Improve performance

Add more documentation to the readme

Author: Techcable

.gitignore

@@ -6,4 +6,7 @@
 DiffUtils.egg-info
 
 # Python
-*.pyc
+*.pyc
+
+# A big file I use to test performance:
+joined.srg

README.md

@@ -1,2 +1,15 @@
 # SrgLib
 A python library for handling srg files and related stuff
+
+## Features
+- Correctly Parses and Outputs srg files
+  - Is able to [fully reproduce](test.py#21) the original srg file from its internal representation
+- Well documented
+  - All public methods have complete documentation
+- Object oriented
+  - Types have objects, which have useful utility methods
+- Fast
+  - Parses 27,000 lines in under a second (on pypy)
+  - Takes slightly over a second on python 3.5
+- Compatible with python 2.7 and 3.5
+  - Also compatible with pypy

profiling.py

@@ -0,0 +1,9 @@
+from srg.output import serialize_srg
+from srg.parser import parse_srg
+
+with open('joined.srg') as file:
+    lines = file.read().splitlines(keepends=False)
+
+mappings = parse_srg(lines)
+
+output_text = list(serialize_srg(mappings))

setup.py

@@ -7,5 +7,6 @@
       description='A python library for handling srg files and related stuff)',
       author='Techcable',
       author_emails='[email protected]',
-      packages=["srg"]
+      packages=["srg"],
+      requires=["enum34"]
       )

srg/__init__.py

@@ -4,7 +4,7 @@
 from srg.types import Type
 
 
-class MethodData:
+class MethodData(object):
     def __init__(self, type, name, args, return_type):
         """
         Create a method data object
@@ -51,7 +51,7 @@ def __eq__(self, other):
                and self.args == other.args\
                and self.return_type == other.return_type
 
-class FieldData:
+class FieldData(object):
     def __init__(self, type, name):
         """
         Create field data
@@ -231,7 +231,7 @@ def validate_method(method):
 
 # String Validator
 
-_identifier_regex = re.compile("[\w_]+")
+_identifier_regex = re.compile("[\w_\$]+")
 _type_regex = re.compile('([\w_]+[\$\.])*([\w_]+)')
 _package_regex = re.compile('([\w_]+\.)*([\w_]+)')
 

srg/types.py

@@ -3,7 +3,7 @@
 import srg
 
 
-class Type:
+class Type(object):
     """
     A java class
 
@@ -139,7 +139,7 @@ def __init__(self, kind):
 
         :param kind: the kind of this primitive
         """
-        super().__init__()
+        super(PrimitiveType, self).__init__()
         self._kind = kind
 
     def get_kind(self):
@@ -167,7 +167,7 @@ def __init__(self, element_type):
 
         :param element_type: the element type of the array
         """
-        super().__init__()
+        super(ArrayType, self).__init__()
         self._type = element_type
 
     def get_element_type(self):
@@ -196,33 +196,27 @@ class PrimitiveKind(Enum):
     void is actually a valid java type (believe it or not)
     """
 
-    BYTE = 0
-    SHORT = 1
-    INT = 2
-    LONG = 3
-    FLOAT = 4
-    DOUBLE = 5
-    CHAR = 6
-    BOOLEAN = 7
-    VOID = 8
+    BYTE = 'B'
+    SHORT = 'S'
+    INT = 'I'
+    LONG = 'J'
+    FLOAT = 'F'
+    DOUBLE = 'D'
+    CHAR = 'C'
+    BOOLEAN = 'Z'
+    VOID = 'V'
 
     def get_descriptor(self):
         """
         Get the bytecode descriptor of this primitive type
 
         :return: the bytecode descriptor of this primitive type
         """
-        return {
-            PrimitiveKind.BYTE: "B",
-            PrimitiveKind.SHORT: "S",
-            PrimitiveKind.INT: "I",
-            PrimitiveKind.LONG: "J",
-            PrimitiveKind.FLOAT: "F",
-            PrimitiveKind.DOUBLE: "D",
-            PrimitiveKind.CHAR: "C",
-            PrimitiveKind.BOOLEAN: "Z",
-            PrimitiveKind.VOID: "V"
-        }[self]
+        return self.value
+
+    @staticmethod
+    def from_descriptor(descriptor):
+        return PrimitiveKind(descriptor)
 
 
 def parse_internal_name(name):
@@ -235,7 +229,6 @@ def parse_internal_name(name):
     """
     return parse_name(name.replace("/", "."))
 
-
 def parse_name(name):
     """
     Parse the given name into a Type
@@ -285,10 +278,10 @@ def parse_descriptor(descriptor):
             return ArrayType(parse_descriptor(descriptor[1:]))
         except ValueError:
             raise ValueError("Couldn't parse descriptor " + descriptor)
-    for primitive in PrimitiveKind:
-        if primitive.get_descriptor() == descriptor:
-            return PrimitiveType(primitive)
     if descriptor.startswith("L") and srg.is_valid_type(descriptor[1:].replace("/", ".")):
         return Type(descriptor[1:-1].replace('/', "."))
+    primitive = PrimitiveKind.from_descriptor(descriptor)
+    if primitive is not None:
+        return PrimitiveType(primitive)
     # Unable to parse
     raise ValueError("Couldn't parse descriptor " + descriptor)