6502/Ophis
1
0
Fork 0
mirror of https://github.com/michaelcmartin/Ophis.git synced 2024-10-12 15:23:38 +00:00
Code Issues Projects Releases Wiki Activity

Major formatting fixes:

Browse Source 
* No more tabs
 * Fix copyright notices to point at right files and name the license right
This commit is contained in:
Michael C. Martin 2012-05-06 20:06:28 -07:00
parent d5ec7bdacd
commit 1df8ad465d
12 changed files with 1508 additions and 1514 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
src/Ophis/CmdLine.py
View File

@ -1,17 +1,17 @@
"""Command line options data.
verbose:
0: Only report errors
1: Announce each file as it is read, and data count (default)
2: As above, but also announce each pass.
3: As above, but print the IR after each pass.
4: As above, but print the labels after each pass.
6510 compatibility and deprecation are handled in Ophis.Main."""
# Copyright 2002 Michael C. Martin.
# You may use, modify, and distribute this file under the BSD
# license: See LICENSE.txt for details.
verbose = 1;
"""Command line options data.
verbose:
0: Only report errors
1: Announce each file as it is read, and data count (default)
2: As above, but also announce each pass.
3: As above, but print the IR after each pass.
4: As above, but print the labels after each pass.
6510 compatibility and deprecation are handled in Ophis.Main."""
# Copyright 2002-2012 Michael C. Martin and additional contributors.
# You may use, modify, and distribute this file under the MIT
# license: See README for details.
verbose = 1;

299
src/Ophis/CorePragmas.py
View File

@ -1,13 +1,10 @@
"""Core pragmas
Provides the core assembler directives. It does not guarantee
compatibility with older versions of P65-Perl."""
Provides the core assembler directives."""
# Copyright 2002 Michael C. Martin.
# You may use, modify, and distribute this file under the BSD
# license: See LICENSE.txt for details.
from __future__ import nested_scopes
# Copyright 2002-2012 Michael C. Martin and additional contributors.
# You may use, modify, and distribute this file under the MIT
# license: See README for details.
import Ophis.IR as IR
import Ophis.Frontend as FE
@ -18,193 +15,193 @@ basecharmap = "".join([chr(x) for x in range(256)])
currentcharmap = basecharmap
def reset():
global loadedfiles, currentcharmap, basecharmap
loadedfiles={}
currentcharmap = basecharmap
global loadedfiles, currentcharmap, basecharmap
loadedfiles={}
currentcharmap = basecharmap
def pragmaInclude(ppt, line, result):
"Includes a source file"
filename = line.expect("STRING").value
line.expect("EOL")
if type(filename)==str: result.append(FE.parse_file(ppt, filename))
"Includes a source file"
filename = line.expect("STRING").value
line.expect("EOL")
if type(filename)==str: result.append(FE.parse_file(ppt, filename))
def pragmaRequire(ppt, line, result):
"Includes a source file at most one time"
filename = line.expect("STRING").value
line.expect("EOL")
if type(filename)==str:
global loadedfiles
if filename not in loadedfiles:
loadedfiles[filename]=1
result.append(FE.parse_file(ppt, filename))
"Includes a source file at most one time"
filename = line.expect("STRING").value
line.expect("EOL")
if type(filename)==str:
global loadedfiles
if filename not in loadedfiles:
loadedfiles[filename]=1
result.append(FE.parse_file(ppt, filename))
def pragmaIncbin(ppt, line, result):
"Includes a binary file"
filename = line.expect("STRING").value
line.expect("EOL")
if type(filename)==str:
try:
f = file(filename, "rb")
bytes = f.read()
f.close()
except IOError:
Err.log ("Could not read "+filename)
return
bytes = [IR.ConstantExpr(ord(x)) for x in bytes]
result.append(IR.Node(ppt, "Byte", *bytes))
"Includes a binary file"
filename = line.expect("STRING").value
line.expect("EOL")
if type(filename)==str:
try:
f = file(filename, "rb")
bytes = f.read()
f.close()
except IOError:
Err.log ("Could not read "+filename)
return
bytes = [IR.ConstantExpr(ord(x)) for x in bytes]
result.append(IR.Node(ppt, "Byte", *bytes))
def pragmaCharmap(ppt, line, result):
"Modify the character map."
global currentcharmap, basecharmap
bytes = readData(line)
if len(bytes) == 0:
currentcharmap = basecharmap
else:
try:
base = bytes[0].data
newsubstr = "".join([chr(x.data) for x in bytes[1:]])
currentcharmap = currentcharmap[:base] + newsubstr + currentcharmap[base+len(newsubstr):]
if len(currentcharmap) != 256 or base < 0 or base > 255:
Err.log("Charmap replacement out of range")
currentcharmap = currentcharmap[:256]
except ValueError:
Err.log("Illegal character in .charmap directive")
"Modify the character map."
global currentcharmap, basecharmap
bytes = readData(line)
if len(bytes) == 0:
currentcharmap = basecharmap
else:
try:
base = bytes[0].data
newsubstr = "".join([chr(x.data) for x in bytes[1:]])
currentcharmap = currentcharmap[:base] + newsubstr + currentcharmap[base+len(newsubstr):]
if len(currentcharmap) != 256 or base < 0 or base > 255:
Err.log("Charmap replacement out of range")
currentcharmap = currentcharmap[:256]
except ValueError:
Err.log("Illegal character in .charmap directive")
def pragmaCharmapbin(ppt, line, result):
"Load a new character map from a file"
global currentcharmap
filename = line.expect("STRING").value
line.expect("EOL")
if type(filename)==str:
try:
f = file(filename, "rb")
bytes = f.read()
f.close()
except IOError:
Err.log ("Could not read "+filename)
return
if len(bytes)==256:
currentcharmap = bytes
else:
Err.log("Character map "+filename+" not 256 bytes long")
"Load a new character map from a file"
global currentcharmap
filename = line.expect("STRING").value
line.expect("EOL")
if type(filename)==str:
try:
f = file(filename, "rb")
bytes = f.read()
f.close()
except IOError:
Err.log ("Could not read "+filename)
return
if len(bytes)==256:
currentcharmap = bytes
else:
Err.log("Character map "+filename+" not 256 bytes long")
def pragmaOrg(ppt, line, result):
"Relocates the PC with no output"
newPC = FE.parse_expr(line)
line.expect("EOL")
result.append(IR.Node(ppt, "SetPC", newPC))
"Relocates the PC with no output"
newPC = FE.parse_expr(line)
line.expect("EOL")
result.append(IR.Node(ppt, "SetPC", newPC))
def pragmaAdvance(ppt, line, result):
"Outputs filler until reaching the target PC"
newPC = FE.parse_expr(line)
line.expect("EOL")
result.append(IR.Node(ppt, "Advance", newPC))
"Outputs filler until reaching the target PC"
newPC = FE.parse_expr(line)
line.expect("EOL")
result.append(IR.Node(ppt, "Advance", newPC))
def pragmaCheckpc(ppt, line, result):
"Enforces that the PC has not exceeded a certain point"
target = FE.parse_expr(line)
line.expect("EOL")
result.append(IR.Node(ppt, "CheckPC", target))
"Enforces that the PC has not exceeded a certain point"
target = FE.parse_expr(line)
line.expect("EOL")
result.append(IR.Node(ppt, "CheckPC", target))
def pragmaAlias(ppt, line, result):
"Assigns an arbitrary label"
lbl = line.expect("LABEL").value
target = FE.parse_expr(line)
result.append(IR.Node(ppt, "Label", lbl, target))
"Assigns an arbitrary label"
lbl = line.expect("LABEL").value
target = FE.parse_expr(line)
result.append(IR.Node(ppt, "Label", lbl, target))
def pragmaSpace(ppt, line, result):
"Reserves space in a data segment for a variable"
lbl = line.expect("LABEL").value
size = line.expect("NUM").value
line.expect("EOL")
result.append(IR.Node(ppt, "Label", lbl, IR.PCExpr()))
result.append(IR.Node(ppt, "SetPC", IR.SequenceExpr([IR.PCExpr(), "+", IR.ConstantExpr(size)])))
"Reserves space in a data segment for a variable"
lbl = line.expect("LABEL").value
size = line.expect("NUM").value
line.expect("EOL")
result.append(IR.Node(ppt, "Label", lbl, IR.PCExpr()))
result.append(IR.Node(ppt, "SetPC", IR.SequenceExpr([IR.PCExpr(), "+", IR.ConstantExpr(size)])))
def pragmaText(ppt, line, result):
"Switches to a text segment"
next = line.expect("LABEL", "EOL")
if next.type == "LABEL":
line.expect("EOL")
segment = next.value
else:
segment = "*text-default*"
result.append(IR.Node(ppt, "TextSegment", segment))
"Switches to a text segment"
next = line.expect("LABEL", "EOL")
if next.type == "LABEL":
line.expect("EOL")
segment = next.value
else:
segment = "*text-default*"
result.append(IR.Node(ppt, "TextSegment", segment))
def pragmaData(ppt, line, result):
"Switches to a data segment (no output allowed)"
next = line.expect("LABEL", "EOL")
if next.type == "LABEL":
line.expect("EOL")
segment = next.value
else:
segment = "*data-default*"
result.append(IR.Node(ppt, "DataSegment", segment))
"Switches to a data segment (no output allowed)"
next = line.expect("LABEL", "EOL")
if next.type == "LABEL":
line.expect("EOL")
segment = next.value
else:
segment = "*data-default*"
result.append(IR.Node(ppt, "DataSegment", segment))
def readData(line):
"Read raw data from a comma-separated list"
if line.lookahead(0).type == "STRING":
data = [IR.ConstantExpr(ord(x)) for x in line.expect("STRING").value.translate(currentcharmap)]
else:
data = [FE.parse_expr(line)]
next = line.expect(',', 'EOL').type
while next == ',':
if line.lookahead(0).type == "STRING":
data.extend([IR.ConstantExpr(ord(x)) for x in line.expect("STRING").value])
else:
data.append(FE.parse_expr(line))
next = line.expect(',', 'EOL').type
return data
"Read raw data from a comma-separated list"
if line.lookahead(0).type == "STRING":
data = [IR.ConstantExpr(ord(x)) for x in line.expect("STRING").value.translate(currentcharmap)]
else:
data = [FE.parse_expr(line)]
next = line.expect(',', 'EOL').type
while next == ',':
if line.lookahead(0).type == "STRING":
data.extend([IR.ConstantExpr(ord(x)) for x in line.expect("STRING").value])
else:
data.append(FE.parse_expr(line))
next = line.expect(',', 'EOL').type
return data
def pragmaByte(ppt, line, result):
"Raw data, a byte at a time"
bytes = readData(line)
result.append(IR.Node(ppt, "Byte", *bytes))
"Raw data, a byte at a time"
bytes = readData(line)
result.append(IR.Node(ppt, "Byte", *bytes))
def pragmaWord(ppt, line, result):
"Raw data, a word at a time, little-endian"
words = readData(line)
result.append(IR.Node(ppt, "Word", *words))
"Raw data, a word at a time, little-endian"
words = readData(line)
result.append(IR.Node(ppt, "Word", *words))
def pragmaDword(ppt, line, result):
"Raw data, a double-word at a time, little-endian"
dwords = readData(line)
result.append(IR.Node(ppt, "Dword", *dwords))
"Raw data, a double-word at a time, little-endian"
dwords = readData(line)
result.append(IR.Node(ppt, "Dword", *dwords))
def pragmaWordbe(ppt, line, result):
"Raw data, a word at a time, big-endian"
words = readData(line)
result.append(IR.Node(ppt, "WordBE", *words))
"Raw data, a word at a time, big-endian"
words = readData(line)
result.append(IR.Node(ppt, "WordBE", *words))
def pragmaDwordbe(ppt, line, result):
"Raw data, a dword at a time, big-endian"
dwords = readData(line)
result.append(IR.Node(ppt, "DwordBE", *dwords))
"Raw data, a dword at a time, big-endian"
dwords = readData(line)
result.append(IR.Node(ppt, "DwordBE", *dwords))
def pragmaScope(ppt, line, result):
"Create a new lexical scoping block"
line.expect("EOL")
result.append(IR.Node(ppt, "ScopeBegin"))
"Create a new lexical scoping block"
line.expect("EOL")
result.append(IR.Node(ppt, "ScopeBegin"))
def pragmaScend(ppt, line, result):
"End the innermost lexical scoping block"
line.expect("EOL")
result.append(IR.Node(ppt, "ScopeEnd"))
"End the innermost lexical scoping block"
line.expect("EOL")
result.append(IR.Node(ppt, "ScopeEnd"))
def pragmaMacro(ppt, line, result):
"Begin a macro definition"
lbl = line.expect("LABEL").value
line.expect("EOL")
result.append(IR.Node(ppt, "MacroBegin", lbl))
"Begin a macro definition"
lbl = line.expect("LABEL").value
line.expect("EOL")
result.append(IR.Node(ppt, "MacroBegin", lbl))
def pragmaMacend(ppt, line, result):
"End a macro definition"
line.expect("EOL")
result.append(IR.Node(ppt, "MacroEnd"))
"End a macro definition"
line.expect("EOL")
result.append(IR.Node(ppt, "MacroEnd"))
def pragmaInvoke(ppt, line, result):
macro = line.expect("LABEL").value
if line.lookahead(0).type == "EOL":
args = []
else:
args = readData(line)
result.append(IR.Node(ppt, "MacroInvoke", macro, *args))
macro = line.expect("LABEL").value
if line.lookahead(0).type == "EOL":
args = []
else:
args = readData(line)
result.append(IR.Node(ppt, "MacroInvoke", macro, *args))

149
src/Ophis/Environment.py
View File

@ -1,75 +1,74 @@
"""Symbol tables and environments for P65.
Implements the symbol lookup, through nested environments -
any non-temporary variable is stored at the top level."""
# Copyright 2002 Michael C. Martin.
# You may use, modify, and distribute this file under the BSD
# license: See LICENSE.txt for details.
from __future__ import nested_scopes
import Ophis.Errors as Err
class Environment:
"""Environment class.
Controls the various scopes and global abstract execution variables."""
def __init__(self):
self.dicts = [{}]
self.stack = [0]
self.pc = 0
self.segmentdict = {}
self.segment = "*text-default*"
self.scopecount = 0
def __contains__(self, item):
if item[0] == '_':
for dict in [self.dicts[i] for i in self.stack]:
if item in dict: return 1
return 0
return item in self.dicts[0]
def __getitem__(self, item):
if item[0] == '_':
for dict in [self.dicts[i] for i in self.stack]:
if item in dict: return dict[item]
else:
if item in self.dicts[0]: return self.dicts[0][item]
Err.log("Unknown label '%s'" % item)
return 0
def __setitem__(self, item, value):
if item[0] == '_':
self.dicts[self.stack[0]][item] = value
else:
self.dicts[0][item] = value
def __str__(self):
return str(self.dicts)
def getPC(self):
return self.pc
def setPC(self, value):
self.pc = value
def incPC(self, amount):
self.pc += amount
def getsegment(self):
return self.segment
def setsegment(self, segment):
self.segmentdict[self.segment] = self.pc
self.segment = segment
self.pc = self.segmentdict.get(segment, 0)
def reset(self):
"Clears out program counter, segment, and scoping information"
self.pc = 0
self.segmentdict = {}
self.segment = "*text-default*"
self.scopecount = 0
if len(self.stack) > 1:
Err.log("Unmatched .scope")
self.stack = [0]
def newscope(self):
"Enters a new scope for temporary labels."
self.scopecount += 1
self.stack.insert(0, self.scopecount)
if len(self.dicts) <= self.scopecount: self.dicts.append({})
def endscope(self):
"Leaves a scope."
if len(self.stack) == 1:
Err.log("Unmatched .scend")
self.stack.pop(0)
"""Symbol tables and environments for Ophis.
Implements the symbol lookup, through nested environments -
any non-temporary variable is stored at the top level."""
# Copyright 2002-2012 Michael C. Martin and additional contributors.
# You may use, modify, and distribute this file under the MIT
# license: See README for details.
import Ophis.Errors as Err
class Environment:
"""Environment class.
Controls the various scopes and global abstract execution variables."""
def __init__(self):
self.dicts = [{}]
self.stack = [0]
self.pc = 0
self.segmentdict = {}
self.segment = "*text-default*"
self.scopecount = 0
def __contains__(self, item):
if item[0] == '_':
for dict in [self.dicts[i] for i in self.stack]:
if item in dict: return 1
return 0
return item in self.dicts[0]
def __getitem__(self, item):
if item[0] == '_':
for dict in [self.dicts[i] for i in self.stack]:
if item in dict: return dict[item]
else:
if item in self.dicts[0]: return self.dicts[0][item]
Err.log("Unknown label '%s'" % item)
return 0
def __setitem__(self, item, value):
if item[0] == '_':
self.dicts[self.stack[0]][item] = value
else:
self.dicts[0][item] = value
def __str__(self):
return str(self.dicts)
def getPC(self):
return self.pc
def setPC(self, value):
self.pc = value
def incPC(self, amount):
self.pc += amount
def getsegment(self):
return self.segment
def setsegment(self, segment):
self.segmentdict[self.segment] = self.pc
self.segment = segment
self.pc = self.segmentdict.get(segment, 0)
def reset(self):
"Clears out program counter, segment, and scoping information"
self.pc = 0
self.segmentdict = {}
self.segment = "*text-default*"
self.scopecount = 0
if len(self.stack) > 1:
Err.log("Unmatched .scope")
self.stack = [0]
def newscope(self):
"Enters a new scope for temporary labels."
self.scopecount += 1
self.stack.insert(0, self.scopecount)
if len(self.dicts) <= self.scopecount: self.dicts.append({})
def endscope(self):
"Leaves a scope."
if len(self.stack) == 1:
Err.log("Unmatched .scend")
self.stack.pop(0)

48
src/Ophis/Errors.py
View File

@ -1,24 +1,24 @@
"""Error logging
Keeps track of the number of errors inflicted so far, and
where in the assembly the errors are occurring."""
# Copyright 2002 Michael C. Martin.
# You may use, modify, and distribute this file under the BSD
# license: See LICENSE.txt for details.
count = 0
currentpoint = "<Top Level>"
def log(err):
"""Reports an error at the current program point, and increases
the global error count."""
global count
count = count+1
print currentpoint+": "+err
def report():
"Print out the number of errors."
if count == 0: print "No errors"
elif count == 1: print "1 error"
else: print str(count)+" errors"
"""Error logging
Keeps track of the number of errors inflicted so far, and
where in the assembly the errors are occurring."""
# Copyright 2002-2012 Michael C. Martin and additional contributors.
# You may use, modify, and distribute this file under the MIT
# license: See README for details.
count = 0
currentpoint = "<Top Level>"
def log(err):
"""Reports an error at the current program point, and increases
the global error count."""
global count
count = count+1
print currentpoint+": "+err
def report():
"Print out the number of errors."
if count == 0: print "No errors"
elif count == 1: print "1 error"
else: print str(count)+" errors"

664
src/Ophis/Frontend.py
View File

@ -1,333 +1,331 @@
"""Lexer and Parser
Constructs a list of IR nodes from a list of input strings."""
from __future__ import nested_scopes
import Ophis.Errors as Err
import Ophis.Opcodes as Ops
import Ophis.IR as IR
import Ophis.CmdLine as Cmd
import os
# Copyright 2002 Michael C. Martin.
# You may use, modify, and distribute this file under the BSD
# license: See LICENSE.txt for details.
class Lexeme:
"Class for lexer tokens. Used by lexer and parser."
def __init__(self, type="UNKNOWN", value=None):
self.type = type.upper()
self.value = value
def __str__(self):
if self.value == None:
return self.type
else:
return self.type+":"+str(self.value)
def __repr__(self):
return "Lexeme("+`self.type`+", "+`self.value`+")"
def matches(self, other):
"1 if Lexemes a and b have the same type."
return self.type == other.type
bases = {"$":("hexadecimal", 16),
"%":("binary", 2),
"0":("octal", 8)}
punctuation = "#,`<>():.+-*/&|^[]"
def lex(point, line):
"""Turns a line of source into a sequence of lexemes."""
Err.currentpoint = point
result = []
def is_opcode(op):
"Tests whether a string is an opcode or an identifier"
return op in Ops.opcodes
def add_token(token):
"Converts a substring into a single lexeme"
if token == "":
return
if token == "0":
result.append(Lexeme("NUM", 0))
return
firstchar = token[0]
rest = token[1:]
if firstchar == '"':
result.append(Lexeme("STRING", rest))
return
elif firstchar in bases:
try:
result.append(Lexeme("NUM", long(rest, bases[firstchar][1])))
return
except ValueError:
Err.log('Invalid '+bases[firstchar][0]+' constant: '+rest)
result.append(Lexeme("NUM", 0))
return
elif firstchar.isdigit():
try:
result.append(Lexeme("NUM", long(token)))
except ValueError:
Err.log('Identifiers may not begin with a number')
result.append(Lexeme("LABEL", "ERROR"))
return
elif firstchar == "'":
if len(rest) == 1:
result.append(Lexeme("NUM", ord(rest)))
else:
Err.log("Invalid character constant '"+rest+"'")
result.append(Lexeme("NUM", 0))
return
elif firstchar in punctuation:
if rest != "":
Err.log("Internal lexer error! '"+token+"' can't happen!")
result.append(Lexeme(firstchar))
return
else: # Label, opcode, or index register
id = token.lower()
if is_opcode(id):
result.append(Lexeme("OPCODE", id))
elif id == "x":
result.append(Lexeme("X"))
elif id == "y":
result.append(Lexeme("Y"))
else:
result.append(Lexeme("LABEL", id))
return
# should never reach here
Err.log("Internal lexer error: add_token fall-through")
def add_EOL():
"Adds an end-of-line lexeme"
result.append(Lexeme("EOL"))
# Actual routine begins here
value = ""
quotemode = 0
backslashmode = 0
for c in line.strip():
if backslashmode:
backslashmode = 0
value = value + c
elif c == "\\":
backslashmode = 1
elif quotemode:
if c == '"':
quotemode = 0
else:
value = value + c
elif c == ';':
add_token(value)
value = ""
break
elif c.isspace():
add_token(value)
value = ""
elif c in punctuation:
add_token(value)
add_token(c)
value = ""
elif c == '"':
add_token(value)
value = '"'
quotemode = 1
else:
value = value + c
if backslashmode:
Err.log("Backslashed newline")
if quotemode:
Err.log("Unterminated string constant")
add_token(value)
add_EOL()
return result
class ParseLine:
"Maintains the parse state of a line of code. Enables arbitrary lookahead."
def __init__(self, lexemes):
self.lexemes = lexemes
self.location = 0
def lookahead(self, i):
"""Returns the token i units ahead in the parse.
lookahead(0) returns the next token; trying to read off the end of
the sequence returns the last token in the sequence (usually EOL)."""
target = self.location+i
if target >= len(self.lexemes): target = -1
return self.lexemes[target]
def pop(self):
"Returns and removes the next element in the line."
old = self.location
if self.location < len(self.lexemes)-1: self.location += 1
return self.lexemes[old]
def expect(self, *tokens):
"""Reads a token from the ParseLine line and returns it if it's of a type
in the sequence tokens. Otherwise, it logs an error."""
token = self.pop()
if token.type not in tokens:
Err.log('Expected: "'+'", "'.join(tokens)+'"')
return token
pragma_modules = []
def parse_expr(line):
"Parses an Ophis arithmetic expression."
def atom():
"Parses lowest-priority expression components."
next = line.lookahead(0).type
if next == "NUM":
return IR.ConstantExpr(line.expect("NUM").value)
elif next == "LABEL":
return IR.LabelExpr(line.expect("LABEL").value)
elif next == "^":
line.expect("^")
return IR.PCExpr()
elif next == "[":
line.expect("[")
result = parse_expr(line)
line.expect("]")
return result
elif next == "+":
offset = 0
while next == "+":
offset += 1
line.expect("+")
next = line.lookahead(0).type
return IR.LabelExpr("*"+str(templabelcount+offset))
elif next == "-":
offset = 1
while next == "-":
offset -= 1
line.expect("-")
next = line.lookahead(0).type
return IR.LabelExpr("*"+str(templabelcount+offset))
elif next == ">":
line.expect(">")
return IR.HighByteExpr(atom())
elif next == "<":
line.expect("<")
return IR.LowByteExpr(atom())
else:
Err.log('Expected: expression')
def precedence_read(constructor, reader, separators):
"""Handles precedence. The reader argument is a function that returns
expressions that bind more tightly than these; separators is a list
of strings naming the operators at this precedence level. The
constructor argument is a class, indicating what node type holds
objects of this precedence level.
Returns a list of Expr objects with separator strings between them."""
result = [reader()] # first object
nextop = line.lookahead(0).type
while (nextop in separators):
line.expect(nextop)
result.append(nextop)
result.append(reader())
nextop = line.lookahead(0).type
if len(result) == 1: return result[0]
return constructor(result)
def term():
"Parses * and /"
return precedence_read(IR.SequenceExpr, atom, ["*", "/"])
def arith():
"Parses + and -"
return precedence_read(IR.SequenceExpr, term, ["+", "-"])
def bits():
"Parses &, |, and ^"
return precedence_read(IR.SequenceExpr, arith, ["&", "|", "^"])
return bits()
def parse_line(ppt, lexemelist):
"Turn a line of source into an IR Node."
Err.currentpoint = ppt
result = []
line = ParseLine(lexemelist)
def aux():
"Accumulates all IR nodes defined by this line."
if line.lookahead(0).type == "EOL":
pass
elif line.lookahead(1).type == ":":
newlabel=line.expect("LABEL").value
line.expect(":")
result.append(IR.Node(ppt, "Label", newlabel, IR.PCExpr()))
aux()
elif line.lookahead(0).type == "*":
global templabelcount
templabelcount = templabelcount + 1
result.append(IR.Node(ppt, "Label", "*"+str(templabelcount), IR.PCExpr()))
line.expect("*")
aux()
elif line.lookahead(0).type == "." or line.lookahead(0).type == "`":
which = line.expect(".", "`").type
if (which == "."): pragma = line.expect("LABEL").value
else: pragma = "invoke"
pragmaFunction = "pragma"+pragma.title()
for mod in pragma_modules:
if hasattr(mod, pragmaFunction):
getattr(mod, pragmaFunction)(ppt, line, result)
break
else:
Err.log("Unknown pragma "+pragma)
else: # Instruction
opcode = line.expect("OPCODE").value
if line.lookahead(0).type == "#":
mode = "Immediate"
line.expect("#")
arg = parse_expr(line)
line.expect("EOL")
elif line.lookahead(0).type == "(":
line.expect("(")
arg = parse_expr(line)
if line.lookahead(0).type == ",":
mode = "PointerX"
line.expect(",")
line.expect("X")
line.expect(")")
line.expect("EOL")
else:
line.expect(")")
tok = line.expect(",", "EOL").type
if tok == "EOL":
mode = "Pointer"
else:
mode = "PointerY"
line.expect("Y")
line.expect("EOL")
elif line.lookahead(0).type == "EOL":
mode = "Implied"
arg = None
else:
arg = parse_expr(line)
tok = line.expect("EOL", ",").type
if tok == ",":
tok = line.expect("X", "Y").type
if tok == "X": mode = "MemoryX"
else: mode = "MemoryY"
line.expect("EOL")
else: mode = "Memory"
result.append(IR.Node(ppt, mode, opcode, arg))
aux()
result = [node for node in result if node is not IR.NullNode]
if len(result) == 0: return IR.NullNode
if len(result) == 1: return result[0]
return IR.SequenceNode(ppt, result)
def parse_file(ppt, filename):
"Loads a .P65 source file, and returns an IR list."
Err.currentpoint = ppt
if Cmd.verbose > 0: print "Loading "+filename
try:
f = file(filename)
linelist = f.readlines()
f.close()
pptlist = ["%s:%d" % (filename, i+1) for i in range(len(linelist))]
lexlist = map(lex, pptlist, linelist)
IRlist = map(parse_line, pptlist, lexlist)
IRlist = [node for node in IRlist if node is not IR.NullNode]
return IR.SequenceNode(ppt, IRlist)
except IOError:
Err.log ("Could not read "+filename)
return IR.NullNode
def parse(filename):
"Top level parsing routine, taking a source file name and returning an IR list."
global templabelcount
templabelcount = 0
return parse_file("<Top Level>", filename)
"""Lexer and Parser
Constructs a list of IR nodes from a list of input strings."""
import Ophis.Errors as Err
import Ophis.Opcodes as Ops
import Ophis.IR as IR
import Ophis.CmdLine as Cmd
import os
# Copyright 2002-2012 Michael C. Martin and additional contributors.
# You may use, modify, and distribute this file under the MIT
# license: See README for details.
class Lexeme:
"Class for lexer tokens. Used by lexer and parser."
def __init__(self, type="UNKNOWN", value=None):
self.type = type.upper()
self.value = value
def __str__(self):
if self.value == None:
return self.type
else:
return self.type+":"+str(self.value)
def __repr__(self):
return "Lexeme("+`self.type`+", "+`self.value`+")"
def matches(self, other):
"1 if Lexemes a and b have the same type."
return self.type == other.type
bases = {"$":("hexadecimal", 16),
"%":("binary", 2),
"0":("octal", 8)}
punctuation = "#,`<>():.+-*/&|^[]"
def lex(point, line):
"""Turns a line of source into a sequence of lexemes."""
Err.currentpoint = point
result = []
def is_opcode(op):
"Tests whether a string is an opcode or an identifier"
return op in Ops.opcodes
def add_token(token):
"Converts a substring into a single lexeme"
if token == "":
return
if token == "0":
result.append(Lexeme("NUM", 0))
return
firstchar = token[0]
rest = token[1:]
if firstchar == '"':
result.append(Lexeme("STRING", rest))
return
elif firstchar in bases:
try:
result.append(Lexeme("NUM", long(rest, bases[firstchar][1])))
return
except ValueError:
Err.log('Invalid '+bases[firstchar][0]+' constant: '+rest)
result.append(Lexeme("NUM", 0))
return
elif firstchar.isdigit():
try:
result.append(Lexeme("NUM", long(token)))
except ValueError:
Err.log('Identifiers may not begin with a number')
result.append(Lexeme("LABEL", "ERROR"))
return
elif firstchar == "'":
if len(rest) == 1:
result.append(Lexeme("NUM", ord(rest)))
else:
Err.log("Invalid character constant '"+rest+"'")
result.append(Lexeme("NUM", 0))
return
elif firstchar in punctuation:
if rest != "":
Err.log("Internal lexer error! '"+token+"' can't happen!")
result.append(Lexeme(firstchar))
return
else: # Label, opcode, or index register
id = token.lower()
if is_opcode(id):
result.append(Lexeme("OPCODE", id))
elif id == "x":
result.append(Lexeme("X"))
elif id == "y":
result.append(Lexeme("Y"))
else:
result.append(Lexeme("LABEL", id))
return
# should never reach here
Err.log("Internal lexer error: add_token fall-through")
def add_EOL():
"Adds an end-of-line lexeme"
result.append(Lexeme("EOL"))
# Actual routine begins here
value = ""
quotemode = 0
backslashmode = 0
for c in line.strip():
if backslashmode:
backslashmode = 0
value = value + c
elif c == "\\":
backslashmode = 1
elif quotemode:
if c == '"':
quotemode = 0
else:
value = value + c
elif c == ';':
add_token(value)
value = ""
break
elif c.isspace():
add_token(value)
value = ""
elif c in punctuation:
add_token(value)
add_token(c)
value = ""
elif c == '"':
add_token(value)
value = '"'
quotemode = 1
else:
value = value + c
if backslashmode:
Err.log("Backslashed newline")
if quotemode:
Err.log("Unterminated string constant")
add_token(value)
add_EOL()
return result
class ParseLine:
"Maintains the parse state of a line of code. Enables arbitrary lookahead."
def __init__(self, lexemes):
self.lexemes = lexemes
self.location = 0
def lookahead(self, i):
"""Returns the token i units ahead in the parse.
lookahead(0) returns the next token; trying to read off the end of
the sequence returns the last token in the sequence (usually EOL)."""
target = self.location+i
if target >= len(self.lexemes): target = -1
return self.lexemes[target]
def pop(self):
"Returns and removes the next element in the line."
old = self.location
if self.location < len(self.lexemes)-1: self.location += 1
return self.lexemes[old]
def expect(self, *tokens):
"""Reads a token from the ParseLine line and returns it if it's of a type
in the sequence tokens. Otherwise, it logs an error."""
token = self.pop()
if token.type not in tokens:
Err.log('Expected: "'+'", "'.join(tokens)+'"')
return token
pragma_modules = []
def parse_expr(line):
"Parses an Ophis arithmetic expression."
def atom():
"Parses lowest-priority expression components."
next = line.lookahead(0).type
if next == "NUM":
return IR.ConstantExpr(line.expect("NUM").value)
elif next == "LABEL":
return IR.LabelExpr(line.expect("LABEL").value)
elif next == "^":
line.expect("^")
return IR.PCExpr()
elif next == "[":
line.expect("[")
result = parse_expr(line)
line.expect("]")
return result
elif next == "+":
offset = 0
while next == "+":
offset += 1
line.expect("+")
next = line.lookahead(0).type
return IR.LabelExpr("*"+str(templabelcount+offset))
elif next == "-":
offset = 1
while next == "-":
offset -= 1
line.expect("-")
next = line.lookahead(0).type
return IR.LabelExpr("*"+str(templabelcount+offset))
elif next == ">":
line.expect(">")
return IR.HighByteExpr(atom())
elif next == "<":
line.expect("<")
return IR.LowByteExpr(atom())
else:
Err.log('Expected: expression')
def precedence_read(constructor, reader, separators):
"""Handles precedence. The reader argument is a function that returns
expressions that bind more tightly than these; separators is a list
of strings naming the operators at this precedence level. The
constructor argument is a class, indicating what node type holds
objects of this precedence level.
Returns a list of Expr objects with separator strings between them."""
result = [reader()] # first object
nextop = line.lookahead(0).type
while (nextop in separators):
line.expect(nextop)
result.append(nextop)
result.append(reader())
nextop = line.lookahead(0).type
if len(result) == 1: return result[0]
return constructor(result)
def term():
"Parses * and /"
return precedence_read(IR.SequenceExpr, atom, ["*", "/"])
def arith():
"Parses + and -"
return precedence_read(IR.SequenceExpr, term, ["+", "-"])
def bits():
"Parses &, |, and ^"
return precedence_read(IR.SequenceExpr, arith, ["&", "|", "^"])
return bits()
def parse_line(ppt, lexemelist):
"Turn a line of source into an IR Node."
Err.currentpoint = ppt
result = []
line = ParseLine(lexemelist)
def aux():
"Accumulates all IR nodes defined by this line."
if line.lookahead(0).type == "EOL":
pass
elif line.lookahead(1).type == ":":
newlabel=line.expect("LABEL").value
line.expect(":")
result.append(IR.Node(ppt, "Label", newlabel, IR.PCExpr()))
aux()
elif line.lookahead(0).type == "*":
global templabelcount
templabelcount = templabelcount + 1
result.append(IR.Node(ppt, "Label", "*"+str(templabelcount), IR.PCExpr()))
line.expect("*")
aux()
elif line.lookahead(0).type == "." or line.lookahead(0).type == "`":
which = line.expect(".", "`").type
if (which == "."): pragma = line.expect("LABEL").value
else: pragma = "invoke"
pragmaFunction = "pragma"+pragma.title()
for mod in pragma_modules:
if hasattr(mod, pragmaFunction):
getattr(mod, pragmaFunction)(ppt, line, result)
break
else:
Err.log("Unknown pragma "+pragma)
else: # Instruction
opcode = line.expect("OPCODE").value
if line.lookahead(0).type == "#":
mode = "Immediate"
line.expect("#")
arg = parse_expr(line)
line.expect("EOL")
elif line.lookahead(0).type == "(":
line.expect("(")
arg = parse_expr(line)
if line.lookahead(0).type == ",":
mode = "PointerX"
line.expect(",")
line.expect("X")
line.expect(")")
line.expect("EOL")
else:
line.expect(")")
tok = line.expect(",", "EOL").type
if tok == "EOL":
mode = "Pointer"
else:
mode = "PointerY"
line.expect("Y")
line.expect("EOL")
elif line.lookahead(0).type == "EOL":
mode = "Implied"
arg = None
else:
arg = parse_expr(line)
tok = line.expect("EOL", ",").type
if tok == ",":
tok = line.expect("X", "Y").type
if tok == "X": mode = "MemoryX"
else: mode = "MemoryY"
line.expect("EOL")
else: mode = "Memory"
result.append(IR.Node(ppt, mode, opcode, arg))
aux()
result = [node for node in result if node is not IR.NullNode]
if len(result) == 0: return IR.NullNode
if len(result) == 1: return result[0]
return IR.SequenceNode(ppt, result)
def parse_file(ppt, filename):
"Loads a .P65 source file, and returns an IR list."
Err.currentpoint = ppt
if Cmd.verbose > 0: print "Loading "+filename
try:
f = file(filename)
linelist = f.readlines()
f.close()
pptlist = ["%s:%d" % (filename, i+1) for i in range(len(linelist))]
lexlist = map(lex, pptlist, linelist)
IRlist = map(parse_line, pptlist, lexlist)
IRlist = [node for node in IRlist if node is not IR.NullNode]
return IR.SequenceNode(ppt, IRlist)
except IOError:
Err.log ("Could not read "+filename)
return IR.NullNode
def parse(filename):
"Top level parsing routine, taking a source file name and returning an IR list."
global templabelcount
templabelcount = 0
return parse_file("<Top Level>", filename)

321
src/Ophis/IR.py
View File

@ -1,161 +1,160 @@
"""P65 Intermediate Representation
Classes for representing the Intermediate nodes upon which the
assembler passes operate."""
# Copyright 2002 Michael C. Martin.
# You may use, modify, and distribute this file under the BSD
# license: See LICENSE.txt for details.
from __future__ import nested_scopes
import Ophis.Errors as Err
class Node:
"""The default IR Node
Instances of Node always have the three fields ppt(Program Point),
nodetype(a string), and data (a list)."""
def __init__(self, ppt, nodetype, *data):
self.ppt = ppt
self.nodetype = nodetype
self.data = list(data)
def accept(self, asmpass, env=None):
"""Implements the Visitor pattern for an assembler pass.
Calls the routine 'asmpass.visitTYPE(self, env)' where
TYPE is the value of self.nodetype."""
Err.currentpoint = self.ppt
routine = getattr(asmpass, "visit"+self.nodetype, asmpass.visitUnknown)
routine(self, env)
def __str__(self):
if self.nodetype != "SEQUENCE":
return str(self.ppt)+": "+self.nodetype+" - "+" ".join(map(str, self.data))
else:
return "\n".join(map(str, self.data))
def __repr__(self):
args = [self.ppt, self.nodetype] + self.data
return "Node(" + ", ".join(map(repr, args)) + ")"
NullNode = Node("<none>", "None")
def SequenceNode(ppt, nodelist):
return Node(ppt, "SEQUENCE", *nodelist)
class Expr:
"""Base class for P65 expressions
All expressions have a field called "data" and a boolean field
called "hardcoded". An expression is hardcoded if it has no
symbolic values in it."""
def __init__(self, data):
self.data = data
self.hardcoded = 0
def __str__(self):
return "<UNKNOWN: "+`self.data`+">"
def valid(self, env=None, PCvalid=0):
"""Returns true if the the expression can be successfully
evaluated in the specified environment."""
return 0
def value(self, env=None):
"Evaluates this expression in the given environment."
return None
class ConstantExpr(Expr):
"Represents a numeric constant"
def __init__(self, data):
self.data = data
self.hardcoded = 1
def __str__(self):
return str(self.data)
def valid(self, env=None, PCvalid=0):
return 1
def value(self, env=None):
return self.data
class LabelExpr(Expr):
"Represents a symbolic constant"
def __init__(self, data):
self.data = data
self.hardcoded = 0
def __str__(self):
return self.data
def valid(self, env=None, PCvalid=0):
return (env is not None) and self.data in env
def value(self, env=None):
return env[self.data]
class PCExpr(Expr):
"Represents the current program counter: ^"
def __init__(self):
self.hardcoded = 0
def __str__(self):
return "^"
def valid(self, env=None, PCvalid=0):
return env is not None and PCvalid
def value(self, env=None):
return env.getPC()
class HighByteExpr(Expr):
"Represents the expression >{data}"
def __init__(self, data):
self.data = data
self.hardcoded = data.hardcoded
def __str__(self):
return ">"+str(self.data)
def valid(self, env=None, PCvalid=0):
return self.data.valid(env, PCvalid)
def value