v8/tools/js2c.py
domenic b40071974e Fix error messages for extra files in js2c.py
https://codereview.chromium.org/1117663002 contained a copy-paste
error; this is a follow-up fixing that so that error messages now
contain the correct filename when using the --extra option to js2c.py.

R=yangguo@chromium.org, jochen@chromium.org

Review URL: https://codereview.chromium.org/1120593002

Cr-Commit-Position: refs/heads/master@{#28158}
2015-04-30 11:24:15 +00:00

603 lines
18 KiB
Python
Executable File

#!/usr/bin/env python
#
# Copyright 2012 the V8 project authors. All rights reserved.
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met:
#
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above
# copyright notice, this list of conditions and the following
# disclaimer in the documentation and/or other materials provided
# with the distribution.
# * Neither the name of Google Inc. nor the names of its
# contributors may be used to endorse or promote products derived
# from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
# This is a utility for converting JavaScript source code into C-style
# char arrays. It is used for embedded JavaScript code in the V8
# library.
import os, re, sys, string
import argparse
import jsmin
import bz2
import textwrap
class Error(Exception):
def __init__(self, msg):
Exception.__init__(self, msg)
def ToCArray(byte_sequence):
result = []
for chr in byte_sequence:
result.append(str(ord(chr)))
joined = ", ".join(result)
return textwrap.fill(joined, 80)
def RemoveCommentsAndTrailingWhitespace(lines):
lines = re.sub(r'//.*\n', '\n', lines) # end-of-line comments
lines = re.sub(re.compile(r'/\*.*?\*/', re.DOTALL), '', lines) # comments.
lines = re.sub(r'\s+\n+', '\n', lines) # trailing whitespace
return lines
def ReadFile(filename):
file = open(filename, "rt")
try:
lines = file.read()
finally:
file.close()
return lines
EVAL_PATTERN = re.compile(r'\beval\s*\(')
WITH_PATTERN = re.compile(r'\bwith\s*\(')
INVALID_ERROR_MESSAGE_PATTERN = re.compile(
r'Make(?!Generic)\w*Error\(([kA-Z]\w+)')
NEW_ERROR_PATTERN = re.compile(r'new \$\w*Error\((?!\))')
def Validate(lines):
# Because of simplified context setup, eval and with is not
# allowed in the natives files.
if EVAL_PATTERN.search(lines):
raise Error("Eval disallowed in natives.")
if WITH_PATTERN.search(lines):
raise Error("With statements disallowed in natives.")
invalid_error = INVALID_ERROR_MESSAGE_PATTERN.search(lines)
if invalid_error:
raise Error("Unknown error message template '%s'" % invalid_error.group(1))
if NEW_ERROR_PATTERN.search(lines):
raise Error("Error constructed without message template.")
# Pass lines through unchanged.
return lines
def ExpandConstants(lines, constants):
for key, value in constants:
lines = key.sub(str(value), lines)
return lines
def ExpandMacroDefinition(lines, pos, name_pattern, macro, expander):
pattern_match = name_pattern.search(lines, pos)
while pattern_match is not None:
# Scan over the arguments
height = 1
start = pattern_match.start()
end = pattern_match.end()
assert lines[end - 1] == '('
last_match = end
arg_index = [0] # Wrap state into array, to work around Python "scoping"
mapping = { }
def add_arg(str):
# Remember to expand recursively in the arguments
replacement = expander(str.strip())
mapping[macro.args[arg_index[0]]] = replacement
arg_index[0] += 1
while end < len(lines) and height > 0:
# We don't count commas at higher nesting levels.
if lines[end] == ',' and height == 1:
add_arg(lines[last_match:end])
last_match = end + 1
elif lines[end] in ['(', '{', '[']:
height = height + 1
elif lines[end] in [')', '}', ']']:
height = height - 1
end = end + 1
# Remember to add the last match.
add_arg(lines[last_match:end-1])
result = macro.expand(mapping)
# Replace the occurrence of the macro with the expansion
lines = lines[:start] + result + lines[end:]
pattern_match = name_pattern.search(lines, start + len(result))
return lines
def ExpandMacros(lines, macros):
# We allow macros to depend on the previously declared macros, but
# we don't allow self-dependecies or recursion.
for name_pattern, macro in reversed(macros):
def expander(s):
return ExpandMacros(s, macros)
lines = ExpandMacroDefinition(lines, 0, name_pattern, macro, expander)
return lines
class TextMacro:
def __init__(self, args, body):
self.args = args
self.body = body
def expand(self, mapping):
result = self.body
for key, value in mapping.items():
result = result.replace(key, value)
return result
class PythonMacro:
def __init__(self, args, fun):
self.args = args
self.fun = fun
def expand(self, mapping):
args = []
for arg in self.args:
args.append(mapping[arg])
return str(self.fun(*args))
CONST_PATTERN = re.compile(r'^define\s+([a-zA-Z0-9_]+)\s*=\s*([^;]*);$')
MACRO_PATTERN = re.compile(r'^macro\s+([a-zA-Z0-9_]+)\s*\(([^)]*)\)\s*=\s*([^;]*);$')
PYTHON_MACRO_PATTERN = re.compile(r'^python\s+macro\s+([a-zA-Z0-9_]+)\s*\(([^)]*)\)\s*=\s*([^;]*);$')
def ReadMacros(lines):
constants = []
macros = []
for line in lines.split('\n'):
hash = line.find('#')
if hash != -1: line = line[:hash]
line = line.strip()
if len(line) is 0: continue
const_match = CONST_PATTERN.match(line)
if const_match:
name = const_match.group(1)
value = const_match.group(2).strip()
constants.append((re.compile("\\b%s\\b" % name), value))
else:
macro_match = MACRO_PATTERN.match(line)
if macro_match:
name = macro_match.group(1)
args = [match.strip() for match in macro_match.group(2).split(',')]
body = macro_match.group(3).strip()
macros.append((re.compile("\\b%s\\(" % name), TextMacro(args, body)))
else:
python_match = PYTHON_MACRO_PATTERN.match(line)
if python_match:
name = python_match.group(1)
args = [match.strip() for match in python_match.group(2).split(',')]
body = python_match.group(3).strip()
fun = eval("lambda " + ",".join(args) + ': ' + body)
macros.append((re.compile("\\b%s\\(" % name), PythonMacro(args, fun)))
else:
raise Error("Illegal line: " + line)
return (constants, macros)
TEMPLATE_PATTERN = re.compile(r'^\s+T\(([A-Z][a-zA-Z]*),')
def ReadMessageTemplates(lines):
templates = []
index = 0
for line in lines.split('\n'):
template_match = TEMPLATE_PATTERN.match(line)
if template_match:
name = "k%s" % template_match.group(1)
value = index
index = index + 1
templates.append((re.compile("\\b%s\\b" % name), value))
return templates
INLINE_MACRO_PATTERN = re.compile(r'macro\s+([a-zA-Z0-9_]+)\s*\(([^)]*)\)\s*\n')
INLINE_MACRO_END_PATTERN = re.compile(r'endmacro\s*\n')
def ExpandInlineMacros(lines):
pos = 0
while True:
macro_match = INLINE_MACRO_PATTERN.search(lines, pos)
if macro_match is None:
# no more macros
return lines
name = macro_match.group(1)
args = [match.strip() for match in macro_match.group(2).split(',')]
end_macro_match = INLINE_MACRO_END_PATTERN.search(lines, macro_match.end());
if end_macro_match is None:
raise Error("Macro %s unclosed" % name)
body = lines[macro_match.end():end_macro_match.start()]
# remove macro definition
lines = lines[:macro_match.start()] + lines[end_macro_match.end():]
name_pattern = re.compile("\\b%s\\(" % name)
macro = TextMacro(args, body)
# advance position to where the macro defintion was
pos = macro_match.start()
def non_expander(s):
return s
lines = ExpandMacroDefinition(lines, pos, name_pattern, macro, non_expander)
INLINE_CONSTANT_PATTERN = re.compile(r'define\s+([a-zA-Z0-9_]+)\s*=\s*([^;\n]+)[;\n]')
def ExpandInlineConstants(lines):
pos = 0
while True:
const_match = INLINE_CONSTANT_PATTERN.search(lines, pos)
if const_match is None:
# no more constants
return lines
name = const_match.group(1)
replacement = const_match.group(2)
name_pattern = re.compile("\\b%s\\b" % name)
# remove constant definition and replace
lines = (lines[:const_match.start()] +
re.sub(name_pattern, replacement, lines[const_match.end():]))
# advance position to where the constant defintion was
pos = const_match.start()
HEADER_TEMPLATE = """\
// Copyright 2011 Google Inc. All Rights Reserved.
// This file was generated from .js source files by GYP. If you
// want to make changes to this file you should either change the
// javascript source files or the GYP script.
#include "src/v8.h"
#include "src/snapshot/natives.h"
#include "src/utils.h"
namespace v8 {
namespace internal {
%(sources_declaration)s\
template <>
int NativesCollection<%(type)s>::GetBuiltinsCount() {
return %(builtin_count)i;
}
template <>
int NativesCollection<%(type)s>::GetDebuggerCount() {
return %(debugger_count)i;
}
template <>
int NativesCollection<%(type)s>::GetIndex(const char* name) {
%(get_index_cases)s\
return -1;
}
template <>
Vector<const char> NativesCollection<%(type)s>::GetScriptSource(int index) {
%(get_script_source_cases)s\
return Vector<const char>("", 0);
}
template <>
Vector<const char> NativesCollection<%(type)s>::GetScriptName(int index) {
%(get_script_name_cases)s\
return Vector<const char>("", 0);
}
template <>
Vector<const char> NativesCollection<%(type)s>::GetScriptsSource() {
return Vector<const char>(sources, %(total_length)i);
}
} // internal
} // v8
"""
SOURCES_DECLARATION = """\
static const char sources[] = { %s };
"""
GET_INDEX_CASE = """\
if (strcmp(name, "%(id)s") == 0) return %(i)i;
"""
GET_SCRIPT_SOURCE_CASE = """\
if (index == %(i)i) return Vector<const char>(sources + %(offset)i, %(source_length)i);
"""
GET_SCRIPT_NAME_CASE = """\
if (index == %(i)i) return Vector<const char>("%(name)s", %(length)i);
"""
def BuildFilterChain(macro_filename, message_template_file):
"""Build the chain of filter functions to be applied to the sources.
Args:
macro_filename: Name of the macro file, if any.
Returns:
A function (string -> string) that reads a source file and processes it.
"""
filter_chain = [ReadFile]
if macro_filename:
(consts, macros) = ReadMacros(ReadFile(macro_filename))
filter_chain.append(lambda l: ExpandConstants(l, consts))
filter_chain.append(lambda l: ExpandMacros(l, macros))
if message_template_file:
message_templates = ReadMessageTemplates(ReadFile(message_template_file))
filter_chain.append(lambda l: ExpandConstants(l, message_templates))
filter_chain.extend([
RemoveCommentsAndTrailingWhitespace,
ExpandInlineMacros,
ExpandInlineConstants,
Validate,
jsmin.JavaScriptMinifier().JSMinify
])
def chain(f1, f2):
return lambda x: f2(f1(x))
return reduce(chain, filter_chain)
def BuildExtraFilterChain():
return lambda x: RemoveCommentsAndTrailingWhitespace(Validate(ReadFile(x)))
class Sources:
def __init__(self):
self.names = []
self.modules = []
self.is_debugger_id = []
def IsDebuggerFile(filename):
return filename.endswith("-debugger.js")
def IsMacroFile(filename):
return filename.endswith("macros.py")
def IsMessageTemplateFile(filename):
return filename.endswith("messages.h")
def PrepareSources(source_files, extra_files, emit_js):
"""Read, prepare and assemble the list of source files.
Args:
source_files: List of JavaScript-ish source files. A file named macros.py
will be treated as a list of macros.
extra_files: List of JavaScript-ish extra source files, passed in
externally from V8. Will not be minified or macro-ified.
emit_js: True if we should skip the byte conversion and just leave the
sources as JS strings.
Returns:
An instance of Sources.
"""
macro_file = None
macro_files = filter(IsMacroFile, source_files)
assert len(macro_files) in [0, 1]
if macro_files:
source_files.remove(macro_files[0])
macro_file = macro_files[0]
message_template_file = None
message_template_files = filter(IsMessageTemplateFile, source_files)
assert len(message_template_files) in [0, 1]
if message_template_files:
source_files.remove(message_template_files[0])
message_template_file = message_template_files[0]
filters = BuildFilterChain(macro_file, message_template_file)
extra_filters = BuildExtraFilterChain()
# Sort 'debugger' sources first.
source_files = sorted(source_files,
lambda l,r: IsDebuggerFile(r) - IsDebuggerFile(l))
result = Sources()
for source in source_files:
try:
lines = filters(source)
except Error as e:
raise Error("In file %s:\n%s" % (source, str(e)))
result.modules.append(lines)
is_debugger = IsDebuggerFile(source)
result.is_debugger_id.append(is_debugger)
name = os.path.basename(source)[:-3]
result.names.append(name if not is_debugger else name[:-9])
for extra in extra_files:
try:
lines = extra_filters(extra)
except Error as e:
raise Error("In file %s:\n%s" % (extra, str(e)))
result.modules.append(lines)
name = os.path.basename(extra)[:-3]
result.names.append(name)
return result
def BuildMetadata(sources, source_bytes, native_type):
"""Build the meta data required to generate a libaries file.
Args:
sources: A Sources instance with the prepared sources.
source_bytes: A list of source bytes.
(The concatenation of all sources; might be compressed.)
native_type: The parameter for the NativesCollection template.
Returns:
A dictionary for use with HEADER_TEMPLATE.
"""
total_length = len(source_bytes)
raw_sources = "".join(sources.modules)
# The sources are expected to be ASCII-only.
assert not filter(lambda value: ord(value) >= 128, raw_sources)
# Loop over modules and build up indices into the source blob:
get_index_cases = []
get_script_name_cases = []
get_script_source_cases = []
offset = 0
for i in xrange(len(sources.modules)):
native_name = "native %s.js" % sources.names[i]
d = {
"i": i,
"id": sources.names[i],
"name": native_name,
"length": len(native_name),
"offset": offset,
"source_length": len(sources.modules[i]),
}
get_index_cases.append(GET_INDEX_CASE % d)
get_script_name_cases.append(GET_SCRIPT_NAME_CASE % d)
get_script_source_cases.append(GET_SCRIPT_SOURCE_CASE % d)
offset += len(sources.modules[i])
assert offset == len(raw_sources)
metadata = {
"builtin_count": len(sources.modules),
"debugger_count": sum(sources.is_debugger_id),
"sources_declaration": SOURCES_DECLARATION % ToCArray(source_bytes),
"total_length": total_length,
"get_index_cases": "".join(get_index_cases),
"get_script_source_cases": "".join(get_script_source_cases),
"get_script_name_cases": "".join(get_script_name_cases),
"type": native_type,
}
return metadata
def PutInt(blob_file, value):
assert(value >= 0 and value < (1 << 28))
if (value < 1 << 6):
size = 1
elif (value < 1 << 14):
size = 2
elif (value < 1 << 22):
size = 3
else:
size = 4
value_with_length = (value << 2) | (size - 1)
byte_sequence = bytearray()
for i in xrange(size):
byte_sequence.append(value_with_length & 255)
value_with_length >>= 8;
blob_file.write(byte_sequence)
def PutStr(blob_file, value):
PutInt(blob_file, len(value));
blob_file.write(value);
def WriteStartupBlob(sources, startup_blob):
"""Write a startup blob, as expected by V8 Initialize ...
TODO(vogelheim): Add proper method name.
Args:
sources: A Sources instance with the prepared sources.
startup_blob_file: Name of file to write the blob to.
"""
output = open(startup_blob, "wb")
debug_sources = sum(sources.is_debugger_id);
PutInt(output, debug_sources)
for i in xrange(debug_sources):
PutStr(output, sources.names[i]);
PutStr(output, sources.modules[i]);
PutInt(output, len(sources.names) - debug_sources)
for i in xrange(debug_sources, len(sources.names)):
PutStr(output, sources.names[i]);
PutStr(output, sources.modules[i]);
output.close()
def JS2C(sources, extra_sources, target, native_type, raw_file, startup_blob, emitJS):
prepared_sources = PrepareSources(sources, extra_sources, emitJS)
sources_output = "".join(prepared_sources.modules)
metadata = BuildMetadata(prepared_sources, sources_output, native_type)
# Optionally emit raw file.
if raw_file:
output = open(raw_file, "w")
output.write(sources_output)
output.close()
if startup_blob:
WriteStartupBlob(prepared_sources, startup_blob)
# Emit resulting source file.
output = open(target, "w")
if emitJS:
output.write(sources_output)
else:
output.write(HEADER_TEMPLATE % metadata)
output.close()
def main():
parser = argparse.ArgumentParser()
parser.add_argument("out.cc",
help="output filename")
parser.add_argument("type",
help="type parameter for NativesCollection template")
parser.add_argument("sources.js",
help="JS internal sources or macros.py.",
nargs="+")
parser.add_argument("--raw",
help="file to write the processed sources array to.")
parser.add_argument("--startup_blob",
help="file to write the startup blob to.")
parser.add_argument("--extra",
help="extra JS sources.",
nargs="*")
parser.add_argument("--js",
help="writes a JS file output instead of a C file",
action="store_true")
args = vars(parser.parse_args())
JS2C(args["sources.js"], args["extra"] or [], args["out.cc"], args["type"], args["raw"], args["startup_blob"],
args["js"])
if __name__ == "__main__":
main()