1 ================================================= 2 Kaleidoscope: Tutorial Introduction and the Lexer 3 ================================================= 4 5 .. contents:: 6 :local: 7 8 Tutorial Introduction 9 ===================== 10 11 Welcome to the "Implementing a language with LLVM" tutorial. This 12 tutorial runs through the implementation of a simple language, showing 13 how fun and easy it can be. This tutorial will get you up and started as 14 well as help to build a framework you can extend to other languages. The 15 code in this tutorial can also be used as a playground to hack on other 16 LLVM specific things. 17 18 The goal of this tutorial is to progressively unveil our language, 19 describing how it is built up over time. This will let us cover a fairly 20 broad range of language design and LLVM-specific usage issues, showing 21 and explaining the code for it all along the way, without overwhelming 22 you with tons of details up front. 23 24 It is useful to point out ahead of time that this tutorial is really 25 about teaching compiler techniques and LLVM specifically, *not* about 26 teaching modern and sane software engineering principles. In practice, 27 this means that we'll take a number of shortcuts to simplify the 28 exposition. For example, the code leaks memory, uses global variables 29 all over the place, doesn't use nice design patterns like 30 `visitors <http://en.wikipedia.org/wiki/Visitor_pattern>`_, etc... but 31 it is very simple. If you dig in and use the code as a basis for future 32 projects, fixing these deficiencies shouldn't be hard. 33 34 I've tried to put this tutorial together in a way that makes chapters 35 easy to skip over if you are already familiar with or are uninterested 36 in the various pieces. The structure of the tutorial is: 37 38 - `Chapter #1 <#language>`_: Introduction to the Kaleidoscope 39 language, and the definition of its Lexer - This shows where we are 40 going and the basic functionality that we want it to do. In order to 41 make this tutorial maximally understandable and hackable, we choose 42 to implement everything in Objective Caml instead of using lexer and 43 parser generators. LLVM obviously works just fine with such tools, 44 feel free to use one if you prefer. 45 - `Chapter #2 <OCamlLangImpl2.html>`_: Implementing a Parser and 46 AST - With the lexer in place, we can talk about parsing techniques 47 and basic AST construction. This tutorial describes recursive descent 48 parsing and operator precedence parsing. Nothing in Chapters 1 or 2 49 is LLVM-specific, the code doesn't even link in LLVM at this point. 50 :) 51 - `Chapter #3 <OCamlLangImpl3.html>`_: Code generation to LLVM IR - 52 With the AST ready, we can show off how easy generation of LLVM IR 53 really is. 54 - `Chapter #4 <OCamlLangImpl4.html>`_: Adding JIT and Optimizer 55 Support - Because a lot of people are interested in using LLVM as a 56 JIT, we'll dive right into it and show you the 3 lines it takes to 57 add JIT support. LLVM is also useful in many other ways, but this is 58 one simple and "sexy" way to shows off its power. :) 59 - `Chapter #5 <OCamlLangImpl5.html>`_: Extending the Language: 60 Control Flow - With the language up and running, we show how to 61 extend it with control flow operations (if/then/else and a 'for' 62 loop). This gives us a chance to talk about simple SSA construction 63 and control flow. 64 - `Chapter #6 <OCamlLangImpl6.html>`_: Extending the Language: 65 User-defined Operators - This is a silly but fun chapter that talks 66 about extending the language to let the user program define their own 67 arbitrary unary and binary operators (with assignable precedence!). 68 This lets us build a significant piece of the "language" as library 69 routines. 70 - `Chapter #7 <OCamlLangImpl7.html>`_: Extending the Language: 71 Mutable Variables - This chapter talks about adding user-defined 72 local variables along with an assignment operator. The interesting 73 part about this is how easy and trivial it is to construct SSA form 74 in LLVM: no, LLVM does *not* require your front-end to construct SSA 75 form! 76 - `Chapter #8 <OCamlLangImpl8.html>`_: Conclusion and other useful 77 LLVM tidbits - This chapter wraps up the series by talking about 78 potential ways to extend the language, but also includes a bunch of 79 pointers to info about "special topics" like adding garbage 80 collection support, exceptions, debugging, support for "spaghetti 81 stacks", and a bunch of other tips and tricks. 82 83 By the end of the tutorial, we'll have written a bit less than 700 lines 84 of non-comment, non-blank, lines of code. With this small amount of 85 code, we'll have built up a very reasonable compiler for a non-trivial 86 language including a hand-written lexer, parser, AST, as well as code 87 generation support with a JIT compiler. While other systems may have 88 interesting "hello world" tutorials, I think the breadth of this 89 tutorial is a great testament to the strengths of LLVM and why you 90 should consider it if you're interested in language or compiler design. 91 92 A note about this tutorial: we expect you to extend the language and 93 play with it on your own. Take the code and go crazy hacking away at it, 94 compilers don't need to be scary creatures - it can be a lot of fun to 95 play with languages! 96 97 The Basic Language 98 ================== 99 100 This tutorial will be illustrated with a toy language that we'll call 101 "`Kaleidoscope <http://en.wikipedia.org/wiki/Kaleidoscope>`_" (derived 102 from "meaning beautiful, form, and view"). Kaleidoscope is a procedural 103 language that allows you to define functions, use conditionals, math, 104 etc. Over the course of the tutorial, we'll extend Kaleidoscope to 105 support the if/then/else construct, a for loop, user defined operators, 106 JIT compilation with a simple command line interface, etc. 107 108 Because we want to keep things simple, the only datatype in Kaleidoscope 109 is a 64-bit floating point type (aka 'float' in O'Caml parlance). As 110 such, all values are implicitly double precision and the language 111 doesn't require type declarations. This gives the language a very nice 112 and simple syntax. For example, the following simple example computes 113 `Fibonacci numbers: <http://en.wikipedia.org/wiki/Fibonacci_number>`_ 114 115 :: 116 117 # Compute the x'th fibonacci number. 118 def fib(x) 119 if x < 3 then 120 1 121 else 122 fib(x-1)+fib(x-2) 123 124 # This expression will compute the 40th number. 125 fib(40) 126 127 We also allow Kaleidoscope to call into standard library functions (the 128 LLVM JIT makes this completely trivial). This means that you can use the 129 'extern' keyword to define a function before you use it (this is also 130 useful for mutually recursive functions). For example: 131 132 :: 133 134 extern sin(arg); 135 extern cos(arg); 136 extern atan2(arg1 arg2); 137 138 atan2(sin(.4), cos(42)) 139 140 A more interesting example is included in Chapter 6 where we write a 141 little Kaleidoscope application that `displays a Mandelbrot 142 Set <OCamlLangImpl6.html#example>`_ at various levels of magnification. 143 144 Lets dive into the implementation of this language! 145 146 The Lexer 147 ========= 148 149 When it comes to implementing a language, the first thing needed is the 150 ability to process a text file and recognize what it says. The 151 traditional way to do this is to use a 152 "`lexer <http://en.wikipedia.org/wiki/Lexical_analysis>`_" (aka 153 'scanner') to break the input up into "tokens". Each token returned by 154 the lexer includes a token code and potentially some metadata (e.g. the 155 numeric value of a number). First, we define the possibilities: 156 157 .. code-block:: ocaml 158 159 (* The lexer returns these 'Kwd' if it is an unknown character, otherwise one of 160 * these others for known things. *) 161 type token = 162 (* commands *) 163 | Def | Extern 164 165 (* primary *) 166 | Ident of string | Number of float 167 168 (* unknown *) 169 | Kwd of char 170 171 Each token returned by our lexer will be one of the token variant 172 values. An unknown character like '+' will be returned as 173 ``Token.Kwd '+'``. If the curr token is an identifier, the value will be 174 ``Token.Ident s``. If the current token is a numeric literal (like 1.0), 175 the value will be ``Token.Number 1.0``. 176 177 The actual implementation of the lexer is a collection of functions 178 driven by a function named ``Lexer.lex``. The ``Lexer.lex`` function is 179 called to return the next token from standard input. We will use 180 `Camlp4 <http://caml.inria.fr/pub/docs/manual-camlp4/index.html>`_ to 181 simplify the tokenization of the standard input. Its definition starts 182 as: 183 184 .. code-block:: ocaml 185 186 (*===----------------------------------------------------------------------=== 187 * Lexer 188 *===----------------------------------------------------------------------===*) 189 190 let rec lex = parser 191 (* Skip any whitespace. *) 192 | [< ' (' ' | '\n' | '\r' | '\t'); stream >] -> lex stream 193 194 ``Lexer.lex`` works by recursing over a ``char Stream.t`` to read 195 characters one at a time from the standard input. It eats them as it 196 recognizes them and stores them in in a ``Token.token`` variant. The 197 first thing that it has to do is ignore whitespace between tokens. This 198 is accomplished with the recursive call above. 199 200 The next thing ``Lexer.lex`` needs to do is recognize identifiers and 201 specific keywords like "def". Kaleidoscope does this with a pattern 202 match and a helper function. 203 204 .. code-block:: ocaml 205 206 (* identifier: [a-zA-Z][a-zA-Z0-9] *) 207 | [< ' ('A' .. 'Z' | 'a' .. 'z' as c); stream >] -> 208 let buffer = Buffer.create 1 in 209 Buffer.add_char buffer c; 210 lex_ident buffer stream 211 212 ... 213 214 and lex_ident buffer = parser 215 | [< ' ('A' .. 'Z' | 'a' .. 'z' | '0' .. '9' as c); stream >] -> 216 Buffer.add_char buffer c; 217 lex_ident buffer stream 218 | [< stream=lex >] -> 219 match Buffer.contents buffer with 220 | "def" -> [< 'Token.Def; stream >] 221 | "extern" -> [< 'Token.Extern; stream >] 222 | id -> [< 'Token.Ident id; stream >] 223 224 Numeric values are similar: 225 226 .. code-block:: ocaml 227 228 (* number: [0-9.]+ *) 229 | [< ' ('0' .. '9' as c); stream >] -> 230 let buffer = Buffer.create 1 in 231 Buffer.add_char buffer c; 232 lex_number buffer stream 233 234 ... 235 236 and lex_number buffer = parser 237 | [< ' ('0' .. '9' | '.' as c); stream >] -> 238 Buffer.add_char buffer c; 239 lex_number buffer stream 240 | [< stream=lex >] -> 241 [< 'Token.Number (float_of_string (Buffer.contents buffer)); stream >] 242 243 This is all pretty straight-forward code for processing input. When 244 reading a numeric value from input, we use the ocaml ``float_of_string`` 245 function to convert it to a numeric value that we store in 246 ``Token.Number``. Note that this isn't doing sufficient error checking: 247 it will raise ``Failure`` if the string "1.23.45.67". Feel free to 248 extend it :). Next we handle comments: 249 250 .. code-block:: ocaml 251 252 (* Comment until end of line. *) 253 | [< ' ('#'); stream >] -> 254 lex_comment stream 255 256 ... 257 258 and lex_comment = parser 259 | [< ' ('\n'); stream=lex >] -> stream 260 | [< 'c; e=lex_comment >] -> e 261 | [< >] -> [< >] 262 263 We handle comments by skipping to the end of the line and then return 264 the next token. Finally, if the input doesn't match one of the above 265 cases, it is either an operator character like '+' or the end of the 266 file. These are handled with this code: 267 268 .. code-block:: ocaml 269 270 (* Otherwise, just return the character as its ascii value. *) 271 | [< 'c; stream >] -> 272 [< 'Token.Kwd c; lex stream >] 273 274 (* end of stream. *) 275 | [< >] -> [< >] 276 277 With this, we have the complete lexer for the basic Kaleidoscope 278 language (the `full code listing <OCamlLangImpl2.html#code>`_ for the 279 Lexer is available in the `next chapter <OCamlLangImpl2.html>`_ of the 280 tutorial). Next we'll `build a simple parser that uses this to build an 281 Abstract Syntax Tree <OCamlLangImpl2.html>`_. When we have that, we'll 282 include a driver so that you can use the lexer and parser together. 283 284 `Next: Implementing a Parser and AST <OCamlLangImpl2.html>`_ 285 286
