignored. By default, test/exn and test/regexp only succeed when the message of the exception matches the supplied string or regular expression.

This variable is the list of all tests that have been run so far, with the most recent test at the head.

2 GC Collector Scheme

#lang planet plai/plai:1:18/collector

GC Collector Scheme is based on PLAI Scheme. It provides additional procedures and syntax for writing garbage collectors.

2.1 Garbage Collector Interface

The GC Collector Scheme language provides the following functions that provide access to the heap and root set:

Returns the size of the heap. The size of the heap is specified by the mutator that uses the garbage collector. See allocator-setup for more information.

(location? v)  boolean?
  v : any/c
Determines if v is an integer between 0 and (- (heap-size) 1) inclusive.

(root? v)  boolean?
  v : any/c
Determines if v is a root.

(heap-value? v)  boolean?
  v : any/c
A value that may be stored on the heap. Roughly corresponds to the contract (or/c boolean? number? procedure? symbol? empty?).

(heap-set! loc val)  void?
  loc : location?
  val : heap-value?
Sets the value at loc to val.

(heap-ref loc)  heap-value?
  loc : location?
Returns the value at loc.

(get-root-set id ...)
Returns the current roots as a list. Local roots are created for the identifiers id as well.

(read-root root)  location?
  root : root?
Returns the location of root.

(set-root! root loc)  void?
  root : root?
  loc : location?
Updates the root to reference the given location.

(procedure-roots proc)  (listof root?)
  proc : procedure?
Given a closure stored on the heap, returns a list of the roots reachable from the closure’s environment. If proc is not reachable, the empty list is returned.

(with-heap heap expr ...)
 
  heap : (vectorof heap-value?)
Evaluates (begin expr ...) in the context of heap. Useful is tests:
  (test (with-heap (make-vector 20)
          (init-allocator)
          (gc:deref (gc:alloc-flat 2)))
        2)

2.2 Garbage Collector Exports

A garbage collector must define the following functions:

init-allocator is called before all other procedures by a mutator. Place any requisite initialization code here.

(gc:deref loc)  heap-value?
  loc : location?
Given the location of a flat Scheme value, this procedure should return that value. If the location does not hold a flat value, this function should signal an error.

(gc:alloc-flat val)  location?
  val : heap-value?
This procedure should allocate a flat Scheme value (number, symbol, boolean, closure or empty list) on the heap, returning its location (a number). The value should occupy a single heap cell, though you may use additional space to store a tag, etc. You are also welcome to pre-allocate common constants (e.g., the empty list). This procedure may need to perform a garbage-collection. If there is still insufficient space, it should signal an error.

Note that closures are flat values. The environment of a closure is internally managed, but contains references to values on the heap. Therefore, during garbage collection, the environment of reachable closures must be updated. The language exposes the environment via the procedure-roots function.

(gc:cons first rest)  location?
  first : location?
  rest : location?
Given the location of the first and rest values, this procedure must allocate a cons cell on the heap. If there is insufficient space to allocate the cons cell, it should signal an error.

(gc:first cons-cell)  location?
  cons-cell : location?
If the given location refers to a cons cell, this should return the first field. Otherwise, it should signal an error.

(gc:rest cons-cell)  location?
  cons-cell : location?
If the given location refers to a cons cell, this should return the rest field. Otherwise, it should signal an error.

(gc:set-first! cons-cell first-value)  void?
  cons-cell : location?
  first-value : location?
If cons-cell refers to a cons cell, set the head of the cons cell to first-value. Otherwise, signal an error.

(gc:set-rest! cons-cell rest-value)  void?
  cons-cell : location?
  rest-value : location?
If cons-cell refers to a cons cell, set the tail of the cons cell to rest-value. Otherwise, signal an error.

(gc:cons? loc)  boolean?
  loc : location?

Returns true if loc refers to a cons cell. This function should never signal an error.

(gc:flat? loc)  boolean?
  loc : location?
Returns true if loc refers to a flat value. This function should never signal an error.

3 GC Mutator Scheme

#lang planet plai/plai:1:18/mutator

The GC Mutator Scheme language is used to test garbage collectors written with the GC Collector Scheme language. Since collectors support a subset of Scheme’s values, the GC Mutator Scheme language supports a subset of procedures and syntax. In addition, many procedures that can be written in the mutator are omitted as they make good test cases. Therefore, the mutator language provides only primitive procedures, such as +, cons, etc.

3.1 Building Mutators

The first expression of a mutator must be:

(allocator-setup collector-module
                 heap-size)
 
heap-size = exact-nonnegative-integer?
collector-module specifies the path to the garbage collector that the mutator should use. The collector must be written in the GC Collector Scheme language.

The rest of a mutator module is a sequence of definitions, expressions and test cases. The GC Mutator Scheme language transforms these definitions and statements to use the collector specified in allocator-setup. In particular, many of the primitive forms, such as cons map directly to procedures such as gc:cons, written in the collector.

3.2 Mutator API

The GC Mutator Scheme language supports the following syntactic forms:

  if and or cond case define define-values let let-values let* set! lambda λ quote error begin

The language also defines the following procedures:

  add1 sub1 zero? + - * / even? odd? = < > <= >= cons first rest
  set-first! set-rest! cons? symbol? number? boolean? empty? eq?

(set-first! c v)  void
  c : cons?
  v : any/c
Sets the first of the cons cell c.

(set-rest! c v)  void
  c : cons?
  v : any/c
Sets the rest of the cons cell c.

The identifier empty is defined to invoke (gc:alloc-flat empty) wherever it is used.

Other common procedures are left undefined as they can be defined in terms of the primitives and may be used to test collectors.

Additional procedures from scheme may be imported with:

Imports the procedures id ... from scheme. Each procedure is transformed to correctly interface with the mutator. That is, its arguments are dereferenced from the mutator’s heap and the result is allocated on the mutator’s heap. The arguments and result must be heap-value?s, even if the imported procedure accepts or produces structured data.

For example, the GC Mutator Scheme language does not define modulo:

  (import-primitives modulo)
  
  (test/value=? (modulo 5 3) 2)

3.3 Testing Mutators

GC Mutator Scheme provides two forms for testing mutators:

(test/location=? mutator-expr1 mutator-expr2)
test/location=? succeeds if mutator-expr1 and mutator-expr2 reference the same location on the heap.

(test/value=? mutator-expr scheme-datum/quoted)
test/value=? succeeds if mutator-expr and scheme-datum/expr are structurally equal. scheme-datum/quoted is not allocated on the mutator’s heap. Futhermore, it must either be a quoted value or a literal value.

(printf format mutator-expr ...)
 
format = literal-string
In GC Mutator Scheme, printf is a syntactic form and not a procedure. The format string, format is not allocated on the mutator’s heap.

4 Web Application Scheme

#lang planet plai/plai:1:18/web

The Web Application Scheme language allows you to write server-side Web applications for the PLT Web Server.

For more information about writing Web applications, see: Web: PLT Web Applications.

A Web application must define a procedure start:

(start initial-request)  response?
  initial-request : request?
The initial request to a Web application is serviced by this procedure.

When you click on the Run button in DrScheme, your Web application is launched in the Web server.

The application is available at http://localhost:8000/servlets/standalone.ss.

The Web Application Scheme language will automatically load this URL in your Web browser.

You may use no-web-browser to prevent the browser from being launched and static-files-path to serve additional static files.

5 Generating Random Mutators

 (require (planet plai/plai:1:18/random-mutator))

This PLAI library provides a facility for generating random mutators, in order to test your garbage collection implementation.

(save-random-mutator file    
  collector-name    
  [#:heap-values heap-values    
  #:iterations iterations    
  #:program-size program-size    
  #:heap-size heap-size]    
  #:plai-version string?)  void?
  file : path-string?
  collector-name : string?
  heap-values : (cons heap-value? (listof heap-value?))
   = (list 0 1 -1 'x 'y #f #t '())
  iterations : exact-positive-integer? = 200
  program-size : exact-positive-integer? = 10
  heap-size : exact-positive-integer? = 100
  string? : "1:18"
Creates a random mutator that uses the collector collector-name and saves it in file.

The mutator is created by first making a random graph whose nodes either have no outgoing edges, two outgoing edges, or some random number of outgoing edges and then picking a random path in the graph that ends at one of the nodes with no edges.

This graph and path are then turned into a PLAI program by creating a let expression that binds one variable per node in the graph. If the node has no outgoing edges, it is bound to a heap-value?. If the node has two outgoing edges, it is bound to a pair and the two edges are put into the first and rest fields. Otherwise, the node is represented as a procedure that accepts an integer index and returns the destination node of the corresponding edge.

Once the let expression has been created, the program creates a bunch of garbage and then traverses the graph, according to the randomly created path. If the result of the path is the expected heap value, the program does this again, up to iterations times. If the result of the path is not the expected heap value, the program terminates with an error.

The keyword arguments control some aspects of the generation of random mutators:
  • Elements from the heap-values argument are used as the base values when creating nodes with no outgoing edges. See also find-heap-values.

  • The iterations argument controls how many times the graph is created (and traversed).

  • The program-size argument is a bound on how big the program it is; it limits the number of nodes, the maximum number of edges, and the length of the path in the graph.

  • The heap-size argument controls the size of the heap in the generated mutator.

  • The plai-version string determines which version of PLAI is put into the #lang line for the generated mutator.

(find-heap-values input)  (listof heap-value?)
  input : (or/c path-string? input-port?)
Processes input looking for occurrences of heap-value?s in the source of the program and returns them. This makes a good start for the heap-values argument to save-random-mutator.

If input is a port, its contents are assumed to be a well-formed PLAI program. If input is a file, the contents of the file are used.