Upstream lexer for Smali.

5 files changed, 147 insertions, 0 deletions

diff --git a/AUTHORS b/AUTHORS
index 6db0490c..f04c56b9 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -63,6 +63,7 @@ Other contributors, listed alphabetically, are:
 * Marek Kubica -- Scheme lexer
 * Jochen Kupperschmidt -- Markdown processor
 * Gerd Kurzbach -- Modelica lexer
+* Jon Larimer, Google Inc. -- Smali lexer
 * Olov Lassus -- Dart lexer
 * Sylvestre Ledru -- Scilab lexer
 * Mark Lee -- Vala lexer
diff --git a/CHANGES b/CHANGES
index 31322f30..230afc60 100644
--- a/CHANGES
+++ b/CHANGES
@@ -26,6 +26,7 @@ Version 1.6
   * VGL (PR#12)
   * SourcePawn (PR#39)
   * Ceylon (PR#86)
+  * Smali (Dalvik assembly)
 
 - Fix Template Haskell highlighting (PR#63)
 
diff --git a/pygments/lexers/_mapping.py b/pygments/lexers/_mapping.py
index 0eca2682..d2cb3135 100644
--- a/pygments/lexers/_mapping.py
+++ b/pygments/lexers/_mapping.py
@@ -237,6 +237,7 @@ LEXERS = {
     'SchemeLexer': ('pygments.lexers.functional', 'Scheme', ('scheme', 'scm'), ('*.scm', '*.ss'), ('text/x-scheme', 'application/x-scheme')),
     'ScilabLexer': ('pygments.lexers.math', 'Scilab', ('scilab',), ('*.sci', '*.sce', '*.tst'), ('text/scilab',)),
     'ScssLexer': ('pygments.lexers.web', 'SCSS', ('scss',), ('*.scss',), ('text/x-scss',)),
+    'SmaliLexer': ('pygments.lexers.dalvik', 'Smali', ('smali',), ('*.smali',), ('text/smali',)),
     'SmalltalkLexer': ('pygments.lexers.other', 'Smalltalk', ('smalltalk', 'squeak'), ('*.st',), ('text/x-smalltalk',)),
     'SmartyLexer': ('pygments.lexers.templates', 'Smarty', ('smarty',), ('*.tpl',), ('application/x-smarty',)),
     'SnobolLexer': ('pygments.lexers.other', 'Snobol', ('snobol',), ('*.snobol',), ('text/x-snobol',)),
diff --git a/pygments/lexers/dalvik.py b/pygments/lexers/dalvik.py
new file mode 100644
index 00000000..7da0953a
--- /dev/null
+++ b/pygments/lexers/dalvik.py
@@ -0,0 +1,104 @@
+# -*- coding: utf-8 -*-
+"""
+    pygments.lexers.dalvik
+    ~~~~~~~~~~~~~~~~~~~~~~
+
+    Pygments lexers for Dalvik VM-related languages.
+
+    :copyright: Copyright 2011-2012 by the Pygments team, see AUTHORS.
+    :license: BSD, see LICENSE for details.
+"""
+
+from pygments.lexer import RegexLexer, include, bygroups, using
+from pygments.token import Keyword, Text, Comment, Name, String, Number, \
+                           Punctuation
+
+__all__ = ['SmaliLexer']
+
+
+class SmaliLexer(RegexLexer):
+    """
+    For `Smali <http://code.google.com/p/smali/>`_ (Android/Dalvik) assembly
+    code.
+
+    *New in Pygments 1.6.*
+    """
+    name = 'Smali'
+    aliases = ['smali']
+    filenames = ['*.smali']
+    mimetypes = ['text/smali']
+
+    tokens = {
+        'root': [
+            include('comment'),
+            include('label'),
+            include('field'),
+            include('method'),
+            include('class'),
+            include('directive'),
+            include('access-modifier'),
+            include('instruction'),
+            include('literal'),
+            include('punctuation'),
+            include('type'),
+            include('whitespace')
+        ],
+        'directive': [
+            (r'^[ \t]*\.(class|super|implements|field|subannotation|annotation|'
+             r'enum|method|registers|locals|array-data|packed-switch|'
+             r'sparse-switch|catchall|catch|line|parameter|local|prologue|'
+             r'epilogue|source)', Keyword),
+            (r'^[ \t]*\.end (field|subannotation|annotation|method|array-data|'
+             'packed-switch|sparse-switch|parameter|local)', Keyword),
+            (r'^[ \t]*\.restart local', Keyword),
+        ],
+        'access-modifier': [
+            (r'(public|private|protected|static|final|synchronized|bridge|'
+             r'varargs|native|abstract|strictfp|synthetic|constructor|'
+             r'declared-synchronized|interface|enum|annotation|volatile|'
+             r'transient)', Keyword),
+        ],
+        'whitespace': [
+            (r'\n', Text),
+            (r'\s+', Text),
+        ],
+        'instruction': [
+            (r'\b[vp]\d+\b', Name.Builtin), # registers
+            (r'\b[a-z][A-Za-z0-9/-]+\s+', Text), # instructions
+        ],
+        'literal': [
+            (r'".*"', String),
+            (r'0x[0-9A-Fa-f]+t?', Number.Hex),
+            (r'[0-9]*\.[0-9]+([eE][0-9]+)?[fd]?', Number.Float),
+            (r'[0-9]+L?', Number.Integer),
+        ],
+        'field': [
+            (r'\$?\b([A-Za-z0-9_$]*)(:)', bygroups(Name.Variable,
+                                                   Punctuation)),
+        ],
+        'method': [
+            (r'<(?:cl)?init>', Name.Function), # constructor
+            (r'\$?\b([A-Za-z0-9_$]*)(\()', bygroups(Name.Function,
+                                                    Punctuation)),
+        ],
+        'label': [
+            (r':[A-Za-z0-9_]+', Name.Label),
+        ],
+        'class': [
+            # class names in the form Lcom/namespace/ClassName;
+            # I only want to color the ClassName part, so the namespace part is
+            # treated as 'Text'
+            (r'(L)((?:[A-Za-z0-9_$]+/)*)([A-Za-z0-9_$]+)(;)',
+                bygroups(Keyword.Type, Text, Name.Class, Text)),
+        ],
+        'punctuation': [
+            (r'->', Punctuation),
+            (r'[{},\(\):=\.-]', Punctuation),
+        ],
+        'type': [
+            (r'[ZBSCIJFDV\[]+', Keyword.Type),
+        ],
+        'comment': [
+            (r'#.*?\n', Comment),
+        ],
+    }
diff --git a/tests/examplefiles/hello.smali b/tests/examplefiles/hello.smali
new file mode 100644
index 00000000..e539f00e
--- /dev/null
+++ b/tests/examplefiles/hello.smali
@@ -0,0 +1,40 @@
+# To Recreate:
+#
+# echo -e 'class hello {\n  public static void main(String[] args) {\n
+# System.out.println("hi");\n  }\n}\n' > hello.java
+# javac -target 1.4 -source 1.4 hello.java
+# dx --dex --output=hello.dex hello.class
+# baksmali hello.dex
+# cat out/hello.smali
+
+.class Lhello;
+.super Ljava/lang/Object;
+.source "hello.java"
+
+
+# direct methods
+.method constructor <init>()V
+    .registers 1
+
+    .prologue
+    .line 1
+    invoke-direct {p0}, Ljava/lang/Object;-><init>()V
+
+    return-void
+.end method
+
+.method public static main([Ljava/lang/String;)V
+    .registers 3
+    .parameter
+
+    .prologue
+    .line 3
+    sget-object v0, Ljava/lang/System;->out:Ljava/io/PrintStream;
+
+    const-string v1, "hi"
+
+    invoke-virtual {v0, v1}, Ljava/io/PrintStream;->println(Ljava/lang/String;)V
+
+    .line 4
+    return-void
+.end method