• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1#!/usr/bin/env python
2#
3# Copyright 2012 the V8 project authors. All rights reserved.
4# Redistribution and use in source and binary forms, with or without
5# modification, are permitted provided that the following conditions are
6# met:
7#
8#     * Redistributions of source code must retain the above copyright
9#       notice, this list of conditions and the following disclaimer.
10#     * Redistributions in binary form must reproduce the above
11#       copyright notice, this list of conditions and the following
12#       disclaimer in the documentation and/or other materials provided
13#       with the distribution.
14#     * Neither the name of Google Inc. nor the names of its
15#       contributors may be used to endorse or promote products derived
16#       from this software without specific prior written permission.
17#
18# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
19# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
20# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
21# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
22# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
23# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
24# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
25# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
26# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
27# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
28# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
29
30# This is a utility for converting JavaScript source code into C-style
31# char arrays. It is used for embedded JavaScript code in the V8
32# library.
33
34import os, re
35import optparse
36import jsmin
37import textwrap
38
39
40class Error(Exception):
41  def __init__(self, msg):
42    Exception.__init__(self, msg)
43
44
45def ToCArray(byte_sequence):
46  result = []
47  for chr in byte_sequence:
48    result.append(str(ord(chr)))
49  joined = ", ".join(result)
50  return textwrap.fill(joined, 80)
51
52
53def RemoveCommentsAndTrailingWhitespace(lines):
54  lines = re.sub(r'//.*\n', '\n', lines) # end-of-line comments
55  lines = re.sub(re.compile(r'/\*.*?\*/', re.DOTALL), '', lines) # comments.
56  lines = re.sub(r'\s+\n+', '\n', lines) # trailing whitespace
57  return lines
58
59
60def ReadFile(filename):
61  file = open(filename, "rt")
62  try:
63    lines = file.read()
64  finally:
65    file.close()
66  return lines
67
68
69EVAL_PATTERN = re.compile(r'\beval\s*\(')
70WITH_PATTERN = re.compile(r'\bwith\s*\(')
71INVALID_ERROR_MESSAGE_PATTERN = re.compile(
72    r'Make(?!Generic)\w*Error\(([kA-Z]\w+)')
73NEW_ERROR_PATTERN = re.compile(r'new \$\w*Error\((?!\))')
74
75def Validate(lines):
76  # Because of simplified context setup, eval and with is not
77  # allowed in the natives files.
78  if EVAL_PATTERN.search(lines):
79    raise Error("Eval disallowed in natives.")
80  if WITH_PATTERN.search(lines):
81    raise Error("With statements disallowed in natives.")
82  invalid_error = INVALID_ERROR_MESSAGE_PATTERN.search(lines)
83  if invalid_error:
84    raise Error("Unknown error message template '%s'" % invalid_error.group(1))
85  if NEW_ERROR_PATTERN.search(lines):
86    raise Error("Error constructed without message template.")
87  # Pass lines through unchanged.
88  return lines
89
90
91def ExpandConstants(lines, constants):
92  for key, value in constants:
93    lines = key.sub(str(value), lines)
94  return lines
95
96
97def ExpandMacroDefinition(lines, pos, name_pattern, macro, expander):
98  pattern_match = name_pattern.search(lines, pos)
99  while pattern_match is not None:
100    # Scan over the arguments
101    height = 1
102    start = pattern_match.start()
103    end = pattern_match.end()
104    assert lines[end - 1] == '('
105    last_match = end
106    arg_index = [0]  # Wrap state into array, to work around Python "scoping"
107    mapping = { }
108    def add_arg(str):
109      # Remember to expand recursively in the arguments
110      if arg_index[0] >= len(macro.args):
111        lineno = lines.count(os.linesep, 0, start) + 1
112        raise Error('line %s: Too many arguments for macro "%s"' % (lineno, name_pattern.pattern))
113      replacement = expander(str.strip())
114      mapping[macro.args[arg_index[0]]] = replacement
115      arg_index[0] += 1
116    while end < len(lines) and height > 0:
117      # We don't count commas at higher nesting levels.
118      if lines[end] == ',' and height == 1:
119        add_arg(lines[last_match:end])
120        last_match = end + 1
121      elif lines[end] in ['(', '{', '[']:
122        height = height + 1
123      elif lines[end] in [')', '}', ']']:
124        height = height - 1
125      end = end + 1
126    # Remember to add the last match.
127    add_arg(lines[last_match:end-1])
128    result = macro.expand(mapping)
129    # Replace the occurrence of the macro with the expansion
130    lines = lines[:start] + result + lines[end:]
131    pattern_match = name_pattern.search(lines, start + len(result))
132  return lines
133
134def ExpandMacros(lines, macros):
135  # We allow macros to depend on the previously declared macros, but
136  # we don't allow self-dependecies or recursion.
137  for name_pattern, macro in reversed(macros):
138    def expander(s):
139      return ExpandMacros(s, macros)
140    lines = ExpandMacroDefinition(lines, 0, name_pattern, macro, expander)
141  return lines
142
143class TextMacro:
144  def __init__(self, args, body):
145    self.args = args
146    self.body = body
147  def expand(self, mapping):
148    result = self.body
149    for key, value in mapping.items():
150        result = result.replace(key, value)
151    return result
152
153class PythonMacro:
154  def __init__(self, args, fun):
155    self.args = args
156    self.fun = fun
157  def expand(self, mapping):
158    args = []
159    for arg in self.args:
160      args.append(mapping[arg])
161    return str(self.fun(*args))
162
163CONST_PATTERN = re.compile(r'^define\s+([a-zA-Z0-9_]+)\s*=\s*([^;]*);$')
164MACRO_PATTERN = re.compile(r'^macro\s+([a-zA-Z0-9_]+)\s*\(([^)]*)\)\s*=\s*([^;]*);$')
165PYTHON_MACRO_PATTERN = re.compile(r'^python\s+macro\s+([a-zA-Z0-9_]+)\s*\(([^)]*)\)\s*=\s*([^;]*);$')
166
167
168def ReadMacros(lines):
169  constants = []
170  macros = []
171  for line in lines.split('\n'):
172    hash = line.find('#')
173    if hash != -1: line = line[:hash]
174    line = line.strip()
175    if len(line) is 0: continue
176    const_match = CONST_PATTERN.match(line)
177    if const_match:
178      name = const_match.group(1)
179      value = const_match.group(2).strip()
180      constants.append((re.compile("\\b%s\\b" % name), value))
181    else:
182      macro_match = MACRO_PATTERN.match(line)
183      if macro_match:
184        name = macro_match.group(1)
185        args = [match.strip() for match in macro_match.group(2).split(',')]
186        body = macro_match.group(3).strip()
187        macros.append((re.compile("\\b%s\\(" % name), TextMacro(args, body)))
188      else:
189        python_match = PYTHON_MACRO_PATTERN.match(line)
190        if python_match:
191          name = python_match.group(1)
192          args = [match.strip() for match in python_match.group(2).split(',')]
193          body = python_match.group(3).strip()
194          fun = eval("lambda " + ",".join(args) + ': ' + body)
195          macros.append((re.compile("\\b%s\\(" % name), PythonMacro(args, fun)))
196        else:
197          raise Error("Illegal line: " + line)
198  return (constants, macros)
199
200
201TEMPLATE_PATTERN = re.compile(r'^\s+T\(([A-Z][a-zA-Z0-9]*),')
202
203def ReadMessageTemplates(lines):
204  templates = []
205  index = 0
206  for line in lines.split('\n'):
207    template_match = TEMPLATE_PATTERN.match(line)
208    if template_match:
209      name = "k%s" % template_match.group(1)
210      value = index
211      index = index + 1
212      templates.append((re.compile("\\b%s\\b" % name), value))
213  return templates
214
215INLINE_MACRO_PATTERN = re.compile(r'macro\s+([a-zA-Z0-9_]+)\s*\(([^)]*)\)\s*\n')
216INLINE_MACRO_END_PATTERN = re.compile(r'endmacro\s*\n')
217
218def ExpandInlineMacros(lines):
219  pos = 0
220  while True:
221    macro_match = INLINE_MACRO_PATTERN.search(lines, pos)
222    if macro_match is None:
223      # no more macros
224      return lines
225    name = macro_match.group(1)
226    args = [match.strip() for match in macro_match.group(2).split(',')]
227    end_macro_match = INLINE_MACRO_END_PATTERN.search(lines, macro_match.end());
228    if end_macro_match is None:
229      raise Error("Macro %s unclosed" % name)
230    body = lines[macro_match.end():end_macro_match.start()]
231
232    # remove macro definition
233    lines = lines[:macro_match.start()] + lines[end_macro_match.end():]
234    name_pattern = re.compile("\\b%s\\(" % name)
235    macro = TextMacro(args, body)
236
237    # advance position to where the macro defintion was
238    pos = macro_match.start()
239
240    def non_expander(s):
241      return s
242    lines = ExpandMacroDefinition(lines, pos, name_pattern, macro, non_expander)
243
244
245INLINE_CONSTANT_PATTERN = re.compile(r'define\s+([a-zA-Z0-9_]+)\s*=\s*([^;\n]+);\n')
246
247def ExpandInlineConstants(lines):
248  pos = 0
249  while True:
250    const_match = INLINE_CONSTANT_PATTERN.search(lines, pos)
251    if const_match is None:
252      # no more constants
253      return lines
254    name = const_match.group(1)
255    replacement = const_match.group(2)
256    name_pattern = re.compile("\\b%s\\b" % name)
257
258    # remove constant definition and replace
259    lines = (lines[:const_match.start()] +
260             re.sub(name_pattern, replacement, lines[const_match.end():]))
261
262    # advance position to where the constant defintion was
263    pos = const_match.start()
264
265
266HEADER_TEMPLATE = """\
267// Copyright 2011 Google Inc. All Rights Reserved.
268
269// This file was generated from .js source files by GYP.  If you
270// want to make changes to this file you should either change the
271// javascript source files or the GYP script.
272
273#include "src/v8.h"
274#include "src/snapshot/natives.h"
275#include "src/utils.h"
276
277namespace v8 {
278namespace internal {
279
280%(sources_declaration)s\
281
282  template <>
283  int NativesCollection<%(type)s>::GetBuiltinsCount() {
284    return %(builtin_count)i;
285  }
286
287  template <>
288  int NativesCollection<%(type)s>::GetDebuggerCount() {
289    return %(debugger_count)i;
290  }
291
292  template <>
293  int NativesCollection<%(type)s>::GetIndex(const char* name) {
294%(get_index_cases)s\
295    return -1;
296  }
297
298  template <>
299  Vector<const char> NativesCollection<%(type)s>::GetScriptSource(int index) {
300%(get_script_source_cases)s\
301    return Vector<const char>("", 0);
302  }
303
304  template <>
305  Vector<const char> NativesCollection<%(type)s>::GetScriptName(int index) {
306%(get_script_name_cases)s\
307    return Vector<const char>("", 0);
308  }
309
310  template <>
311  Vector<const char> NativesCollection<%(type)s>::GetScriptsSource() {
312    return Vector<const char>(sources, %(total_length)i);
313  }
314}  // internal
315}  // v8
316"""
317
318SOURCES_DECLARATION = """\
319  static const char sources[] = { %s };
320"""
321
322
323GET_INDEX_CASE = """\
324    if (strcmp(name, "%(id)s") == 0) return %(i)i;
325"""
326
327
328GET_SCRIPT_SOURCE_CASE = """\
329    if (index == %(i)i) return Vector<const char>(sources + %(offset)i, %(source_length)i);
330"""
331
332
333GET_SCRIPT_NAME_CASE = """\
334    if (index == %(i)i) return Vector<const char>("%(name)s", %(length)i);
335"""
336
337
338def BuildFilterChain(macro_filename, message_template_file):
339  """Build the chain of filter functions to be applied to the sources.
340
341  Args:
342    macro_filename: Name of the macro file, if any.
343
344  Returns:
345    A function (string -> string) that processes a source file.
346  """
347  filter_chain = []
348
349  if macro_filename:
350    (consts, macros) = ReadMacros(ReadFile(macro_filename))
351    filter_chain.append(lambda l: ExpandMacros(l, macros))
352    filter_chain.append(lambda l: ExpandConstants(l, consts))
353
354  if message_template_file:
355    message_templates = ReadMessageTemplates(ReadFile(message_template_file))
356    filter_chain.append(lambda l: ExpandConstants(l, message_templates))
357
358  filter_chain.extend([
359    RemoveCommentsAndTrailingWhitespace,
360    ExpandInlineMacros,
361    ExpandInlineConstants,
362    Validate,
363    jsmin.JavaScriptMinifier().JSMinify
364  ])
365
366  def chain(f1, f2):
367    return lambda x: f2(f1(x))
368
369  return reduce(chain, filter_chain)
370
371def BuildExtraFilterChain():
372  return lambda x: RemoveCommentsAndTrailingWhitespace(Validate(x))
373
374class Sources:
375  def __init__(self):
376    self.names = []
377    self.modules = []
378    self.is_debugger_id = []
379
380
381def IsDebuggerFile(filename):
382  return "debug" in filename
383
384def IsMacroFile(filename):
385  return filename.endswith("macros.py")
386
387def IsMessageTemplateFile(filename):
388  return filename.endswith("messages.h")
389
390
391def PrepareSources(source_files, native_type, emit_js):
392  """Read, prepare and assemble the list of source files.
393
394  Args:
395    source_files: List of JavaScript-ish source files. A file named macros.py
396        will be treated as a list of macros.
397    native_type: String corresponding to a NativeType enum value, allowing us
398        to treat different types of sources differently.
399    emit_js: True if we should skip the byte conversion and just leave the
400        sources as JS strings.
401
402  Returns:
403    An instance of Sources.
404  """
405  macro_file = None
406  macro_files = filter(IsMacroFile, source_files)
407  assert len(macro_files) in [0, 1]
408  if macro_files:
409    source_files.remove(macro_files[0])
410    macro_file = macro_files[0]
411
412  message_template_file = None
413  message_template_files = filter(IsMessageTemplateFile, source_files)
414  assert len(message_template_files) in [0, 1]
415  if message_template_files:
416    source_files.remove(message_template_files[0])
417    message_template_file = message_template_files[0]
418
419  filters = None
420  if native_type in ("EXTRAS", "EXPERIMENTAL_EXTRAS"):
421    filters = BuildExtraFilterChain()
422  else:
423    filters = BuildFilterChain(macro_file, message_template_file)
424
425  # Sort 'debugger' sources first.
426  source_files = sorted(source_files,
427                        lambda l,r: IsDebuggerFile(r) - IsDebuggerFile(l))
428
429  source_files_and_contents = [(f, ReadFile(f)) for f in source_files]
430
431  # Have a single not-quite-empty source file if there are none present;
432  # otherwise you get errors trying to compile an empty C++ array.
433  # It cannot be empty (or whitespace, which gets trimmed to empty), as
434  # the deserialization code assumes each file is nonempty.
435  if not source_files_and_contents:
436    source_files_and_contents = [("dummy.js", "(function() {})")]
437
438  result = Sources()
439
440  for (source, contents) in source_files_and_contents:
441    try:
442      lines = filters(contents)
443    except Error as e:
444      raise Error("In file %s:\n%s" % (source, str(e)))
445
446    result.modules.append(lines)
447
448    is_debugger = IsDebuggerFile(source)
449    result.is_debugger_id.append(is_debugger)
450
451    name = os.path.basename(source)[:-3]
452    result.names.append(name)
453
454  return result
455
456
457def BuildMetadata(sources, source_bytes, native_type):
458  """Build the meta data required to generate a libaries file.
459
460  Args:
461    sources: A Sources instance with the prepared sources.
462    source_bytes: A list of source bytes.
463        (The concatenation of all sources; might be compressed.)
464    native_type: The parameter for the NativesCollection template.
465
466  Returns:
467    A dictionary for use with HEADER_TEMPLATE.
468  """
469  total_length = len(source_bytes)
470  raw_sources = "".join(sources.modules)
471
472  # The sources are expected to be ASCII-only.
473  assert not filter(lambda value: ord(value) >= 128, raw_sources)
474
475  # Loop over modules and build up indices into the source blob:
476  get_index_cases = []
477  get_script_name_cases = []
478  get_script_source_cases = []
479  offset = 0
480  for i in xrange(len(sources.modules)):
481    native_name = "native %s.js" % sources.names[i]
482    d = {
483        "i": i,
484        "id": sources.names[i],
485        "name": native_name,
486        "length": len(native_name),
487        "offset": offset,
488        "source_length": len(sources.modules[i]),
489    }
490    get_index_cases.append(GET_INDEX_CASE % d)
491    get_script_name_cases.append(GET_SCRIPT_NAME_CASE % d)
492    get_script_source_cases.append(GET_SCRIPT_SOURCE_CASE % d)
493    offset += len(sources.modules[i])
494  assert offset == len(raw_sources)
495
496  metadata = {
497    "builtin_count": len(sources.modules),
498    "debugger_count": sum(sources.is_debugger_id),
499    "sources_declaration": SOURCES_DECLARATION % ToCArray(source_bytes),
500    "total_length": total_length,
501    "get_index_cases": "".join(get_index_cases),
502    "get_script_source_cases": "".join(get_script_source_cases),
503    "get_script_name_cases": "".join(get_script_name_cases),
504    "type": native_type,
505  }
506  return metadata
507
508
509def PutInt(blob_file, value):
510  assert(value >= 0 and value < (1 << 28))
511  if (value < 1 << 6):
512    size = 1
513  elif (value < 1 << 14):
514    size = 2
515  elif (value < 1 << 22):
516    size = 3
517  else:
518    size = 4
519  value_with_length = (value << 2) | (size - 1)
520
521  byte_sequence = bytearray()
522  for i in xrange(size):
523    byte_sequence.append(value_with_length & 255)
524    value_with_length >>= 8;
525  blob_file.write(byte_sequence)
526
527
528def PutStr(blob_file, value):
529  PutInt(blob_file, len(value));
530  blob_file.write(value);
531
532
533def WriteStartupBlob(sources, startup_blob):
534  """Write a startup blob, as expected by V8 Initialize ...
535    TODO(vogelheim): Add proper method name.
536
537  Args:
538    sources: A Sources instance with the prepared sources.
539    startup_blob_file: Name of file to write the blob to.
540  """
541  output = open(startup_blob, "wb")
542
543  debug_sources = sum(sources.is_debugger_id);
544  PutInt(output, debug_sources)
545  for i in xrange(debug_sources):
546    PutStr(output, sources.names[i]);
547    PutStr(output, sources.modules[i]);
548
549  PutInt(output, len(sources.names) - debug_sources)
550  for i in xrange(debug_sources, len(sources.names)):
551    PutStr(output, sources.names[i]);
552    PutStr(output, sources.modules[i]);
553
554  output.close()
555
556
557def JS2C(sources, target, native_type, raw_file, startup_blob, emit_js):
558  prepared_sources = PrepareSources(sources, native_type, emit_js)
559  sources_output = "".join(prepared_sources.modules)
560  metadata = BuildMetadata(prepared_sources, sources_output, native_type)
561
562  # Optionally emit raw file.
563  if raw_file:
564    output = open(raw_file, "w")
565    output.write(sources_output)
566    output.close()
567
568  if startup_blob:
569    WriteStartupBlob(prepared_sources, startup_blob)
570
571  # Emit resulting source file.
572  output = open(target, "w")
573  if emit_js:
574    output.write(sources_output)
575  else:
576    output.write(HEADER_TEMPLATE % metadata)
577  output.close()
578
579
580def main():
581  parser = optparse.OptionParser()
582  parser.add_option("--raw",
583                    help="file to write the processed sources array to.")
584  parser.add_option("--startup_blob",
585                    help="file to write the startup blob to.")
586  parser.add_option("--js",
587                    help="writes a JS file output instead of a C file",
588                    action="store_true", default=False, dest='js')
589  parser.add_option("--nojs", action="store_false", default=False, dest='js')
590  parser.set_usage("""js2c out.cc type sources.js ...
591        out.cc: C code to be generated.
592        type: type parameter for NativesCollection template.
593        sources.js: JS internal sources or macros.py.""")
594  (options, args) = parser.parse_args()
595  JS2C(args[2:],
596       args[0],
597       args[1],
598       options.raw,
599       options.startup_blob,
600       options.js)
601
602
603if __name__ == "__main__":
604  main()
605