第120页 | The Art of UNIX Programming | 阅读 ‧ 电子书库

Part營I.燚esign

Prev    Next

Design

Table of Contents

4. Modularity

Encapsulation and Optimal Module SizeCompactness and OrthogonalityCompactnessOrthogonalityThe SPOT RuleCompactness and the Strong Single CenterThe Value of DetachmentSoftware Is a Many-Layered ThingTop-Down versus Bottom-UpGlue LayersCase Study: C Considered as Thin GlueLibrariesCase Study: GIMP PluginsUnix and Object-Oriented LanguagesCoding for Modularity

5. Textuality

The Importance of Being TextualCase Study: Unix Password File FormatCase Study: .newsrc FormatCase Study: The PNG Graphics File FormatData File MetaformatsDSV StyleRFC 822 FormatCookie-Jar FormatRecord-Jar FormatXMLWindows INI FormatUnix Textual File Format ConventionsThe Pros and Cons of File CompressionApplication Protocol DesignCase Study: SMTP, the Simple Mail Transfer ProtocolCase Study: POP3, the Post Office ProtocolCase Study: IMAP, the Internet Message Access ProtocolApplication Protocol MetaformatsThe Classical Internet Application MetaprotocolHTTP as a Universal Application ProtocolBEEP: Blocks Extensible Exchange ProtocolXML-RPC, SOAP, and Jabber

6. Transparency

Studying CasesCase Study: audacityCase Study: fetchmail's -v optionCase Study: GCCCase Study: kmailCase Study: SNGCase Study: The Terminfo DatabaseCase Study: Freeciv Data FilesDesigning for Transparency and DiscoverabilityThe Zen of TransparencyCoding for Transparency and DiscoverabilityTransparency and Avoiding OverprotectivenessTransparency and Editable RepresentationsTransparency, Fault Diagnosis, and Fault RecoveryDesigning for Maintainability

7. Multiprogramming

Separating Complexity Control from Performance TuningTaxonomy of Unix IPC MethodsHanding off Tasks to Specialist ProgramsPipes, Redirection, and FiltersWrappersSecurity Wrappers and Bernstein ChainingSlave ProcessesPeer-to-Peer Inter-Process CommunicationProblems and Methods to AvoidObsolescent Unix IPC MethodsRemote Procedure CallsThreads — Threat or Menace?Process Partitioning at the Design Level

8. Minilanguages

Understanding the Taxonomy of LanguagesApplying MinilanguagesCase Study: sngCase Study: Regular ExpressionsCase Study: GladeCase Study: m4Case Study: XSLTCase Study: The Documenter's Workbench ToolsCase Study: fetchmail Run-Control SyntaxCase Study: awkCase Study: PostScriptCase Study: bc and dcCase Study: Emacs LispCase Study: JavaScriptDesigning MinilanguagesChoosing the Right Complexity LevelExtending and Embedding LanguagesWriting a Custom GrammarMacros — Beware!Language or Application Protocol?

9. Generation

Data-Driven ProgrammingCase Study: asciiCase Study: Statistical Spam FilteringCase Study: Metaclass Hacking in fetchmailconfAd-hoc Code GenerationCase Study: Generating Code for the ascii DisplaysCase Study: Generating HTML Code for a Tabular List

10. Configuration

What Should Be Configurable?Where Configurations LiveRun-Control FilesCase Study: The .netrc FilePortability to Other Operating SystemsEnvironment VariablesSystem Environment VariablesUser Environment VariablesWhen to Use Environment VariablesPortability to Other Operating SystemsCommand-Line OptionsThe -a to -z of Command-Line OptionsPortability to Other Operating SystemsHow to Choose among the MethodsCase Study: fetchmailCase Study: The XFree86 ServerOn Breaking These Rules

11. Interfaces

Applying the Rule of Least SurpriseHistory of Interface Design on UnixEvaluating Interface DesignsTradeoffs between CLI and Visual InterfacesCase Study: Two Ways to Write a Calculator ProgramTransparency, Expressiveness, and ConfigurabilityUnix Interface Design PatternsThe Filter PatternThe Cantrip PatternThe Source PatternThe Sink PatternThe Compiler PatternThe ed patternThe Roguelike PatternThe ‘Separated Engine and Interface’ PatternThe CLI Server PatternLanguage-Based Interface PatternsApplying Unix Interface-Design PatternsThe Polyvalent-Program PatternThe Web Browser as a Universal Front EndSilence Is Golden

12. Optimization

Don't Just Do Something, Stand There!Measure before OptimizingNonlocality Considered HarmfulThroughput vs. LatencyBatching OperationsOverlapping OperationsCaching Operation Results

13. Complexity

Speaking of ComplexityThe Three Sources of ComplexityTradeoffs between Interface and Implementation ComplexityEssential, Optional, and Accidental ComplexityMapping ComplexityWhen Simplicity Is Not EnoughA Tale of Five EditorsedviSamEmacsWilyThe Right Size for an EditorIdentifying the Complexity ProblemsCompromise Doesn't WorkIs Emacs an Argument against the Unix Tradition?The Right Size of Software

 

 

 

Prev  Up Next

What Goes Around, Comes Around  Home 燙hapter�Modularity