Posts in this category
- Current State of Exceptions in Rakudo and Perl 6
- Meet DBIish, a Perl 6 Database Interface
- doc.perl6.org and p6doc
- Exceptions Grant Report for May 2012
- Exceptions Grant Report -- Final update
- Perl 6 Hackathon in Oslo: Be Prepared!
- Localization for Exception Messages
- News in the Rakudo 2012.05 release
- News in the Rakudo 2012.06 release
- Perl 6 Hackathon in Oslo: Report From The First Day
- Perl 6 Hackathon in Oslo: Report From The Second Day
- Quo Vadis Perl?
- Rakudo Hack: Dynamic Export Lists
- SQLite support for DBIish
- Stop The Rewrites!
- Upcoming Perl 6 Hackathon in Oslo, Norway
- A small regex optimization for NQP and Rakudo
- Pattern Matching and Unpacking
- Rakudo's Abstract Syntax Tree
- The REPL trick
- First day at YAPC::Europe 2013 in Kiev
- YAPC Europe 2013 Day 2
- YAPC Europe 2013 Day 3
- A new Perl 6 community server - call for funding
- New Perl 6 community server now live, accepting signups
- A new Perl 6 community server - update
- All Perl 6 modules in a box
- doc.perl6.org: some stats, future directions
- Profiling Perl 6 code on IRC
- Why is it hard to write a compiler for Perl 6?
- Writing docs helps you take the user's perspective
- A shiny perl6.org site
- Creating an entry point for newcomers
- An offer for software developers: free IRC logging
- Announcing try.rakudo.org, an interactive Perl 6 shell in your browser
- Another perl6.org iteration
- Blackjack and Perl 6
- Why I commit Crud to the Perl 6 Test Suite
- This Week's Contribution to Perl 6 Week 5: Implement Str.trans
- This Week's Contribution to Perl 6
- This Week's Contribution to Perl 6 Week 8: Implement $*ARGFILES for Rakudo
- This Week's Contribution to Perl 6 Week 6: Improve Book markup
- This Week's Contribution to Perl 6 Week 2: Fix up a test
- This Week's Contribution to Perl 6 Week 9: Implement Hash.pick for Rakudo
- This Week's Contribution to Perl 6 Week 11: Improve an error message for Hyper Operators
- This Week's Contribution to Perl 6 - Lottery Intermission
- This Week's Contribution to Perl 6 Week 3: Write supporting code for the MAIN sub
- This Week's Contribution to Perl 6 Week 1: A website for proto
- This Week's Contribution to Perl 6 Week 4: Implement :samecase for .subst
- This Week's Contribution to Perl 6 Week 10: Implement samespace for Rakudo
- This Week's Contribution to Perl 6 Week 7: Implement try.rakudo.org
- What is the "Cool" class in Perl 6?
- Report from the Perl 6 Hackathon in Copenhagen
- Custom operators in Rakudo
- A Perl 6 Date Module
- Defined Behaviour with Undefined Values
- Dissecting the "Starry obfu"
- The case for distributed version control systems
- Perl 6: Failing Softly with Unthrown Exceptions
- Perl 6 Compiler Feature Matrix
- The first Perl 6 module on CPAN
- A Foray into Perl 5 land
- Gabor: Keep going
- First Grant Report: Structured Error Messages
- Second Grant Report: Structured Error Messages
- Third Grant Report: Structured Error Messages
- Fourth Grant Report: Structured Error Messages
- Google Summer of Code Mentor Recap
- How core is core?
- How fast is Rakudo's "nom" branch?
- Building a Huffman Tree With Rakudo
- Immutable Sigils and Context
- Is Perl 6 really Perl?
- Mini-Challenge: Write Your Prisoner's Dilemma Strategy
- Longest Palindrome by Regex
- Perl 6: Lost in Wonderland
- Lots of momentum in the Perl 6 community
- Monetize Perl 6?
- Musings on Rakudo's spectest chart
- My first executable from Perl 6
- My first YAPC - YAPC::EU 2010 in Pisa
- Trying to implement new operators - failed
- Programming Languages Are Not Zero Sum
- Perl 6 notes from February 2011
- Notes from the YAPC::EU 2010 Rakudo hackathon
- Let's build an object
- Perl 6 is optimized for fun
- How to get a parse tree for a Perl 6 Program
- Pascal's Triangle in Perl 6
- Perl 6 in 2009
- Perl 6 in 2010
- Perl 6 in 2011 - A Retrospection
- Perl 6 ticket life cycle
- The Perl Survey and Perl 6
- The Perl 6 Advent Calendar
- Perl 6 Questions on Perlmonks
- Physical modeling with Math::Model and Perl 6
- How to Plot a Segment of a Circle with SVG
- Results from the Prisoner's Dilemma Challenge
- Protected Attributes Make No Sense
- Publicity for Perl 6
- PVC - Perl 6 Vocabulary Coach
- Fixing Rakudo Memory Leaks
- Rakudo architectural overview
- Rakudo Rocks
- Rakudo "star" announced
- My personal "I want a PONIE" wish list for Rakudo Star
- Rakudo's rough edges
- Rats and other pets
- The Real World Strikes Back - or why you shouldn't forbid stuff just because you think it's wrong
- Releasing Rakudo made easy
- Set Phasers to Stun!
- Starry Perl 6 obfu
- Recent Perl 6 Developments August 2008
- The State of Regex Modifiers in Rakudo
- Strings and Buffers
- Subroutines vs. Methods - Differences and Commonalities
- A SVG plotting adventure
- A Syntax Highlighter for Perl 6
- Test Suite Reorganization: How to move tests
- The Happiness of Design Convergence
- Thoughts on masak's Perl 6 Coding Contest
- The Three-Fold Function of the Smart Match Operator
- Perl 6 Tidings from September and October 2008
- Perl 6 Tidings for November 2008
- Perl 6 Tidings from December 2008
- Perl 6 Tidings from January 2009
- Perl 6 Tidings from February 2009
- Perl 6 Tidings from March 2009
- Perl 6 Tidings from April 2009
- Perl 6 Tidings from May 2009
- Perl 6 Tidings from May 2009 (second iteration)
- Perl 6 Tidings from June 2009
- Perl 6 Tidings from August 2009
- Perl 6 Tidings from October 2009
- Timeline for a syntax change in Perl 6
- Visualizing match trees
- Want to write shiny SVG graphics with Perl 6? Port Scruffy!
- We write a Perl 6 book for you
- When we reach 100% we did something wrong
- Where Rakudo Lives Now
- Why Rakudo needs NQP
- Why was the Perl 6 Advent Calendar such a Success?
- What you can write in Perl 6 today
- Why you don't need the Y combinator in Perl 6
- You are good enough!
Sat, 14 Mar 2015
Why is it hard to write a compiler for Perl 6?
Russian translation available; Пост доступен на сайте softdroid.net: Почему так трудно написать компилятор для Perl 6?.
Today's deceptively simple question on #perl6: is it harder to write a compiler for Perl 6 than for any other programming language?
The answer is simple: yes, it's harder (and more work) than for many other languages. The more involved question is: why?
So, let's take a look. The first point is organizational: Perl 6 isn't yet fully explored and formally specified; it's much more stable than it used to be, but less stable than, say, targeting C89.
But even if you disregard this point, and target the subset that for example the Rakudo Perl 6 compiler implements right now, or the wait a year and target the first Perl 6 language release, the point remains valid.
So let's look at some technical aspects.
Static vs. Dynamic
Perl 6 has both static and dynamic corners. For example, lexical lookups are statical, in the sense that they can be resolved at compile time. But that's not optional. For a compiler to properly support native types, it must resolve them at compile time. We also expect the compiler to notify us of certain errors at compile time, so there must be a fair amount of static analysis.
On the other hand, type annotations are optional pretty much anywhere, and methods are late bound. So the compiler must also support features typically found in dynamic languages.
And even though method calls are late bound, composing roles into classes is a compile time operation, with mandatory compile time analysis.
The Perl 6 grammar can change during a parse, for example by newly defined operators, but also through more invasive operations such as defining slangs or macros. Speaking of slangs: Perl 6 doesn't have a single grammar, it switches back and forth between the "main" language, regexes, character classes inside regexes, quotes, and all the other dialects you might think of.
Since the grammar extensions are done with, well, Perl 6 grammars, it forces the parser to be interoperable with Perl 6 regexes and grammars. At which point you might just as well use them for parsing the whole thing, and you get some level of minimally required self-hosting.
In a language like C++, the behavior of the object system is hard-coded into the language, and so the compiler can work under this assumption, and optimize the heck out of it.
In Perl 6, the object system is defined by other objects and classes, the meta objects. So there is another layer of indirection that must be handled.
Mixing of compilation and run time
Declarations like classes, but also
BEGIN blocks and the
right-hand side of
constant declarations are run as soon as they
are parsed. Which means the compiler must be able to run Perl 6 code while
compiling Perl 6 code. And also the other way round, through
More importantly, it must be able to run Perl 6 code before it has finished compiling the whole compilation unit. That means it hasn't even fully constructed the lexical pads, and hasn't initialized all the variables. So it needs special "static lexpads" to which compile-time usages of variables can fall back to. Also the object system has to be able to work with types that haven't been fully declared yet.
So, lots of trickiness involved.
Types are objects defined through their meta objects. That means that when you precompile a module (or even just the setting, that is, the mass of built-ins), the compiler has to serialize the types and their meta objects. Including closures. Do you have any idea how hard it is to correctly serialize closures?
But, classes are mutable. So another module might load a precompiled module, and add another method to it, or otherwise mess with it. Now the compiler has to serialize the fact that, if the second module is loaded, the object from the first module is modified. We say that the serialization context from the second module repossesses the type.
And there are so many ways in which this can go wrong.
One of the many Perl 6 mottos is "torture the implementor on behalf of the user". So it demands not only both static and dynamic typing, but also functional features, continuations, exceptions, lazy lists, a powerful grammar engine, named arguments, variadic arguments, introspection of call frames, closures, lexical and dynamic variables, packed types (for direct interfacing with C libraries, for example), and phasers (code that is automatically run at different phases of the program).
All of these features aren't too hard to implement in isolation, but in combination they are a real killer. And you want it to be fast, right?