DOCSLIB.ORG

Biglook: a Widget Library for the Scheme Programming Language

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Biglook: a Widget Library for the Scheme Programming Language Biglook: a Widget Library for the Scheme Programming Language Erick Gallesio Manuel Serrano Universite´ de Nice – Sophia Antipolis Inria Sophia Antipolis 950 route des Colles, B.P. 145 2004 route des Lucioles – B.P. 93 F-06903 Sophia Antipolis, Cedex F-06902 Sophia-Antipolis, Cedex ¢¡¤£¦¥¨§ © ¢ ¦¦£¦£¦¥©¡ ¤ ¦¢© ¦¡¡ ¦£!¢¡£¦ ©¡ Abstract Biglook is an Object Oriented Scheme library for constructing GUIs. It uses classes of a CLOS- like object layer to represent widgets and Scheme closures to handle events. Combining functional and object-oriented programming styles yields an original application programming interface that advocates a strict separation between the implementation of the graphical interfaces and the user-associated commands, enabling compact source code. The Biglook implementation separates the Scheme programming interface and the na- Figure 1: Two Biglook applications: a code tive back-end. This permits different ports browser on the left, “dock” applications on the for Biglook. The current version uses GTK " right and Swing graphical toolkits, while the pre- vious release used Tk. It is available at: #%$&$&')(+*,*-/.,0%1324)5 6,472&8¤9:5<;=3*>72?/10&0- . shots of Biglook applications: i) -&B,=/0¤CD¤9 on the left and ii) the Biglook dock applications, a` la NextStep, on the right. These Biglook applica- tions are used on a daily basis. Introduction By contrast to previous work, no attempt has been made to make that library familiar to pro- We have studied the problem of construct- grammers used to imperative or purely object ori- ing GUI in functional languages by designing a ented programming style. On the contrary, our li- widget library for the Scheme programming lan- brary introduces an original application program- guage, called Biglook. In this paper we focus on ming interface (API) that benefits from the high how to apply the functional style to GUIs pro- level constructions of the extended Scheme im- gramming. plementation named Bigloo [25] which is open source and freely available since 1992. The main Biglook’s primary use is to implement graph- Bigloo component is an optimizing compiler that 1&0&.&A ical applications (e.g., @ , editors a` la delivers small and efficient applications for the Emacs, browser a` la Netscape, programming UnixTM operating system. Bigloo is able to pro- tools such as -&B%=%0¤CD¤9 , our graphical Scheme duce native code (via C) and JVM bytecode. Cur- code browser). Figure 1 presents two screen rently Biglook uses GTK " [20] associated with the Bigloo C back-end and Swing [29] with the 1 Bigloo Bigloo JVM back-end. The previous release of Biglook [12] used Tk [19]. Bigloo is an open implementation of the Bigloo implements an object layer inspired Scheme programming language. From the be- by CLOS [1]. It is a class-based model where ginning, the aim was to propose a realistic and methods override generic functions and are not pragmatic alternative to the strict Scheme [17]. declared inside classes as in Smalltalk [13], Bigloo does not implement “all” of Scheme; for O’Caml [22] or Java [14]. example, the execution of tail-recursion may al- locate memory. On the other hand, Bigloo imple- Biglook is implemented as a wrapping layer ments numerous extensions: support for lexical on top of native widget libraries (that we name and syntactic analysis, pattern matching, an ex- henceforth the back-end). This software architec- ception mechanism, a foreign interface, an object ture saves the effort of implementing low-level layer and a module language. In this Section, we constructions (pixel switching, clipping, event present Bigloo’s modules and its object model; handling and so on) allowing to focus on the that is, class declarations and generic functions. Scheme implementation of new features. Virtual slots, which are heavily used in Biglook, are presented in Section 2. When designing the Biglook API, we always had to decide which model to choose: the func- tional model or the object model. We think that 1.1 Modules these two models are not contradictory but com- plementary. For instance, if the widget hierarchy naturally fits a class hierarchy, user call-backs are Bigloo modules have two basic purposes: one naturally implemented by the means of Scheme is to allow separate compilation and the second is closures. to increase the number of errors that can be de- tected by the compiler. Bigloo modules are sim- In Section 1 we briefly present the Bigloo sys- ple and they have been designed with the concern tem emphasizing its module system and its object of an easy implementation. layer. This section is required for readers unfa- miliar with the CLOS model and it also serves as A module is a compilation unit for Bigloo. It an introduction to virtual slots. Virtual slots are is represented by one or more files and has the a new Bigloo construction that is required in or- following syntax: der to separate the Biglook API made of classes and the native back-end. They are presented in ¢¡¤£¦¥¨§ ©¨ module-name ¡ £¦ Section 2. In Section 3 we present the Biglook import ¦ ¨¨ £¦ library. We start showing a simple Biglook ap- export ¦ ¨ ¦ static plication and its associated source code. Then, © !¨" static #¦ we detail the Biglook programming principles (widgets creation, event handling, etc.) motivat- opt-body ing our design orientations by programming lan- guage considerations. In this section we are con- Import clauses are used to import bindings in the ducted to compare the functional programming module. In order to import, one just needs to state style and the object oriented one. In Section 4 we the identifier to be imported and its source mod- present the Biglook implementation. At last, in ule. Note that a shorthand exists to import all the Section 5 we present a comparison with related bindings of a module. work. Export and static clauses play a close role. They point out to the compiler that the module implements some bindings and distinguish those that can be used within other modules (they are exported) and those that cannot (they are static). These clauses do not contain identifiers but proto- types. It is then possible to export variables (mu- ¢¡¤£¦¥¨§ ©¨ ¡¤£¦¥¨§ ©¨ £ ¨©¨ £ table bindings) or functions (read-only bindings). ¨¨ £¦ £ ¤ Static clauses are optional (the bindings of a mod- ¥ £§! ©¨ ¥ ¦§ ©¦ ¡¢¡ ule which are not referenced in a clause are, by " ¥ £§! ©¨ ¥ ¦§ ©¦ ¡¢¡ default, static). ¨©¨ £ ¨¥ £ ¡¢¡ ! ¥ £§! ©¨ ¥ ¦§ ©¦ ¡¢¡ Library clauses enable programs to use Bigloo £ ¤ libraries. A Bigloo library is merely a collec- ¥ ©¨ ¨ ¨£§¢ tion of pre-compiled modules. Using a library is ©¨ ¡ !¥ £!" £ ¡¢¡ ¤# ¨£§¢ !$&% ¨£§¢ equivalent to importing all the modules compos- ¦ ¨ (') £* £ ' ¨£§¢ ing the library. Detailed information on Bigloo ¡ modules may be found in the two papers [25, 26]. 1.2.2 Instances 1.2 Object layer When declaring a class +-,. , Bigloo automat- / ically generates the predicate 81&. , an allo- 2¦4D¢$/.4%$2.$%9 (,( cator +-,. , an instance cloner 1/2&8¤.$/9:(,( 81&.1 In this paper we assume a CLOS-like object A6,' @ +-,0. , accessors (e.g., for %D¤9$32 model with single inheritance and single dis- 8¤1&. @ a slot @ ), modifiers (e.g., for a patch. The object layer implemented in Bigloo slot @ ) and an abbreviated special access form, ¤&.38,8¤9/D,D (,( is a restricted version of CLOS [1] inspired, to a C2$&# +4,0. , to allow accessing and great extent, by MEROON [21]. writing slots simply by using their name. Here is how to allocate and access an instance of ©657¦8-¥9 9:¤ : ©¨ ¨ * ¨ £ ¨¥ ¡¢¡ "<; =%9 1.2.1 Class declarations ;; The initialization value of a slot can be omitted ;; from the arguments list if it has a default value; ;; this is the case for the slot of class > ?@BADCFEHG¨I . KJ¤ L0¨ ¦ £ ¨¥ "M Classes can be declared static or exported. It ¡¢¡ FN !$ FN FN " FNO is then possible to make a declaration accessible from another module or to limit its scope to one C2¦$&#P&.38&8¤93D,D module. The abbreviated syntax of a class decla- 2¦4D¢$%.¤4%$2.¤$/9 ration is: The and special forms are implemented by the means of macros ¨©¨ ¡¢¡ that statically resolve the keyword parameters class-id super-class-id opt-init opt-slots Q R (such as @ , and ). For instance, the above ex- ample is expanded into: ¢¡*S £ ¨¥T%9O<;O A class can inherit from a single super class. ©¨ ¨ FN !$ FN £ ¨¥¦ Classes with no specified super class inherit from 0¤B¤£,9/8¢$ FN £ ¨¥¦" the class. The type associated with a sub- FN £ ¨¥* class is a subtype of the type of the super class. A class may be provided with an initialization As we can see, since macros are expanded at function (opt-init) that is automatically called compile-time, there is no run-time penalty asso- each time an instance of class-id is created. Ini- ciated with keyword parameters. tialization functions accept one argument, the created instance. 1.2.3 Generic functions and methods A slot may be typed (with the annotation ¦¥¨§¨© (,( ) and may have a default value ( A%9&;%.¤631¤$ option). Here are some possible decla- Generic function declarations are function decla- 2¦4%$ '30 24%$¤&A ?/9¤439=72&8 rations for the traditional '/0 and : rations annotated by the keyword. They ¥ ¡ ¨L £¦¥ BL £FJ £ £ ¡¢¡ KJ¤ L0¨ ¦ £ £ " can be exported, which means that they can be ¡¢¡ '©¨D' used from other modules and that methods can '"©¨D' " be added to those functions from other mod- ules. They can also be static, that is, not acces- ¥ ¡ ¨L £¦¥ BL £FJ £ £ ¨¥ ¡¢¡ KJ¤ L0¨ ¦ £ ¨¥ £ ! sible from within other modules, hich means that ¡¢¡ ¢F ¨¡ ¨L £¦¥ no other modules can add methods. The CLOS ¦©© ' ©¨D' model fits harmoniously with the traditional func- tional programming style because a function can be thought as a generic function overridden with Hereafter is an example of a call to the D#30¤C exactly one method.
Load more

Recommended publications
  • An Optimized R5RS Macro Expander
    An Optimized R5RS Macro Expander Sean Reque A thesis submitted to the faculty of Brigham Young University in partial fulfillment of the requirements for the degree of Master of Science Jay McCarthy, Chair Eric Mercer Quinn Snell Department of Computer Science Brigham Young University February 2013 Copyright c 2013 Sean Reque All Rights Reserved ABSTRACT An Optimized R5RS Macro Expander Sean Reque Department of Computer Science, BYU Master of Science Macro systems allow programmers abstractions over the syntax of a programming language. This gives the programmer some of the same power posessed by a programming language designer, namely, the ability to extend the programming language to fit the needs of the programmer. The value of such systems has been demonstrated by their continued adoption in more languages and platforms. However, several barriers to widespread adoption of macro systems still exist. The language Racket [6] defines a small core of primitive language constructs, including a powerful macro system, upon which all other features are built. Because of this design, many features of other programming languages can be implemented through libraries, keeping the core language simple without sacrificing power or flexibility. However, slow macro expansion remains a lingering problem in the language's primary implementation, and in fact macro expansion currently dominates compile times for Racket modules and programs. Besides the typical problems associated with slow compile times, such as slower testing feedback, increased mental disruption during the programming process, and unscalable build times for large projects, slow macro expansion carries its own unique problems, such as poorer performance for IDEs and other software analysis tools.
    [Show full text]
  • The Evolution of Lisp
    1 The Evolution of Lisp Guy L. Steele Jr. Richard P. Gabriel Thinking Machines Corporation Lucid, Inc. 245 First Street 707 Laurel Street Cambridge, Massachusetts 02142 Menlo Park, California 94025 Phone: (617) 234-2860 Phone: (415) 329-8400 FAX: (617) 243-4444 FAX: (415) 329-8480 E-mail: [email protected] E-mail: [email protected] Abstract Lisp is the world’s greatest programming language—or so its proponents think. The structure of Lisp makes it easy to extend the language or even to implement entirely new dialects without starting from scratch. Overall, the evolution of Lisp has been guided more by institutional rivalry, one-upsmanship, and the glee born of technical cleverness that is characteristic of the “hacker culture” than by sober assessments of technical requirements. Nevertheless this process has eventually produced both an industrial- strength programming language, messy but powerful, and a technically pure dialect, small but powerful, that is suitable for use by programming-language theoreticians. We pick up where McCarthy’s paper in the first HOPL conference left off. We trace the development chronologically from the era of the PDP-6, through the heyday of Interlisp and MacLisp, past the ascension and decline of special purpose Lisp machines, to the present era of standardization activities. We then examine the technical evolution of a few representative language features, including both some notable successes and some notable failures, that illuminate design issues that distinguish Lisp from other programming languages. We also discuss the use of Lisp as a laboratory for designing other programming languages. We conclude with some reflections on the forces that have driven the evolution of Lisp.
    [Show full text]
  • Structured Foreign Types
    Ftypes: Structured foreign types Andrew W. Keep R. Kent Dybvig Indiana University fakeep,[email protected] Abstract When accessing scalar elements within an ftype, the value of the High-level programming languages, like Scheme, typically repre- scalar is automatically marshaled into the equivalent Scheme rep- sent data in ways that differ from the host platform to support resentation. When setting scalar elements, the Scheme value is consistent behavior across platforms and automatic storage man- checked to ensure compatibility with the specified foreign type, and agement, i.e., garbage collection. While crucial to the program- then marshaled into the equivalent foreign representation. Ftypes ming model, differences in data representation can complicate in- are well integrated into the system, with compiler support for ef- teraction with foreign programs and libraries that employ machine- ficient access to foreign data and debugger support for convenient dependent data structures that do not readily support garbage col- inspection of foreign data. lection. To bridge this gap, many high-level languages feature for- The ftype syntax is convenient and flexible. While it is similar in eign function interfaces that include some ability to interact with some ways to foreign data declarations in other Scheme implemen- foreign data, though they often do not provide complete control tations, and language design is largely a matter of taste, we believe over the structure of foreign data, are not always fully integrated our syntax is cleaner and more intuitive than most. Our system also into the language and run-time system, and are often not as effi- has a more complete set of features, covering all C data types along cient as possible.
    [Show full text]
  • Chez Scheme Version 5 Release Notes Copyright C 1994 Cadence Research Systems All Rights Reserved October 1994 Overview 1. Funct
    Chez Scheme Version 5 Release Notes Copyright c 1994 Cadence Research Systems All Rights Reserved October 1994 Overview This document outlines the changes made to Chez Scheme for Version 5 since Version 4. New items include support for syntax-case and revised report high-level lexical macros, support for multiple values, support for guardians and weak pairs, support for generic ports, improved trace output, support for shared incremental heaps, support for passing single floats to and returning single floats from foreign procedures, support for passing structures to and returning structures from foreign procedures on MIPS-based computers, and various performance enhancements, including a dramatic improvement in the time it takes to load Scheme source and object files on DEC Ultrix MIPS-based computers. Overall performance of generated code has increased by an average of 15–50 percent over Version 4 depending upon optimize level and programming style. Programs that make extensive use of locally-defined procedures will benefit more than programs that rely heavily on globally-defined procedures, and programs compiled at higher optimize levels will improve more than programs compiled at lower optimize levels. Compile time is approximately 25 percent faster. Refer to the Chez Scheme System Manual, Revision 2.5 for detailed descriptions of the new or modified procedures and syntactic forms mentioned in these notes. Version 5 is available for the following platforms: • Sun Sparc SunOS 4.X & Solaris 2.X • DEC Alpha AXP OSF/1 V2.X • Silicon Graphics IRIX 5.X • Intel 80x86 NeXT Mach 3.2 • Motorola Delta MC88000 SVR3 & SVR4 • Decstation/Decsystem Ultrix 4.3A • Intel 80x86 BSDI BSD/386 1.1 • Intel 80x86 Linux This document contains four sections describing (1) functionality enhancements, (2) performance enhance- ments, (3) bugs fixed, and (4) compatibility issues.
    [Show full text]
  • Drscheme: a Pedagogic Programming Environment for Scheme
    DrScheme A Pedagogic Programming Environment for Scheme Rob ert Bruce Findler Cormac Flanagan Matthew Flatt Shriram Krishnamurthi and Matthias Felleisen Department of Computer Science Rice University Houston Texas Abstract Teaching intro ductory computing courses with Scheme el evates the intellectual level of the course and thus makes the sub ject more app ealing to students with scientic interests Unfortunatelythe p o or quality of the available programming environments negates many of the p edagogic advantages Toovercome this problem wehavedevel op ed DrScheme a comprehensive programming environmentforScheme It fully integrates a graphicsenriched editor a multilingua l parser that can pro cess a hierarchyofsyntactically restrictivevariants of Scheme a functional readevalprint lo op and an algebraical ly sensible printer The environment catches the typical syntactic mistakes of b eginners and pinp oints the exact source lo cation of runtime exceptions DrScheme also provides an algebraic stepp er a syntax checker and a static debugger The rst reduces Scheme programs including programs with assignment and control eects to values and eects The to ol is useful for explainin g the semantics of linguistic facilities and for studying the b ehavior of small programs The syntax c hecker annotates programs with fontandcolorchanges based on the syntactic structure of the pro gram It also draws arrows on demand that p oint from b ound to binding o ccurrences of identiers The static debugger roughly sp eaking pro vides a typ e inference system
    [Show full text]
  • Supplementary Material Forrebuilding Racket on Chez Scheme
    Supplementary Material for Rebuilding Racket on Chez Scheme (Experience Report) MATTHEW FLATT, University of Utah, USA CANER DERICI, Indiana University, USA R. KENT DYBVIG, Cisco Systems, Inc., USA ANDREW W. KEEP, Cisco Systems, Inc., USA GUSTAVO E. MASSACCESI, Universidad de Buenos Aires, Argentina SARAH SPALL, Indiana University, USA SAM TOBIN-HOCHSTADT, Indiana University, USA JON ZEPPIERI, independent researcher, USA All benchmark measurements were performed on an Intel Core i7-2600 3.4GHz processor running 64-bit Linux. Except as specifically noted, we used Chez Scheme 9.5.3 commit 7df2fb2e77 at github:cicso/ChezScheme, modified as commit 6d05b70e86 at github:racket/ChezScheme, and Racket 7.3.0.3 as commit ff95f1860a at github:racket/racket. 1 TRADITIONAL SCHEME BENCHMARKS The traditional Scheme benchmarks in figure1 are based on a suite of small programs that have been widely used to compare Scheme implementations. The benchmark sources are in the "common" directory of the racket-benchmarks package in the Racket GitHub repository. The results are in two groups, where the group starting with scheme-c uses mutable pairs, so they are run in Racket as #lang r5rs programs; for Racket CS we expect overhead due to the use of a record datatype for mutable pairs, instead of Chez Scheme’s built-in pairs. The groups are sorted by the ratio of times for Chez Scheme and the current Racket imple- mentation. Note that the break-even point is near the end of the first group. The racket collatz benchmark turns out to mostly measure the performance of the built-in division operator for rational numbers, while fft and nucleic benefit from flonum unboxing.
    [Show full text]
  • 23 Things I Know About Modules for Scheme
    23 things I know about modules for Scheme Christian Queinnec Université Paris 6 — Pierre et Marie Curie LIP6, 4 place Jussieu, 75252 Paris Cedex — France [email protected] ABSTRACT difficult to deliver (or even rebuild) stand-alone systems. The benefits of modularization are well known. However, Interfaces as collection of names — If modules are about shar- modules are not standard in Scheme. This paper accompanies ing, what should be shared ? Values, locations (that is an invited talk at the Scheme Workshop 2002 on the current variables), types, classes (and their cortege` of accessors, state of modules for Scheme. Implementation is not addressed, constructors and predicates) ? only linguistic features are covered. Cave lector, this paper only reflects my own and instanta- The usual answer in Scheme is to share locations with neous biases! (quite often) two additional properties: (i) these loca- tions can only be mutated from the body of their defin- ing modules (this favors block compilation), (ii) they 1. MODULES should hold functions (and this should be statically (and The benefits of modularization within conventional languages easily) discoverable). This restricts linking with other are well known. Modules dissociate interfaces and implemen- (foreign) languages that may export locations holding tations; they allow separate compilation (or at least indepen- non-functional data (the errno location for instance). dent compilation a` la C). Modules tend to favor re-usability, This is not a big restriction since modern interfaces (Corba common libraries and cross language linkage. for example) tend to exclusively use functions (or meth- Modules discipline name spaces with explicit names expo- ods).
    [Show full text]
  • Programming Languages As Operating Systems
    Programming Languages as Op erating Systems (or Revenge of the Son of the Lisp Machine) Matthew Flatt Rob ert Bruce Findler Shriram Krishnamurthi Matthias Felleisen Department of Computer Science Rice University Houston, Texas 77005-1892 reclaim the program's resources|even though the program Abstract and DrScheme share a single virtual machine. The MrEd virtual machine serves b oth as the implementa- To address this problem, MrEd provides a small set of tion platform for the DrScheme programming environment, new language constructs. These constructs allow a program- and as the underlying Scheme engine for executing expres- running program, suchasDrScheme, to run nested programs sions and programs entered into DrScheme's read-eval-print directly on the MrEd virtual machine without sacri cing lo op. We describ e the key elements of the MrEd virtual control over the nested programs. As a result, DrScheme machine for building a programming environment, and we can execute a copyofDrScheme that is executing its own step through the implementation of a miniature version of copy of DrScheme (see Figure 1). The inner and middle DrScheme in MrEd. More generally,weshowhow MrEd de- DrSchemes cannot interfere with the op eration of the outer nes a high-level op erating system for graphical programs. DrScheme, and the middle DrScheme cannot interfere with the outer DrScheme's control over the inner DrScheme. In this pap er, we describ e the key elements of the MrEd 1 MrEd: A Scheme Machine virtual machine, and we step through the implementation of a miniature version of DrScheme in MrEd. More gen- The DrScheme programming environment [10] provides stu- erally,weshowhow MrEd de nes a high-level op erating dents and programmers with a user-friendly environment system (OS) for graphical programs.
    [Show full text]
  • Dissertation
    HOWTOGENERATEACTIONABLE ADVICEABOUTPERFORMANCE PROBLEMS Vincent St-Amour Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy College of Computer and Information Science Northeastern University Boston, Massachusetts April 2015 Vincent St-Amour: How to Generate Actionable Advice about Performance Problems, Doctor of Philosophy, Northeastern University, Boston, Massachusetts © April 2015 ABSTRACT Performance engineering is an important activity regardless of ap- plication domain, as critical for server software as for mobile appli- cations. This activity, however, demands advanced, specialized skills that require a significant time investment to acquire, and are therefore absent from most programmers’ toolboxes. My thesis is that tool support can make performance engineering both accessible and time-efficient for non-expert programmers. To support this claim, this dissertation introduces two novel families of performance tools that are designed specifically to provide actionable information to programmers: optimization coaches and feature-specific profilers. This dissertation presents blueprints for building tools in these families, and provides examples from tools that I have built. v ACKNOWLEDGMENTS I would like to thank first and foremost my advisor Matthias Felleisen, who has taught me most of what I know today about the craft of research, writing, and design. He is the best advisor and ally I could have wished for, and I am incredibly grateful for of all the help and guidance he has provided me over the years. This work would not have been possible without the help of my collaborators, with whom I have worked on the projects described in these pages and others along the way: Leif Andersen, Eric Dob- son, Matthias Felleisen, Robby Findler, Matthew Flatt, Shu-yu Guo, Asumu Takikawa, Sam Tobin-Hochstadt, and Neil Toronto.
    [Show full text]
  • Bugloo: a Source Level Debugger for Scheme Programs Compiled Into JVM Bytecode
    Bugloo: A Source Level Debugger for Scheme Programs Compiled into JVM Bytecode Damien Ciabrini Manuel Serrano INRIA Sophia Antipolis INRIA Sophia Antipolis 2004 route des Lucioles - BP 93 2004 route des Lucioles - BP 93 F-06902 Sophia Antipolis, Cedex F-06902 Sophia Antipolis, Cedex [email protected] [email protected] ABSTRACT strated by a static debugger. Our work focuses on the dy- namic debugging approach and on the Scheme [20] program- This paper presents Bugloo, a source level debugger for ming language. Scheme programs. It enables debugging of programs com- piled into JVM bytecode by the Bigloo compiler. It aims 1.1 Motivation of Our Work at being easy to use because it provides a user interface. It In practice, programmers hardly use debuggers. Obviously, aims at being practical to use because it is easy to deploy they prefer ad hoc debugging techniques such as inserting and because it is efficient enough to debug large programs prints in the source code. We think that the reason of such as the Bigloo compiler itself. this disinterest is twofold: (i) many debuggers are not ef- ficient enough to debug large programs, (ii), features of- The JVM platform provides two standard APIs for imple- fered by many debuggers are unsufficient for the type of bug menting debuggers and profilers: JVMDI and JVMPI. One programmers have to correct. In consequence, they often of the motivations for providing the Bigloo compiler with a conclude that “prints are simpler and quicker”. In other JVM back-end was to benefit from these two APIs to imple- words, the overall benefits of debuggers might not be worth ment a debugger for Scheme.
    [Show full text]
  • Proceedings of the FREENIX Track: 2002 USENIX Annual Technical Conference
    USENIX Association Proceedings of the FREENIX Track: 2002 USENIX Annual Technical Conference Monterey, California, USA June 10-15, 2002 THE ADVANCED COMPUTING SYSTEMS ASSOCIATION © 2002 by The USENIX Association All Rights Reserved For more information about the USENIX Association: Phone: 1 510 528 8649 FAX: 1 510 548 5738 Email: [email protected] WWW: http://www.usenix.org Rights to individual papers remain with the author or the author's employer. Permission is granted for noncommercial reproduction of the work for educational or research purposes. This copyright notice must be included in the reproduced paper. USENIX acknowledges all trademarks herein. Biglook: a Widget Library for the Scheme Programming Language Erick Gallesio Manuel Serrano Universite´ de Nice – Sophia Antipolis Inria Sophia Antipolis 950 route des Colles, B.P. 145 2004 route des Lucioles – B.P. 93 F-06903 Sophia Antipolis, Cedex F-06902 Sophia-Antipolis, Cedex EÖickºGaÐÐe×iÓ@ÙÒiceºfÖ ÅaÒÙeкËeÖÖaÒÓ@iÒÖiaºfÖ Abstract Biglook is an Object Oriented Scheme library for constructing GUIs. It uses classes of a CLOS- like object layer to represent widgets and Scheme closures to handle events. Combining functional and object-oriented programming styles yields an original application programming interface that advocates a strict separation between the implementation of the graphical interfaces and the user-associated commands, enabling compact source code. The Biglook implementation separates the Scheme programming interface and the na- Figure 1: Two Biglook applications: a code tive back-end. This permits different ports browser on the left, “dock” applications on the for Biglook. The current version uses GTK· right and Swing graphical toolkits, while the pre- vious release used Tk.
    [Show full text]
  • Gene Finding
    Programming Languages Rhys Price Jones Class home page is linked from: http://www.seas.gwu.edu/~rhyspj Friday, December 31, 2010 CSCI 3221 Programming Languages Rhys Price Jones Class home page is linked from: http://www.seas.gwu.edu/~rhyspj Friday, December 31, 2010 Textbooks: Little Schemer absolutely essential to read and write (!) this book this week SICP Structure and Interpretation of Computer Programs a timeless classic buy the paper book, or use the online free version 2 Friday, December 31, 2010 Software - free downloads: Petite chez scheme http://www.scheme.com emacs www.gnu.org/software/emacs You can plug in stuff to Eclipse so it supports scheme but my class presentations and materials will be emacs based 3 Friday, December 31, 2010 Platform specifics Windows nice to have cygwin www.cygwin.com Mac all set to go BUT do not use the Emacs that comes with mac os. Linux ... 4 Friday, December 31, 2010 Exams etc One midterm One final more or less weekly quizzes reading assignments lab assignments 5 Friday, December 31, 2010 Overview of course Learn Scheme Use Scheme Ignore Scheme and concentrate on principles of computation Implement Scheme Modify Scheme Implement other languages and experiment with language features 6 Friday, December 31, 2010 Reading assignment linked from class home page This first week. Little Schemer Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 7 Friday, December 31, 2010 This week’s lab play with scheme What is Scheme? a dialect of Lisp What is Lisp? ... 8
    [Show full text]