Introduction to Java 7 (OSCON 2012)
Download as KEY, PDF2 likes910 views
This document provides an overview and agenda for a session on the "Well-Grounded Java Developer". The session will cover introductions to Java 7 features like Project Coin, NIO.2, and Method Handles. It will also cover polyglot and functional programming and modern Java concurrency practices like Fork/Join. The session aims to help developers code more rapidly, concisely, and take advantage of compiler, JVM, and GC improvements in Java 7.
Introduction to Java 7 (OSCON 2012)
- 1. The Well-Grounded Java Developer Introduction to Java 7
- 2. This is not an Oracle legal slide 2
- 3. No, we’re not in sales! • Get in touch - Ben is: @kittylist - Martijn is: @karianna (and @diabolicaldev) • What we do - Authors of “The Well-Grounded Java Developer” - CEO/CTO of jClarity - JVM/Hardware/Performance stuff • We co-lead the LJC (London’s Java User Group) - Hold a seat on the JCP SE/EE Executive Committee - Run Adopt a JSR and Adopt OpenJDK programmes
- 5. How this session is going to work • This session may be a little different to advertised... - Being a “Well-Grounded Java Developer” changes over time • This session will go fairly quickly - But the slides will be made available - And you can always get hold of us for any questions • There are USB keys going around - Copy over the WGJD into $WORKSPACE - Install Java 7 if you haven't done so already! - Your major IDEs should work • Pair/Group programming works best - Have the least experienced person do the actual typing!
- 6. Attempted Timetable Introduction to Java 7 - 0900-0955 (~55 mins) – BREAK (5 minutes) Polyglot and functional programming - 1000-1030 (~30 mins) BREAK (30 minutes) Polyglot and functional programming - 1100-1125 (~25 mins) – BREAK (5 minutes) Modern Java Concurrency - 1130-1230(~60 mins)
- 7. Real developers code through bathroom breaks 7
- 8. The Well-Grounded Java Developer... • Is not just a Java language whizz - Understands the basics of the JVM - Understands software development is a social activity - Utilises software craftsmanship approaches - Understands physical and logical architectures • Is using Java 7 - At least at home if not at work • Is looking at polyglot and functional programming • Is looking at modern concurrency practices
- 9. Java 7 - Why is it important • The WGJD wants to code rapidly • The WGJD wants to code concisely • The WGJD wants to take advantage of: - The compiler (including JIT) - The JVM (including GC) • Java 7 gives you many improvements in these areas
- 10. OpenJDK - The Java 7/8 split • 20 Sep 2010 - Mark Reinhold announces Plan B - Splits OpenJDK plan into Java 7 (July 2011) & Java 8 (2013) - Popular choice with the community • July 2011 - Java 7 goes gold - Wow, a whole year ago now! • Some new features to be delayed until JDK 8 - Lambda Expressions - Modularisation (aka Jigsaw) • JRockit features to be merged into OpenJDK - Enhanced management of the JVM - Ongoing merge of the VM engineering groups
- 11. Contents of JDK 7 Release Project Coin NIO.2 Method Handles invokedynamic Concurrency Refresh Odds & Ends
- 12. Project Coin • “Small” changes • Language Level, not VM • Stay away from the type system • Developed in a very OSS manner - Was an ‘interesting’ experience for all involved
- 13. Project Coin - Highlights Strings in switch try-with-resources (aka ARM or TWR) Diamond Syntax Multi-catch with precise re-throw Enhanced syntax for numeric literals Varargs / autoboxing warning
- 14. Strings in switch • Code along exercise: FeedingSchedule.java
- 16. Diamond syntax • Code along exercise: HordeOfOtters.java
- 17. Personally I’d stick to Java 1.4 17
- 18. NIO.2 - New I/O version 2 • A new file system and path abstraction • Based on Path - An abstract view of a ‘file like’ system • Files class contains many helper methods, including - File manipulation (copy, move, rename etc) - Walking directory trees - Native file system support (e.g. symbolic links) • Works with existing java.io.File code
- 19. NIO.2 - New I/O version 2 • Asynchronous (non-blocking) I/O • For sockets and files • Mainly utilises java.util.concurrent.Future • New NetworkChannel - Socket/Channel construct - Binding, options and multicast
- 20. Path and Files - Some examples • Code along exercise: Housekeeping.java
- 21. URL stream to file - Java 6 style NIO
- 22. URL stream to file in Java 7 • Code along exercise: VetNotes.java
- 23. NIO.2 - Future base file I/O • See ExceptionExamples.java as well (multi-catch)
- 24. Method Handles • We already have Reflection API - It’s all pretty horrible • JDK 7 introduces a new way to inspect at runtime • Method Handles - Represent the ability to call a method - Implement subclass java.lang.invoke.MethodHandle - Are strongly-typed and typesafe - Needed for invokedynamic - but also used standalone
- 28. invokedynamic • invokedynamic is a key new JVM feature - It’s the first new bytecode since Java 1.0 - Joins invokevirtual, invokestatic, invokeinterface and invokespecial • Relaxes a key part of the static typing system - User code can determine dispatch at runtime • Aims to be as fast as regular method dispatch - e.g. invokevirtual • No Java syntax in Java 7 - But maybe for Java 8
- 29. invokedynamic contd. • JRuby, Jython, Groovy, Clojure et al all benefit - JRuby gains a lot, Clojure not so much - Even Scala is using it now! • JRuby works closely with the JSR-292 team - Charlie Nutter (winner of JAX special jury award) - Big wins - Makes JRuby enticing to Ruby developers • Java 8 also gets a boost - Lambdas will be implemented with invokedynamic
- 30. Other JVM Languages? Meh - Use the Golden Hammer! 30
- 31. Java 7 Concurrency Refresh • Java 7 has a number of new concurrency toys • The headline item is Fork / Join - Similar to Map/Reduce, useful for a certain class of problems - Fork and join executions are not necessarily threads • We’ll cover this in the final hour!
- 32. Java 7 - That’s not all • Nimbus Look & Feel for Swing • New helper classes including Objects - deepEquals() and friends • JDBC with try-with-resources support • Better unicode support - Thanks in part to Tom Christiansen of Perl fame • More small changes in niche areas
- 33. New Faces in the OpenJDK
- 34. What We Didn’t Have Time To Talk About • Java 8 - Java FX 2.x • Java EE 6/7 • Java ME • JDK Enhancement Proposals (JEPs) • Java Community Process (JCP) - Java Specification Requests (JSRs) - Java User Groups (JUGs) and 'Adopt' programs
- 35. What? You still here? Go take a break will you! 35
Editor's Notes
- #2: TODO: Overall if we have time, replace images of code with actual code\n
- #3: First laugh of the day\n
- #4: We're (Fairly) old (mostly FOSS) hackers and cause trouble at conferences\nBen is known for his performance & concurrency work\nMartijn is also known as the “Diabolical Developer”\n
- #5: * Charities, financial services and beer.\n
- #6: \n
- #7: * No plan survives contact with the enemy, that would be you guys!\n
- #8: In reference to timeline\n
- #9: \n
- #10: \n
- #11: * Surprising fact, but the largest group of engineers ever were deployed on Java 7\nWe'll cover Java 8 in our talk later this week!\nClosures, Lambdas, SAM literals\n
- #12: \n
- #13: Most developers make f&^^%n terrible language designers\nWant to get involved? Talk to us about Adopt OpenJDK\n
- #14: \n
- #15: * Incidentally JSR 310 provides much better date and time handling, join our Adopt a JSR program to learn more!\n
- #16: \n
- #17: \n
- #18: In reference to how awesome the coin features are\n
- #19: \n
- #20: \n
- #21: \n
- #22: \n
- #23: * "So we had a bit of a fail with our domain hosting agency..."\n* TODO - We need to provide some text files hosted at a URL\n
- #24: * Do the “who knows j.u.c”? question at the start\n* Spot the numeric literal! \n
- #25: TODO: “I want my Coffee!” physical theatre\n
- #26: See & use interfaces/classes/methods and fields at runtime.\n* Call methods without knowing their names at runtime\n* Think “Reflection done in a safe and modern way\n
- #27: TODO: Do we have a simpler example? Perhaps we could extract the Runnable into a local var for readability?\nTODO: Was this the one with the code bug?\n
- #28: * Spot the setAccessible() - we need to make the private method accessible\n
- #29: Lookup happens in the context where it’s called from.\nMeans that if you lookup() from inside a class, you can see all the private methods.\nThen you can selectively hand out a reference for others to call them.\n* Method handle calls can be inlined through\n* Also unreflect\n
- #30: * Method names / signatures will not need to be known at compile time\n* uses MethodHandle\n
- #31: \n
- #32: In reference to how awesome invokedynamic is\n
- #33: \n
- #34: * Talk to the pattern in the many Java libraries - Foo / AbstractFoo / Foos\n
- #35: \n
- #36: \n
- #37: \n