|
#!/usr/bin/env python |
|
# -*- coding: utf-8 -*- |
|
|
|
# cy.py: a traspiler for an alternate syntax for C. |
|
# Note: this is an incomplete work-in-progress. |
|
|
|
|
|
# Data structures: |
|
|
|
# A token is a tuple. |
|
# First element is a string 'token'. |
|
# Second element is a string which is the token type (e.g. 'OBRACKET'). |
|
# Third element is a string which is the source text of the token (e.g. '['). |
|
# e.g. ('token', 'OBRACKET', '[') |
|
# e.g. ('token', 'IDENTIFIER', 'foo') |
|
# e.g. ('token', 'WS', ' ') |
|
|
|
|
|
# An AST node is a tuple. |
|
# First element is a string 'ast'. |
|
# Second element is a string which is the AST node type (e.g. 'vardecl'). |
|
# Third element is an array which are the child tuples (AST nodes or tokens). |
|
# e.g. ('ast', 'type', [('token', 'IDENTIFIER', 'int')]) |
|
|
|
|
|
# Grammar: |
|
# |
|
# program = statement [WS] { statement [WS] } |
|
# statement = declassign | assign | fundecl | vardecl | return | funcall | if | scope |
|
# fundecl = FUNC WS IDENTIFIER fundeclargs [ fundeclret ] scope |
|
# fundeclargs = OPAREN [ [WS] vardecl { COMMA WS vardecl } [WS] ] CPAREN |
|
# fundeclret = WS ARROW WS type |
|
# vardecl = IDENTIFIER COLON WS type |
|
# scope = COLON INDENT statement { WS statement } DEDENT |
|
# return = RETURN [ expr ] |
|
# declassign = vardecl WS EQ WS expr |
|
# assign = IDENTIFIER WS aoperator WS expr |
|
# aoperator = PLUSEQ | MINUSEQ | STAREQ | SLASHEQ | PERCENTEQ | EQ |
|
# unary = uoperator OPAREN expr CPAREN |
|
# uoperator = MINUS | AMP | STAR | BANG |
|
# binary = OPAREN expr WS boperator WS expr CPAREN |
|
# boperator = PLUS | MINUS | STAR | SLASH | PERCENT |
|
# | LT | LTEQ | GT | GTEQ | EQEQ | BANGEQ |
|
# | AMPAMP | BARBAR | BANG |
|
# | AMP | BAR | LTLT | GTGT | TILDE | CARAT |
|
# if = IF WS expr scope { elif } [else] |
|
# elif = ELIF WS expr scope |
|
# else = ELSE scope |
|
# expr = funcall | binary | unary | IDENTIFIER |
|
# | INTLIT | FLOATLIT | STRINGLIT | BOOLLIT |
|
# funcall = IDENTIFIER OPAREN [ [WS] expr { COMMA WS expr } [WS] ] CPAREN |
|
# type = POINTER LT type GT |
|
# | ARRAY LT type GT |
|
# | FUNCTION LT type GT |
|
# | ARRAY OBRACKET INTLIT CBRACKET LT type GT |
|
# | IDENTIFIER |
|
|
|
# The above EBNF notation is: |
|
# |
|
# Concatenation: rule1 = rule2 rule3 rule4 |
|
# Alternation: rule1 = rule2 | rule3 | rule4 |
|
# Repetition: rule1 = rule2 { rule3 } |
|
# |
|
# Note: lowercase are AST nodes. |
|
# Note: ALLCAPS are tokens (terminals). |
|
|
|
|
|
# Token data structure utils: |
|
|
|
def is_token(token): |
|
"""Returns whether the arg is a token.""" |
|
return isinstance(token, tuple) and len(token) > 0 and token[0] == 'token' |
|
|
|
|
|
def token_type(token): |
|
"""Returns the type of the token (e.g. 'IDENTIFIER').""" |
|
assert is_token(token) |
|
return token[1] |
|
|
|
|
|
def token_text(token): |
|
"""Returns the source text of the token (e.g. 'foo' for a 'IDENTIFIER' token).""" |
|
assert is_token(token) |
|
return token[2] |
|
|
|
|
|
def is_toktype(token, toktype): |
|
""" |
|
Returns whether the token is of the given type. |
|
_toktype_ is e.g. 'IDENTIFIER', 'COLON', etc. |
|
""" |
|
assert is_token(token) |
|
return token_type(token) == toktype |
|
|
|
|
|
|
|
# Tokenizer implementation: |
|
|
|
def load_tokendefs(fpath): |
|
""" |
|
Load the token definitions from the file at fpath. |
|
|
|
Returns a list of tuples of the form (<token type>, <compiled regex>). |
|
|
|
The format of the tokendefs file should be pairs of lines: |
|
- a line which is the token type (e.g. 'IDENTIFIER', 'OBRACKET', etc), |
|
- followed by a line which is the regex which recognizes that token. |
|
|
|
Example tokendefs.txt: |
|
IDENTIFIER |
|
[a-zA-Z][a-zA-Z0-9]* |
|
OPAREN |
|
\( |
|
CPAREN |
|
) |
|
""" |
|
import re |
|
tokendefs = [] |
|
with open(fpath) as f: |
|
for line in f: |
|
token_name = line.rstrip('\n') |
|
pattern = f.next().rstrip('\n') |
|
regex = re.compile(pattern) |
|
tokendef = (token_name, regex) |
|
tokendefs.append(tokendef) |
|
return tokendefs |
|
|
|
|
|
def load_keywords(fpath): |
|
""" |
|
Load the keyword definitions from the file at the path. |
|
|
|
Returns a dict which maps keywords to toktypes. |
|
|
|
The format of the keywords file should be pairs of lines: |
|
- a keywork (e.g. 'true'), |
|
- followed by a line which is the toktype of that keyword (e.g. 'BOOLLIT'). |
|
|
|
Example keywords.txt file: |
|
true |
|
BOOLLIT |
|
false |
|
BOOLLIT |
|
array |
|
ARRAY |
|
return |
|
RETURN |
|
""" |
|
keyword_map = {} |
|
with open(fpath) as f: |
|
for line in f: |
|
keyword = line.rstrip('\n') |
|
toktype = f.next().rstrip('\n') |
|
keyword_map[keyword] = toktype |
|
return keyword_map |
|
|
|
|
|
INDENT_UNIT=4 |
|
def offside_rule(token, previous, indent_level): |
|
""" |
|
Implement the "offside rule" indentation-based scoping mechanism. |
|
See https://en.wikipedia.org/wiki/Off-side_rule |
|
|
|
Args: |
|
- token: the current token |
|
- previous: the previous token |
|
- indent_level: the indent level before encounting this token |
|
|
|
If this is a WS (containing a newline) following a COLON and indentation has increased, |
|
return an INDENT token. |
|
If this is a WS (containing a newline) and indentation has decreased, |
|
return (possibly multiple) DEDENT token(s). |
|
Otherwise, |
|
return the token unmodified. |
|
""" |
|
tokens2 = [token] |
|
|
|
def get_indent_level(text): |
|
""" |
|
Determine the indentation level of the text. |
|
TODO: this implementation is pretty naive. |
|
""" |
|
indentation = text.split('\n')[-1] |
|
assert len(indentation) % INDENT_UNIT == 0 |
|
return len(indentation) / INDENT_UNIT |
|
|
|
assert is_token(token) |
|
assert is_token(previous) or previous is None |
|
text = token_text(token) |
|
if is_toktype(token, 'WS') and '\n' in text: |
|
new_level = get_indent_level(text) |
|
if is_toktype(previous, 'COLON') and new_level > indent_level: |
|
tokens2 = [('token', 'INDENT', text)] |
|
indent_level = new_level |
|
elif new_level < indent_level: |
|
# emit as many DEDENT tokens as needed. |
|
tokens2 = [] |
|
for _ in range(indent_level - new_level): |
|
tokens2.append(('token', 'DEDENT', text)) |
|
indent_level = new_level |
|
return (tokens2, indent_level) |
|
|
|
|
|
def keyword_specialcases(token, keyword_map): |
|
""" |
|
Add a special-case which turns IDENTIFIER tokens into keyword tokens. |
|
|
|
e.g.: |
|
- 'array' changes from an IDENTIFIER token to an ARRAY token. |
|
- 'true' changes from an IDENTIFIER token to a BOOLLIT token. |
|
""" |
|
if is_toktype(token, 'IDENTIFIER'): |
|
text = token_text(token) |
|
if text in keyword_map.keys(): |
|
toktype = keyword_map[text] |
|
return ('token', toktype, text) |
|
return token |
|
|
|
|
|
def tokenize(tokendefs, keyword_map, input, indent=0): |
|
"""Uses tokendefs to tokenize the 'input' string and return a list of tokens""" |
|
tokens = [] |
|
offset = 0 |
|
previous = None |
|
while offset < len(input): |
|
for (token_name, regex) in tokendefs: |
|
m = regex.match(input, offset) |
|
if m is not None: |
|
matched_text = m.group(0) |
|
offset += len(matched_text) |
|
token = ('token', token_name, matched_text) |
|
|
|
# special-case: distinguish keywords from identifiers |
|
token = keyword_specialcases(token, keyword_map) |
|
|
|
# special-case: implement "offside-rule" |
|
(tokens2, indent) = offside_rule(token, previous, indent) |
|
|
|
tokens += tokens2 |
|
previous = token |
|
break |
|
else: |
|
raise Exception("Couldn't tokenize starting at '%s...'" % input[offset:offset+16]) |
|
return tokens |
|
|
|
|
|
# AST node data structure utils: |
|
|
|
def is_ast(ast): |
|
"""Returns whether the arg is an AST node.""" |
|
return isinstance(ast, tuple) and len(ast) > 0 and ast[0] == 'ast' |
|
|
|
|
|
def ast_type(ast): |
|
"""Returns the type of the AST (e.g. 'vardecl', 'statement', etc.)""" |
|
assert is_ast(ast) |
|
return ast[1] |
|
|
|
|
|
def is_asttype(ast, atype): |
|
""" |
|
Returns whether the AST node is of the given type. |
|
atype is e.g. 'vardecl', 'statement', etc. |
|
""" |
|
return ast_type(ast) == atype |
|
|
|
|
|
def ast_children(ast): |
|
"""Return the child nodes of the given AST.""" |
|
assert is_ast(ast) |
|
return ast[2] |
|
|
|
|
|
# Parser helpers: |
|
# See https://en.wikipedia.org/wiki/Extended_Backus%E2%80%93Naur_form#Table_of_symbols |
|
|
|
def alternation(tokens, funcs): |
|
"""Try each of the parsing functions until one succeeds.""" |
|
failure = (None, tokens) |
|
for f in funcs: |
|
(ast, tokens) = f(tokens) |
|
if ast is not None: |
|
return (ast, tokens) |
|
return failure |
|
|
|
|
|
def concatenation(tokens, funcs): |
|
""" |
|
All of the parsing functions must succeed as a whole. |
|
Returns (<array of AST's>, <remaining tokens>). |
|
""" |
|
failure = (None, tokens) |
|
asts = [] |
|
for f in funcs: |
|
(ast, tokens) = f(tokens) |
|
if ast is None: |
|
return failure |
|
asts.append(ast) |
|
return (asts, tokens) |
|
|
|
|
|
# Parsing functions: |
|
|
|
# Note: all parse_x functions share a similar format: |
|
# - Their first argument is the list of tokens. |
|
# - They return a tuple of (<parsed result>, <remaining tokens>). |
|
# - <parsed result> is either an AST node or a token (a terminal). |
|
# - <remaining tokens> is the list of unconsumed tokens. |
|
# - If the parse fails, (None, <tokens>) is returned, where <tokens> |
|
# is the list of tokens which was passed in. |
|
|
|
|
|
# fundecl examples: |
|
# func f(): |
|
# func f(a: int): |
|
# func f() -> int: |
|
# func f(a: int) -> int: |
|
# func f(a: int, b: int): |
|
# func f(a: int, b: int) -> int: |
|
def parse_fundecl(tokens): |
|
""" |
|
Attempts to parse a function declaration. |
|
|
|
Grammar: |
|
fundecl = FUNC WS IDENTIFIER fundeclargs [ fundeclret ] scope |
|
""" |
|
failure = (None, tokens) |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_FUNC, |
|
parse_WS, |
|
parse_IDENTIFIER, |
|
parse_fundeclargs |
|
]) |
|
if asts is None: |
|
return failure |
|
identifier_token = asts[2] |
|
args_ast = asts[3] |
|
(ret_ast, tokens) = parse_fundeclret(tokens) |
|
(scope_ast, tokens) = parse_scope(tokens) |
|
if scope_ast is None: |
|
return failure |
|
ast = ('ast', 'fundecl', [identifier_token, args_ast, ret_ast, scope_ast]) |
|
return (ast, tokens) |
|
|
|
|
|
def parse_fundeclargs(tokens): |
|
""" |
|
Attempts to parse a list of function declaration arguments. |
|
|
|
Grammar: |
|
fundeclargs = OPAREN [ [WS] vardecl { COMMA WS vardecl } [WS] ] CPAREN |
|
""" |
|
failure = (None, tokens) |
|
|
|
def list_of_args(tokens): |
|
""" |
|
Grammar: [WS] vardecl { COMMA WS vardecl } [WS] |
|
Returns (<list of args>, <remaining tokens>) |
|
""" |
|
args = [] |
|
(_, tokens) = parse_WS(tokens) |
|
(vardecl_ast, tokens) = parse_vardecl(tokens) |
|
if vardecl_ast is None: |
|
return (args, tokens) |
|
args.append(vardecl_ast) |
|
|
|
while True: |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_COMMA, |
|
parse_WS, |
|
parse_vardecl |
|
]) |
|
if asts is None: |
|
break |
|
vardecl_ast = asts[2] |
|
args.append(vardecl_ast) |
|
|
|
(_, tokens) = parse_WS(tokens) |
|
return (args, tokens) |
|
|
|
(ast, tokens) = parse_OPAREN(tokens) |
|
if ast is None: |
|
return failure |
|
|
|
(args, tokens) = list_of_args(tokens) |
|
|
|
(ast, tokens) = parse_CPAREN(tokens) |
|
if ast is None: |
|
return failure |
|
|
|
ast = ('ast', 'fundeclargs', args) |
|
return (ast, tokens) |
|
|
|
|
|
def parse_fundeclret(tokens): |
|
""" |
|
Attempts to parse a function declaration return value. |
|
|
|
Grammar: |
|
fundeclret = WS ARROW WS type |
|
""" |
|
failure = (None, tokens) |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_WS, |
|
parse_ARROW, |
|
parse_WS, |
|
parse_type |
|
]) |
|
if asts is None: |
|
return failure |
|
type_ast = asts[3] |
|
ast = ('ast', 'fundeclret', [type_ast]) |
|
return (ast, tokens) |
|
|
|
|
|
def parse_scope(tokens): |
|
""" |
|
Attempts to parse a scoped sequence of statements. |
|
|
|
Grammar: |
|
scope = COLON INDENT statement { WS statement } DEDENT |
|
""" |
|
failure = (None, tokens) |
|
statements = [] |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_COLON, |
|
parse_INDENT, |
|
parse_statement |
|
]) |
|
if asts is None: |
|
return failure |
|
statements.append(asts[2]) |
|
|
|
while True: |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_WS, |
|
parse_statement |
|
]) |
|
if asts is None: |
|
break |
|
statements.append(asts[1]) |
|
|
|
(ast, tokens) = parse_DEDENT(tokens) |
|
if ast is None: |
|
return failure |
|
|
|
ast = ('ast', 'scope', statements) |
|
return (ast, tokens) |
|
|
|
|
|
def parse_vardecl(tokens): |
|
""" |
|
Attempts to parse a vardecl AST node. |
|
|
|
Grammar: |
|
vardecl = IDENTIFIER COLON WS type |
|
""" |
|
failure = (None, tokens) |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_IDENTIFIER, |
|
parse_COLON, |
|
parse_WS, |
|
parse_type |
|
]) |
|
if asts is None: |
|
return failure |
|
identifier_token = asts[0] |
|
type_ast = asts[3] |
|
vardecl_ast = ('ast', 'vardecl', [identifier_token, type_ast]) |
|
return (vardecl_ast, tokens) |
|
|
|
|
|
def parse_type(tokens): |
|
""" |
|
Attemps to parse a type declaration. |
|
|
|
Grammar: |
|
type = POINTER LT type GT |
|
| ARRAY LT type GT |
|
| FUNCTION LT type GT |
|
| ARRAY OBRACKET INTLIT CBRACKET LT type GT |
|
| IDENTIFIER |
|
""" |
|
failure = (None, tokens) |
|
|
|
def parse_type1(tokens, toktype): |
|
""" |
|
Grammar fragments: |
|
POINTER LT type GT |
|
ARRAY LT type GT |
|
|
|
'array<int>' becomes: |
|
('ast', 'type', [ |
|
('token', 'ARRAY', 'array'), |
|
('ast', 'type', [ |
|
('token', 'IDENTIFIER', 'int') |
|
]) |
|
]) |
|
|
|
'pointer<char>' becomes: |
|
('ast', 'type', [ |
|
('token', 'POINTER', 'pointer'), |
|
('ast', 'type', [ |
|
('token', 'IDENTIFIER', 'char') |
|
]) |
|
]) |
|
|
|
'array<pointer<char>>' becomes: |
|
('ast', 'type', [ |
|
('token', 'ARRAY', 'array'), |
|
('ast', 'type', [ |
|
('token', 'POINTER', 'pointer'), |
|
('ast', 'type', [ |
|
('token', 'IDENTIFIER', 'char') |
|
]) |
|
]) |
|
]) |
|
""" |
|
failure = (None, tokens) |
|
(identifier_token, tokens) = parse_terminal(tokens, toktype) |
|
if identifier_token is None: |
|
return failure |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_LT, |
|
parse_type, |
|
parse_GT |
|
]) |
|
if asts is None: |
|
return failure |
|
subtype_ast = asts[1] |
|
ast = ('ast', 'type', [identifier_token, subtype_ast]) |
|
return (ast, tokens) |
|
|
|
def parse_type2(tokens): |
|
""" |
|
Grammar fragment: |
|
ARRAY OBRACKET INTLIT CBRACKET LT type GT |
|
|
|
'array[8]<int>' becomes: |
|
('ast', 'type', [ |
|
('token', 'ARRAY', 'array'), |
|
('token', 'INTLIT', '8'), |
|
('ast', 'type', [ |
|
('token', 'IDENTIFIER', 'int') |
|
]) |
|
]) |
|
""" |
|
failure = (None, tokens) |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_ARRAY, |
|
parse_OBRACKET, |
|
parse_INTLIT, |
|
parse_CBRACKET, |
|
parse_LT, |
|
parse_type, |
|
parse_GT |
|
]) |
|
if asts is None: |
|
return failure |
|
identifier_token = asts[0] |
|
intlit_token = asts[2] |
|
subtype_ast = asts[5] |
|
ast = ('ast', 'type', [identifier_token, intlit_token, subtype_ast]) |
|
return (ast, tokens) |
|
|
|
def parse_type3(tokens): |
|
""" |
|
Grammar fragment: |
|
IDENTIFIER |
|
|
|
'int' becomes: |
|
('ast', 'type', [('token', 'IDENTIFIER', 'int')]) |
|
|
|
'char' becomes: |
|
('ast', 'type', [('token', 'IDENTIFIER', 'char')]) |
|
""" |
|
failure = (None, tokens) |
|
(identifier_token, tokens) = parse_terminal(tokens, 'IDENTIFIER') |
|
if identifier_token is None: |
|
return failure |
|
type_ast = ('ast', 'type', [identifier_token]) |
|
return (type_ast, tokens) |
|
|
|
(ast, tokens) = parse_type1(tokens, 'POINTER') |
|
if ast is not None: |
|
return (ast, tokens) |
|
(ast, tokens) = parse_type1(tokens, 'ARRAY') |
|
if ast is not None: |
|
return (ast, tokens) |
|
(ast, tokens) = parse_type1(tokens, 'FUNCTION') |
|
if ast is not None: |
|
return (ast, tokens) |
|
(ast, tokens) = parse_type2(tokens) |
|
if ast is not None: |
|
return (ast, tokens) |
|
(ast, tokens) = parse_type3(tokens) |
|
if ast is not None: |
|
return (ast, tokens) |
|
return failure |
|
|
|
|
|
def parse_return(tokens): |
|
""" |
|
Attempts to parse a return statement. |
|
|
|
Grammar: |
|
return = RETURN [ WS expr ] |
|
""" |
|
failure = (None, tokens) |
|
(ast, tokens) = parse_terminal(tokens, 'RETURN') |
|
if ast is None: |
|
return failure |
|
return_ast = ('ast', 'return', []) |
|
|
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_WS, |
|
parse_expr |
|
]) |
|
if asts is not None: |
|
expr_ast = asts[1] |
|
return_ast = ('ast', 'return', [expr_ast]) |
|
|
|
return (return_ast, tokens) |
|
|
|
|
|
def parse_expr(tokens): |
|
""" |
|
Attempts to parse an expression. |
|
Expressions result in a value. |
|
|
|
Grammar: |
|
expr = funcall | binary | unary | IDENTIFIER |
|
| INTLIT | FLOATLIT | STRINGLIT | BOOLLIT |
|
""" |
|
failure = (None, tokens) |
|
(ast, tokens) = alternation(tokens, [ |
|
parse_funcall, |
|
parse_binary, |
|
parse_unary, |
|
parse_IDENTIFIER, |
|
parse_INTLIT, |
|
parse_FLOATLIT, |
|
parse_STRINGLIT, |
|
parse_BOOLLIT |
|
]) |
|
if ast is None: |
|
return failure |
|
expr_ast = ('ast', 'expr', [ast]) |
|
return (expr_ast, tokens) |
|
|
|
|
|
def parse_funcall(tokens): |
|
""" |
|
Attempts to parse a funcion call. |
|
|
|
Grammar: |
|
funcall = IDENTIFIER OPAREN [ [WS] expr { COMMA WS expr } [WS] ] CPAREN |
|
""" |
|
failure = (None, tokens) |
|
|
|
def list_of_exprs(tokens): |
|
""" |
|
Grammar: [WS] expr { COMMA WS expr } [WS] |
|
Returns (<list of exprs>, <remaining tokens>) |
|
""" |
|
exprs = [] |
|
(_, tokens) = parse_WS(tokens) |
|
(expr_ast, tokens) = parse_expr(tokens) |
|
if expr_ast is None: |
|
return (exprs, tokens) |
|
exprs.append(expr_ast) |
|
|
|
while True: |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_COMMA, |
|
parse_WS, |
|
parse_expr |
|
]) |
|
if asts is None: |
|
break |
|
expr_ast = asts[2] |
|
exprs.append(expr_ast) |
|
|
|
(_, tokens) = parse_WS(tokens) |
|
return (exprs, tokens) |
|
|
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_IDENTIFIER, |
|
parse_OPAREN |
|
]) |
|
if asts is None: |
|
return failure |
|
identifier_token = asts[0] |
|
|
|
(exprs, tokens) = list_of_exprs(tokens) |
|
|
|
(ast, tokens) = parse_CPAREN(tokens) |
|
if ast is None: |
|
return failure |
|
|
|
funcall_ast = ('ast', 'funcall', [identifier_token] + exprs) |
|
return (funcall_ast, tokens) |
|
|
|
|
|
def parse_declassign(tokens): |
|
""" |
|
Attempts to parse a declaration-assignment statement. |
|
|
|
Grammar: |
|
declassign = vardecl WS EQ WS expr |
|
""" |
|
failure = (None, tokens) |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_vardecl, |
|
parse_WS, |
|
parse_EQ, |
|
parse_WS, |
|
parse_expr |
|
]) |
|
if asts is None: |
|
return failure |
|
vardecl_ast = asts[0] |
|
expr_ast = asts[4] |
|
ast = ('ast', 'declassign', [vardecl_ast, expr_ast]) |
|
return (ast, tokens) |
|
|
|
|
|
def parse_assign(tokens): |
|
""" |
|
Attempts to parse an assignment statement. |
|
|
|
Grammar: |
|
assign = IDENTIFIER WS aoperator WS expr |
|
""" |
|
failure = (None, tokens) |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_IDENTIFIER, |
|
parse_WS, |
|
parse_aoperator, |
|
parse_WS, |
|
parse_expr |
|
]) |
|
if asts is None: |
|
return failure |
|
identifier_token = asts[0] |
|
op_ast = asts[2] |
|
expr_ast = asts[4] |
|
ast = ('ast', 'assign', [identifier_token, op_ast, expr_ast]) |
|
return (ast, tokens) |
|
|
|
|
|
def parse_aoperator(tokens): |
|
""" |
|
Attempts to parse an assignment operator. |
|
|
|
Grammar: |
|
aoperator = PLUSEQ | MINUSEQ | STAREQ | SLASHEQ | PERCENTEQ | EQ |
|
""" |
|
failure = (None, tokens) |
|
(token, tokens) = alternation(tokens, [ |
|
parse_PLUSEQ, |
|
parse_MINUSEQ, |
|
parse_STAREQ, |
|
parse_SLASHEQ, |
|
parse_PERCENTEQ, |
|
parse_EQ |
|
]) |
|
if token is None: |
|
return failure |
|
ast = ('ast', 'aoperator', [token]) |
|
return (ast, tokens) |
|
|
|
|
|
def parse_unary(tokens): |
|
""" |
|
Attempts to parse a unary expression. |
|
|
|
Grammar: |
|
unary = uoperator OPAREN expr CPAREN |
|
""" |
|
failure = (None, tokens) |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_uoperator, |
|
parse_OPAREN, |
|
parse_expr, |
|
parse_CPAREN |
|
]) |
|
if asts is None: |
|
return failure |
|
op = asts[0] |
|
expr = asts[2] |
|
ast = ('ast', 'unary', [op, expr]) |
|
return (ast, tokens) |
|
|
|
|
|
def parse_uoperator(tokens): |
|
""" |
|
Attempts to parse a unary operator. |
|
|
|
Grammar: |
|
uoperator = MINUS | AMP | STAR | BANG |
|
""" |
|
failure = (None, tokens) |
|
(token, tokens) = alternation(tokens, [ |
|
parse_MINUS, |
|
parse_AMP, |
|
parse_STAR, |
|
parse_BANG |
|
]) |
|
if token is None: |
|
return failure |
|
ast = ('ast', 'uoperator', [token]) |
|
return (ast, tokens) |
|
|
|
|
|
def parse_binary(tokens): |
|
""" |
|
Attempts to parse a binary expression. |
|
|
|
Grammar: |
|
binary = OPAREN expr WS boperator WS expr CPAREN |
|
""" |
|
failure = (None, tokens) |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_OPAREN, |
|
parse_expr, |
|
parse_WS, |
|
parse_boperator, |
|
parse_WS, |
|
parse_expr, |
|
parse_CPAREN |
|
]) |
|
if asts is None: |
|
return failure |
|
expr1 = asts[1] |
|
op = asts[3] |
|
expr2 = asts[5] |
|
ast = ('ast', 'binary', [expr1, op, expr2]) |
|
return (ast, tokens) |
|
|
|
|
|
def parse_boperator(tokens): |
|
""" |
|
Attempts to parse a binary operator. |
|
|
|
Grammar: |
|
boperator = PLUS | MINUS | STAR | SLASH | PERCENT |
|
| LT | LTEQ | GT | GTEQ | EQEQ | BANGEQ |
|
| AMPAMP | BARBAR | BANG |
|
| AMP | BAR | LTLT | GTGT | TILDE | CARAT |
|
""" |
|
failure = (None, tokens) |
|
(token, tokens) = alternation(tokens, [ |
|
parse_PLUS, |
|
parse_MINUS, |
|
parse_STAR, |
|
parse_SLASH, |
|
parse_LT, |
|
parse_LTEQ, |
|
parse_GT, |
|
parse_GTEQ, |
|
parse_EQEQ, |
|
parse_BANGEQ, |
|
parse_AMPAMP, |
|
parse_BARBAR, |
|
parse_BANG, |
|
parse_AMP, |
|
parse_BAR, |
|
parse_LTLT, |
|
parse_GTGT, |
|
parse_TILDE, |
|
parse_CARAT, |
|
]) |
|
if token is None: |
|
return failure |
|
ast = ('ast', 'boperator', [token]) |
|
return (ast, tokens) |
|
|
|
|
|
def parse_statement(tokens): |
|
""" |
|
Attempts to parse a statement. |
|
|
|
Grammar: |
|
statement = declassign | assign | fundecl | vardecl | return | funcall | if | scope |
|
""" |
|
failure = (None, tokens) |
|
(ast, tokens) = alternation(tokens, [ |
|
parse_declassign, |
|
parse_assign, |
|
parse_fundecl, |
|
parse_vardecl, |
|
parse_return, |
|
parse_funcall, |
|
parse_if, |
|
parse_scope |
|
]) |
|
if ast is None: |
|
return failure |
|
statement_ast = ('ast', 'statement', [ast]) |
|
return (statement_ast, tokens) |
|
|
|
|
|
def parse_if(tokens): |
|
""" |
|
Attempts to parse an if statement. |
|
|
|
Grammar: |
|
if = IF WS expr scope { elif } [else] |
|
""" |
|
failure = (None, tokens) |
|
|
|
def parse_elif(tokens): |
|
""" |
|
Attempts to parse an elif statement. |
|
|
|
Grammar: |
|
elif = ELIF WS expr scope |
|
""" |
|
failure = (None, tokens) |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_ELIF, |
|
parse_WS, |
|
parse_expr, |
|
parse_scope |
|
]) |
|
if asts is None: |
|
return failure |
|
ast = ('ast', 'elif', [asts[2], asts[3]]) |
|
return (ast, tokens) |
|
|
|
def parse_else(tokens): |
|
""" |
|
Attempts to parse an else statement. |
|
|
|
Grammar: |
|
else = ELSE scope |
|
""" |
|
failure = (None, tokens) |
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_ELSE, |
|
parse_scope |
|
]) |
|
if asts is None: |
|
return failure |
|
ast = ('ast', 'else', [asts[1]]) |
|
return (ast, tokens) |
|
|
|
(asts, tokens) = concatenation(tokens, [ |
|
parse_IF, |
|
parse_WS, |
|
parse_expr, |
|
parse_scope |
|
]) |
|
if asts is None: |
|
return failure |
|
sub_asts = [asts[2], asts[3]] |
|
|
|
elifs = [] |
|
while True: |
|
(ast, tokens) = parse_elif(tokens) |
|
if ast is None: |
|
break |
|
elifs.append(ast) |
|
sub_asts += elifs |
|
|
|
(else_ast, tokens) = parse_else(tokens) |
|
if else_ast is not None: |
|
sub_asts.append(else_ast) |
|
|
|
if_ast = ('ast', 'if', sub_asts) |
|
return (if_ast, tokens) |
|
|
|
|
|
def parse_program(tokens): |
|
""" |
|
Attempts to parse a program. |
|
|
|
Grammar: |
|
statement [WS] { statement [WS] } |
|
""" |
|
failure = (None, tokens) |
|
statement_asts = [] |
|
|
|
(ast, tokens) = parse_statement(tokens) |
|
if ast is None: |
|
return failure |
|
statement_asts.append(ast) |
|
(_, tokens) = parse_WS(tokens) |
|
|
|
while True: |
|
(ast, tokens) = parse_statement(tokens) |
|
if ast is None: |
|
break |
|
statement_asts.append(ast) |
|
(_, tokens) = parse_WS(tokens) |
|
|
|
ast = ('ast', 'program', statement_asts) |
|
return (ast, tokens) |
|
|
|
|
|
def parse(tokens): |
|
"""Attempts to parse the tokens into an AST.""" |
|
(ast, tokens) = parse_program(tokens) |
|
if ast is None: |
|
raise Exception("Couldn't parse starting at %s" % tokens[:8]) |
|
if len(tokens) > 0: |
|
raise Exception("Leftover tokens: %s" % tokens) |
|
return ast |
|
|
|
|
|
# Terminal parsing: parsing individual tokens. |
|
|
|
def parse_terminal(tokens, toktype): |
|
""" |
|
Attempts to parse a terminal node of type _toktype_. |
|
Note that the parsed result of a terminal is the token itself, not an AST node. |
|
""" |
|
failure = (None, tokens) |
|
if len(tokens) > 0 and is_toktype(tokens[0], toktype): |
|
return (tokens[0], tokens[1:]) |
|
else: |
|
return failure |
|
|
|
def parse_FUNC(tokens): |
|
return parse_terminal(tokens, 'FUNC') |
|
|
|
def parse_ARRAY(tokens): |
|
return parse_terminal(tokens, 'ARRAY') |
|
|
|
def parse_COMMA(tokens): |
|
return parse_terminal(tokens, 'COMMA') |
|
|
|
def parse_WS(tokens): |
|
return parse_terminal(tokens, 'WS') |
|
|
|
def parse_ARROW(tokens): |
|
return parse_terminal(tokens, 'ARROW') |
|
|
|
def parse_IDENTIFIER(tokens): |
|
return parse_terminal(tokens, 'IDENTIFIER') |
|
|
|
def parse_COLON(tokens): |
|
return parse_terminal(tokens, 'COLON') |
|
|
|
def parse_INTLIT(tokens): |
|
return parse_terminal(tokens, 'INTLIT') |
|
|
|
def parse_FLOATLIT(tokens): |
|
return parse_terminal(tokens, 'FLOATLIT') |
|
|
|
def parse_STRINGLIT(tokens): |
|
return parse_terminal(tokens, 'STRINGLIT') |
|
|
|
def parse_BOOLLIT(tokens): |
|
return parse_terminal(tokens, 'BOOLLIT') |
|
|
|
def parse_OPAREN(tokens): |
|
return parse_terminal(tokens, 'OPAREN') |
|
|
|
def parse_CPAREN(tokens): |
|
return parse_terminal(tokens, 'CPAREN') |
|
|
|
def parse_OBRACKET(tokens): |
|
return parse_terminal(tokens, 'OBRACKET') |
|
|
|
def parse_CBRACKET(tokens): |
|
return parse_terminal(tokens, 'CBRACKET') |
|
|
|
def parse_MINUS(tokens): |
|
return parse_terminal(tokens, 'MINUS') |
|
|
|
def parse_AMP(tokens): |
|
return parse_terminal(tokens, 'AMP') |
|
|
|
def parse_STAR(tokens): |
|
return parse_terminal(tokens, 'STAR') |
|
|
|
def parse_BANG(tokens): |
|
return parse_terminal(tokens, 'BANG') |
|
|
|
def parse_PLUS(tokens): |
|
return parse_terminal(tokens, 'PLUS') |
|
|
|
def parse_SLASH(tokens): |
|
return parse_terminal(tokens, 'SLASH') |
|
|
|
def parse_LT(tokens): |
|
return parse_terminal(tokens, 'LT') |
|
|
|
def parse_LTEQ(tokens): |
|
return parse_terminal(tokens, 'LTEQ') |
|
|
|
def parse_GT(tokens): |
|
return parse_terminal(tokens, 'GT') |
|
|
|
def parse_GTEQ(tokens): |
|
return parse_terminal(tokens, 'GTEQ') |
|
|
|
def parse_EQ(tokens): |
|
return parse_terminal(tokens, 'EQ') |
|
|
|
def parse_EQEQ(tokens): |
|
return parse_terminal(tokens, 'EQEQ') |
|
|
|
def parse_BANGEQ(tokens): |
|
return parse_terminal(tokens, 'BANGEQ') |
|
|
|
def parse_AMPAMP(tokens): |
|
return parse_terminal(tokens, 'AMPAMP') |
|
|
|
def parse_BARBAR(tokens): |
|
return parse_terminal(tokens, 'BARBAR') |
|
|
|
def parse_BAR(tokens): |
|
return parse_terminal(tokens, 'BAR') |
|
|
|
def parse_LTLT(tokens): |
|
return parse_terminal(tokens, 'LTLT') |
|
|
|
def parse_GTGT(tokens): |
|
# So, I'm not super happy about this, but here's why this has to be special-cased: |
|
# If we just treated this like all of the other tokenization and parsing functions, |
|
# we would have a token for '>>', and this would be a one-liner: |
|
# return parse_terminal(tokens, 'GTGT') |
|
# However, that would break `pointer<pointer<char>>`, because the last two |
|
# characters would be a single GTGT token, rathe than two separate GT tokens. |
|
# So we have to implement parse_GTGT by looking for two distinct GT tokens. |
|
failure = (None, tokens) |
|
(gt_tokens, tokens) = concatenation(tokens, [parse_GT, parse_GT]) |
|
if gt_tokens is None: |
|
return failure |
|
token = ('token', 'GTGT', '>>') |
|
|
|
def parse_TILDE(tokens): |
|
return parse_terminal(tokens, 'TILDE') |
|
|
|
def parse_CARAT(tokens): |
|
return parse_terminal(tokens, 'CARAT') |
|
|
|
def parse_PLUSEQ(tokens): |
|
return parse_terminal(tokens, 'PLUSEQ') |
|
|
|
def parse_MINUSEQ(tokens): |
|
return parse_terminal(tokens, 'MINUSEQ') |
|
|
|
def parse_STAREQ(tokens): |
|
return parse_terminal(tokens, 'STAREQ') |
|
|
|
def parse_SLASHEQ(tokens): |
|
return parse_terminal(tokens, 'SLASHEQ') |
|
|
|
def parse_PERCENTEQ(tokens): |
|
return parse_terminal(tokens, 'PERCENTEQ') |
|
|
|
def parse_INDENT(tokens): |
|
return parse_terminal(tokens, 'INDENT') |
|
|
|
def parse_DEDENT(tokens): |
|
return parse_terminal(tokens, 'DEDENT') |
|
|
|
def parse_IF(tokens): |
|
return parse_terminal(tokens, 'IF') |
|
|
|
def parse_ELIF(tokens): |
|
return parse_terminal(tokens, 'ELIF') |
|
|
|
def parse_ELSE(tokens): |
|
return parse_terminal(tokens, 'ELSE') |
|
|
|
|
|
# Output generator implementation: |
|
|
|
# Note: all generate_x functions share a similar format: |
|
# - Their first argument is as AST. |
|
# - They return a string (of C code). |
|
# If the passed AST is valid, then errors are not possible. |
|
# Exceptions are thrown otherwise. |
|
|
|
def generate_type(ast, identifier=""): |
|
""" |
|
Generate a C type declaration. |
|
|
|
If identifier is not given, this is an abstract type declaration (e.g. a cast). |
|
""" |
|
def did_switch_direction(previous, current): |
|
""" |
|
Detect a change in 'direction' while intepreting the type stack. |
|
|
|
C type declaration operator precedence requires that we "go right" first. |
|
i.e. 'int *foo[]' is "array of pointer to int", not "pointer to array of int". |
|
In order to express "pointer to array of int", we have to use parenthesis |
|
to change overcome operator precedence, i.e. 'int (*foo)[]'. |
|
Any time we need to "change direction", we need to wrap in parenthesis. |
|
See http://unixwiz.net/techtips/reading-cdecl.html |
|
""" |
|
lefts = ['pointer'] |
|
rights = ['array', 'function'] |
|
return (previous in lefts and current in rights) \ |
|
or (previous in rights and current in lefts) |
|
|
|
assert is_asttype(ast, 'type') |
|
types = type_ast_as_list(ast) |
|
output = identifier |
|
previous = None |
|
for t in types[:-1]: |
|
if t == 'pointer': |
|
if did_switch_direction(previous, t): |
|
output = "*(%s)" % output |
|
else: |
|
output = "*%s" % output |
|
elif t == 'array': |
|
if did_switch_direction(previous, t): |
|
output = "(%s)[]" % output |
|
else: |
|
output = "%s[]" % output |
|
elif t == 'function': |
|
if did_switch_direction(previous, t): |
|
output = "(%s)()" % output |
|
else: |
|
output = "%s()" % output |
|
elif t.startswith('array:'): |
|
array_size = int(t.split(':')[1]) |
|
if did_switch_direction(previous, 'array'): |
|
output = "(%s)[%i]" % (output, array_size) |
|
else: |
|
output = "%s[%i]" % (output, array_size) |
|
else: |
|
raise Exception("generate_type: unexpected type '%s'." % t) |
|
|
|
if t.startswith('array:'): |
|
previous = 'array' |
|
else: |
|
previous = t |
|
|
|
base_type = types[-1] |
|
if identifier == "": |
|
output = "%s%s" % (base_type, output) |
|
else: |
|
output = "%s %s" % (base_type, output) |
|
return output |
|
|
|
|
|
def type_ast_as_list(ast): |
|
""" |
|
Return the type AST as a list of types. |
|
|
|
e.g. |
|
- 'int' results in ['int'] |
|
- 'pointer<array<char>>' results in ['pointer', 'array', 'char'] |
|
- 'array[8]<int>' results in ['array:8', 'int'] |
|
""" |
|
assert is_asttype(ast, 'type') |
|
children = ast_children(ast) |
|
if len(children) == 1: |
|
# if there is only one child, it is the "base" type (e.g. int, char, etc). |
|
assert is_token(children[0]) |
|
token = children[0] |
|
assert is_toktype(token, 'IDENTIFIER') |
|
# return e.g. ['int'], ['char'], etc. |
|
return [token_text(token)] |
|
elif len(children) == 2: |
|
# if there are 2 children, this is either array or pointer or function. |
|
assert is_toktype(children[0], 'ARRAY') \ |
|
or is_toktype(children[0], 'POINTER') \ |
|
or is_toktype(children[0], 'FUNCTION') |
|
token = children[0] |
|
assert is_ast(children[1]) |
|
sub_ast = children[1] |
|
assert is_asttype(ast, 'type') |
|
# return e.g. ['array', <recursive call>] |
|
return [token_text(token)] + type_ast_as_list(sub_ast) |
|
elif len(children) == 3: |
|
# if there are three children, this is a dimensioned array (e.g. array[8]). |
|
assert is_toktype(children[0], 'ARRAY') |
|
assert is_toktype(children[1], 'INTLIT') |
|
intlit_token = children[1] |
|
array_size = int(token_text(intlit_token)) |
|
assert is_ast(children[2]) |
|
sub_ast = children[2] |
|
assert is_asttype(sub_ast, 'type') |
|
# return e.g. ['array:8', <recursive call>] |
|
return ["array:%s" % (array_size)] + type_ast_as_list(sub_ast) |
|
else: |
|
raise Exception("type_ast_as_list: type AST node with more than 3 children!") |
|
|
|
|
|
def generate_vardecl(ast): |
|
""" |
|
Generate a C variable declaration. |
|
|
|
Note: we don't append the trailing ';' here because this is also used for function args. |
|
""" |
|
assert is_asttype(ast, 'vardecl') |
|
children = ast_children(ast) |
|
|
|
identifier_token = children[0] |
|
assert is_toktype(identifier_token, 'IDENTIFIER') |
|
var_name = token_text(identifier_token) |
|
|
|
type_ast = children[1] |
|
assert is_asttype(type_ast, 'type') |
|
|
|
# e.g. 'int a', 'char* b', 'int c[8]', 'char** d', 'int* d[3]', etc. |
|
return "%s" % generate_type(type_ast, var_name) |
|
|
|
|
|
def generate_fundecl(ast): |
|
"""Generate a C function declaration.""" |
|
assert is_asttype(ast, 'fundecl') |
|
|
|
identifier_token = ast_children(ast)[0] |
|
assert is_toktype(identifier_token, 'IDENTIFIER') |
|
identifier = token_text(identifier_token) |
|
|
|
fundeclargs_ast = ast_children(ast)[1] |
|
assert is_asttype(fundeclargs_ast, 'fundeclargs') |
|
arg_asts = ast_children(fundeclargs_ast) |
|
args = [] |
|
for vardecl_ast in arg_asts: |
|
arg = generate_vardecl(vardecl_ast) |
|
args.append(arg) |
|
args_output = ", ".join(args) |
|
|
|
fundeclret_ast = ast_children(ast)[2] |
|
if fundeclret_ast is None: |
|
ret = "void" |
|
else: |
|
ret_ast = ast_children(fundeclret_ast)[0] |
|
ret = generate_type(ret_ast) |
|
|
|
scope_ast = ast_children(ast)[3] |
|
scope = generate_scope(scope_ast) |
|
|
|
output = "%s %s(%s) %s" % (ret, identifier, args_output, scope) |
|
return output |
|
|
|
|
|
def indent_text(text): |
|
"""Add indentation to the given block of text.""" |
|
indent = " " * INDENT_UNIT |
|
indented_lines = [indent + line for line in text.splitlines()] |
|
return "\n".join(indented_lines) |
|
|
|
|
|
def generate_scope(ast): |
|
"""Generate a C scope.""" |
|
assert is_ast(ast) |
|
assert is_asttype(ast, 'scope') |
|
output = "{" |
|
for statement_ast in ast_children(ast): |
|
statement = generate_statement(statement_ast) |
|
output += ("\n" + indent_text(statement)) |
|
line = "\n}" |
|
output += line |
|
return output |
|
|
|
|
|
def generate_funcall(ast): |
|
"""Generate a C function call.""" |
|
assert is_asttype(ast, 'funcall') |
|
children = ast_children(ast) |
|
identifier_token = children[0] |
|
assert is_toktype(identifier_token, 'IDENTIFIER') |
|
identifier = token_text(identifier_token) |
|
exprs = [generate_expr(a) for a in children[1:]] |
|
return "%s(%s)" % (identifier, ", ".join(exprs)) |
|
|
|
|
|
def generate_expr(ast): |
|
"""Generate a C expression.""" |
|
assert is_asttype(ast, 'expr') |
|
child = ast_children(ast)[0] |
|
if is_ast(child): |
|
if is_asttype(child, 'funcall'): |
|
return generate_funcall(child) |
|
if is_asttype(child, 'binary'): |
|
return generate_binary(child) |
|
elif is_asttype(child, 'unary'): |
|
return generate_unary(child) |
|
elif is_token(child): |
|
token = child |
|
if is_toktype(token, 'INTLIT') \ |
|
or is_toktype(token, 'FLOATLIT') \ |
|
or is_toktype(token, 'STRINGLIT') \ |
|
or is_toktype(token, 'BOOLLIT') \ |
|
or is_toktype(token, 'IDENTIFIER') \ |
|
: |
|
output = token_text(token) |
|
return output |
|
else: |
|
raise Exception("generate_expr: unexpcted ast %s" % (ast,)) |
|
|
|
|
|
def generate_return(ast): |
|
"""Generate a C return statement.""" |
|
assert is_asttype(ast, 'return') |
|
children = ast_children(ast) |
|
if len(children) == 0: |
|
return "return;" |
|
else: |
|
expr_ast = children[0] |
|
return "return %s;" % generate_expr(expr_ast) |
|
|
|
|
|
def generate_declassign(ast): |
|
"""Generate a C declaration-assignment statement.""" |
|
assert is_asttype(ast, 'declassign') |
|
children = ast_children(ast) |
|
vardecl_ast = children[0] |
|
assert is_asttype(vardecl_ast, 'vardecl') |
|
vardecl = generate_vardecl(vardecl_ast) |
|
expr_ast = children[1] |
|
assert is_asttype(expr_ast, 'expr') |
|
expr = generate_expr(expr_ast) |
|
return "%s = %s;" % (vardecl, expr) |
|
|
|
|
|
def generate_assign(ast): |
|
"""Generate a C assignment statement.""" |
|
assert is_asttype(ast, 'assign') |
|
children = ast_children(ast) |
|
|
|
identifier_token = children[0] |
|
assert is_toktype(identifier_token, 'IDENTIFIER') |
|
identifier = token_text(identifier_token) |
|
|
|
op_ast = children[1] |
|
assert is_asttype(op_ast, 'aoperator') |
|
op_token = ast_children(op_ast)[0] |
|
assert is_token(op_token) |
|
op = token_text(op_token) |
|
|
|
expr_ast = children[2] |
|
assert is_asttype(expr_ast, 'expr') |
|
expr = generate_expr(expr_ast) |
|
return "%s %s %s;" % (identifier, op, expr) |
|
|
|
|
|
def generate_unary(ast): |
|
"""Generate a unary-operator C statement.""" |
|
assert is_asttype(ast, 'unary') |
|
children = ast_children(ast) |
|
op_ast = children[0] |
|
assert is_asttype(op_ast, 'uoperator') |
|
op_token = ast_children(op_ast)[0] |
|
assert is_token(op_token) |
|
op = token_text(op_token) |
|
expr_ast = children[1] |
|
assert is_asttype(expr_ast, 'expr') |
|
expr = generate_expr(expr_ast) |
|
return "%s(%s)" % (op, expr) |
|
|
|
|
|
def generate_binary(ast): |
|
"""Generate a binary-operator C statement.""" |
|
assert is_asttype(ast, 'binary') |
|
children = ast_children(ast) |
|
|
|
expr1_ast = children[0] |
|
assert is_asttype(expr1_ast, 'expr') |
|
expr1 = generate_expr(expr1_ast) |
|
|
|
op_ast = children[1] |
|
assert is_asttype(op_ast, 'boperator') |
|
op_token = ast_children(op_ast)[0] |
|
assert is_token(op_token) |
|
op = token_text(op_token) |
|
|
|
expr2_ast = children[2] |
|
assert is_asttype(expr2_ast, 'expr') |
|
expr2 = generate_expr(expr2_ast) |
|
|
|
return "(%s %s %s)" % (expr1, op, expr2) |
|
|
|
|
|
def generate_if(ast): |
|
"""Generate a C if statement""" |
|
assert is_asttype(ast, 'if') |
|
children = ast_children(ast) |
|
|
|
predicate_ast = children[0] |
|
assert is_asttype(predicate_ast, 'expr') |
|
predicate = generate_expr(predicate_ast) |
|
output = "if (%s) " % (predicate) |
|
|
|
scope_ast = children[1] |
|
assert is_asttype(scope_ast, 'scope') |
|
output += generate_scope(scope_ast) |
|
|
|
for ast in children[2:]: |
|
assert is_ast(ast) |
|
if ast_type(ast) == 'elif': |
|
elif_children = ast_children(ast) |
|
|
|
predicate_ast = elif_children[0] |
|
assert is_asttype(predicate_ast, 'expr') |
|
predicate = generate_expr(predicate_ast) |
|
output += " else if (%s) " % (predicate) |
|
|
|
scope_ast = elif_children[1] |
|
assert is_asttype(scope_ast, 'scope') |
|
output += generate_scope(scope_ast) |
|
|
|
elif ast_type(ast) == 'else': |
|
else_children = ast_children(ast) |
|
|
|
scope_ast = else_children[0] |
|
assert is_asttype(scope_ast, 'scope') |
|
output += " else " + generate_scope(scope_ast) |
|
|
|
else: |
|
raise Exception("generate_if: unexpected AST type %s" % ast_type(ast)) |
|
|
|
return output |
|
|
|
|
|
def generate_statement(ast): |
|
"""Generate a C statement.""" |
|
assert is_asttype(ast, 'statement') |
|
sub_ast = ast_children(ast)[0] |
|
assert is_ast(sub_ast) |
|
if is_asttype(sub_ast, 'declassign'): |
|
return generate_declassign(sub_ast) |
|
elif is_asttype(sub_ast, 'assign'): |
|
return generate_assign(sub_ast) |
|
elif is_asttype(sub_ast, 'fundecl'): |
|
return generate_fundecl(sub_ast) |
|
elif is_asttype(sub_ast, 'vardecl'): |
|
return generate_vardecl(sub_ast) + ';' |
|
elif is_asttype(sub_ast, 'return'): |
|
return generate_return(sub_ast) |
|
elif is_asttype(sub_ast, 'funcall'): |
|
return generate_funcall(sub_ast) + ';' |
|
elif is_asttype(sub_ast, 'if'): |
|
return generate_if(sub_ast) |
|
elif is_asttype(sub_ast, 'scope'): |
|
return generate_scope(sub_ast) |
|
else: |
|
raise Exception("generate_statement: don't know how to generate %s" % sub_ast) |
|
|
|
|
|
def generate_program(ast): |
|
"""Generate a C program.""" |
|
assert is_asttype(ast, 'program') |
|
outputs = [] |
|
for statement_ast in ast_children(ast): |
|
statement = generate_statement(statement_ast) |
|
if statement.endswith('\n}'): |
|
statement += '\n' # extra newline after function bodies. |
|
outputs.append(statement) |
|
output = '\n'.join(outputs) |
|
if not output.endswith('\n'): |
|
output += '\n' # always include a trailing newline. |
|
return output |
|
|
|
|
|
def generate(ast): |
|
"""Generate C code from the given AST.""" |
|
return generate_program(ast) |
|
|
|
|
|
if __name__ == "__main__": |
|
import sys |
|
import pprint |
|
tdefs = load_tokendefs("tokendefs.txt") |
|
keywords = load_keywords("keywords.txt") |
|
input = None |
|
if len(sys.argv) > 1 and not sys.argv[-1].startswith('--'): |
|
input = open(sys.argv[-1]).read() |
|
else: |
|
input = sys.stdin.read() |
|
tokens = tokenize(tdefs, keywords, input) |
|
if '--tokens' in sys.argv: |
|
pprint.pprint(tokens) |
|
sys.exit(0) |
|
ast = parse(tokens) |
|
if '--ast' in sys.argv: |
|
pprint.pprint(ast) |
|
sys.exit(0) |
|
output = generate(ast) |
|
sys.stdout.write(output) |