grit/gather/rc.py

#!/usr/bin/env python
# Copyright (c) 2012 The Chromium Authors. All rights reserved.
# Use of this source code is governed by a BSD-style license that can be
# found in the LICENSE file.

'''Support for gathering resources from RC files.
'''


import re

from grit import exception
from grit import lazy_re
from grit import tclib

from grit.gather import regexp


# Find portions that need unescaping in resource strings.  We need to be
# careful that a \\n is matched _first_ as a \\ rather than matching as
# a \ followed by a \n.
# TODO(joi) Handle ampersands if we decide to change them into <ph>
# TODO(joi) May need to handle other control characters than \n
_NEED_UNESCAPE = lazy_re.compile(r'""|\\\\|\\n|\\t')

# Find portions that need escaping to encode string as a resource string.
_NEED_ESCAPE = lazy_re.compile(r'"|\n|\t|\\|\&nbsp\;')

# How to escape certain characters
_ESCAPE_CHARS = {
  '"' : '""',
  '\n' : '\\n',
  '\t' : '\\t',
  '\\' : '\\\\',
  '&nbsp;' : ' '
}

# How to unescape certain strings
_UNESCAPE_CHARS = dict([[value, key] for key, value in _ESCAPE_CHARS.items()])



class Section(regexp.RegexpGatherer):
  '''A section from a resource file.'''

  @staticmethod
  def Escape(text):
    '''Returns a version of 'text' with characters escaped that need to be
    for inclusion in a resource section.'''
    def Replace(match):
      return _ESCAPE_CHARS[match.group()]
    return _NEED_ESCAPE.sub(Replace, text)

  @staticmethod
  def UnEscape(text):
    '''Returns a version of 'text' with escaped characters unescaped.'''
    def Replace(match):
      return _UNESCAPE_CHARS[match.group()]
    return _NEED_UNESCAPE.sub(Replace, text)

  def _RegExpParse(self, rexp, text_to_parse):
    '''Overrides _RegExpParse to add shortcut group handling.  Otherwise
    the same.
    '''
    super(Section, self)._RegExpParse(rexp, text_to_parse)

    if not self.is_skeleton and len(self.GetTextualIds()) > 0:
      group_name = self.GetTextualIds()[0]
      for c in self.GetCliques():
        c.AddToShortcutGroup(group_name)

  def ReadSection(self):
    rc_text = self._LoadInputFile()

    out = ''
    begin_count = 0
    assert self.extkey
    first_line_re = re.compile(r'\s*' + self.extkey + r'\b')
    for line in rc_text.splitlines(True):
      if out or first_line_re.match(line):
        out += line

      # we stop once we reach the END for the outermost block.
      begin_count_was = begin_count
      if len(out) > 0 and line.strip() == 'BEGIN':
        begin_count += 1
      elif len(out) > 0 and line.strip() == 'END':
        begin_count -= 1
      if begin_count_was == 1 and begin_count == 0:
        break

    if len(out) == 0:
      raise exception.SectionNotFound('%s in file %s' % (self.extkey, self.rc_file))

    self.text_ = out.strip()


class Dialog(Section):
  '''A resource section that contains a dialog resource.'''

  # A typical dialog resource section looks like this:
  #
  # IDD_ABOUTBOX DIALOGEX 22, 17, 230, 75
  # STYLE DS_SETFONT | DS_MODALFRAME | WS_CAPTION | WS_SYSMENU
  # CAPTION "About"
  # FONT 8, "System", 0, 0, 0x0
  # BEGIN
  #     ICON            IDI_KLONK,IDC_MYICON,14,9,20,20
  #     LTEXT           "klonk Version ""yibbee"" 1.0",IDC_STATIC,49,10,119,8,
  #                     SS_NOPREFIX
  #     LTEXT           "Copyright (C) 2005",IDC_STATIC,49,20,119,8
  #     DEFPUSHBUTTON   "OK",IDOK,195,6,30,11,WS_GROUP
  #     CONTROL         "Jack ""Black"" Daniels",IDC_RADIO1,"Button",
  #                     BS_AUTORADIOBUTTON,46,51,84,10
  # END

  # We are using a sorted set of keys, and we assume that the
  # group name used for descriptions (type) will come after the "text"
  # group in alphabetical order. We also assume that there cannot be
  # more than one description per regular expression match.
  # If that's not the case some descriptions will be clobbered.
  dialog_re_ = lazy_re.compile('''
    # The dialog's ID in the first line
    (?P<id1>[A-Z0-9_]+)\s+DIALOG(EX)?
    |
    # The caption of the dialog
    (?P<type1>CAPTION)\s+"(?P<text1>.*?([^"]|""))"\s
    |
    # Lines for controls that have text and an ID
    \s+(?P<type2>[A-Z]+)\s+"(?P<text2>.*?([^"]|"")?)"\s*,\s*(?P<id2>[A-Z0-9_]+)\s*,
    |
    # Lines for controls that have text only
    \s+(?P<type3>[A-Z]+)\s+"(?P<text3>.*?([^"]|"")?)"\s*,
    |
    # Lines for controls that reference other resources
    \s+[A-Z]+\s+[A-Z0-9_]+\s*,\s*(?P<id3>[A-Z0-9_]*[A-Z][A-Z0-9_]*)
    |
    # This matches "NOT SOME_STYLE" so that it gets consumed and doesn't get
    # matched by the next option (controls that have only an ID and then just
    # numbers)
    \s+NOT\s+[A-Z][A-Z0-9_]+
    |
    # Lines for controls that have only an ID and then just numbers
    \s+[A-Z]+\s+(?P<id4>[A-Z0-9_]*[A-Z][A-Z0-9_]*)\s*,
    ''', re.MULTILINE | re.VERBOSE)

  def Parse(self):
    '''Knows how to parse dialog resource sections.'''
    self.ReadSection()
    self._RegExpParse(self.dialog_re_, self.text_)


class Menu(Section):
  '''A resource section that contains a menu resource.'''

  # A typical menu resource section looks something like this:
  #
  # IDC_KLONK MENU
  # BEGIN
  #     POPUP "&File"
  #     BEGIN
  #         MENUITEM "E&xit",                       IDM_EXIT
  #         MENUITEM "This be ""Klonk"" me like",   ID_FILE_THISBE
  #         POPUP "gonk"
  #         BEGIN
  #             MENUITEM "Klonk && is ""good""",           ID_GONK_KLONKIS
  #         END
  #     END
  #     POPUP "&Help"
  #     BEGIN
  #         MENUITEM "&About ...",                  IDM_ABOUT
  #     END
  # END

  # Description used for the messages generated for menus, to explain to
  # the translators how to handle them.
  MENU_MESSAGE_DESCRIPTION = (
    'This message represents a menu. Each of the items appears in sequence '
    '(some possibly within sub-menus) in the menu. The XX01XX placeholders '
    'serve to separate items. Each item contains an & (ampersand) character '
    'in front of the keystroke that should be used as a shortcut for that item '
    'in the menu. Please make sure that no two items in the same menu share '
    'the same shortcut.'
  )

  # A dandy regexp to suck all the IDs and translateables out of a menu
  # resource
  menu_re_ = lazy_re.compile('''
    # Match the MENU ID on the first line
    ^(?P<id1>[A-Z0-9_]+)\s+MENU
    |
    # Match the translateable caption for a popup menu
    POPUP\s+"(?P<text1>.*?([^"]|""))"\s
    |
    # Match the caption & ID of a MENUITEM
    MENUITEM\s+"(?P<text2>.*?([^"]|""))"\s*,\s*(?P<id2>[A-Z0-9_]+)
    ''', re.MULTILINE | re.VERBOSE)

  def Parse(self):
    '''Knows how to parse menu resource sections.  Because it is important that
    menu shortcuts are unique within the menu, we return each menu as a single
    message with placeholders to break up the different menu items, rather than
    return a single message per menu item.  we also add an automatic description
    with instructions for the translators.'''
    self.ReadSection()
    self.single_message_ = tclib.Message(description=self.MENU_MESSAGE_DESCRIPTION)
    self._RegExpParse(self.menu_re_, self.text_)


class Version(Section):
  '''A resource section that contains a VERSIONINFO resource.'''

  # A typical version info resource can look like this:
  #
  # VS_VERSION_INFO VERSIONINFO
  #  FILEVERSION 1,0,0,1
  #  PRODUCTVERSION 1,0,0,1
  #  FILEFLAGSMASK 0x3fL
  # #ifdef _DEBUG
  #  FILEFLAGS 0x1L
  # #else
  #  FILEFLAGS 0x0L
  # #endif
  #  FILEOS 0x4L
  #  FILETYPE 0x2L
  #  FILESUBTYPE 0x0L
  # BEGIN
  #     BLOCK "StringFileInfo"
  #     BEGIN
  #         BLOCK "040904e4"
  #         BEGIN
  #             VALUE "CompanyName", "TODO: <Company name>"
  #             VALUE "FileDescription", "TODO: <File description>"
  #             VALUE "FileVersion", "1.0.0.1"
  #             VALUE "LegalCopyright", "TODO: (c) <Company name>.  All rights reserved."
  #             VALUE "InternalName", "res_format_test.dll"
  #             VALUE "OriginalFilename", "res_format_test.dll"
  #             VALUE "ProductName", "TODO: <Product name>"
  #             VALUE "ProductVersion", "1.0.0.1"
  #         END
  #     END
  #     BLOCK "VarFileInfo"
  #     BEGIN
  #         VALUE "Translation", 0x409, 1252
  #     END
  # END
  #
  #
  # In addition to the above fields, VALUE fields named "Comments" and
  # "LegalTrademarks" may also be translateable.

  version_re_ = lazy_re.compile('''
    # Match the ID on the first line
    ^(?P<id1>[A-Z0-9_]+)\s+VERSIONINFO
    |
    # Match all potentially translateable VALUE sections
    \s+VALUE\s+"
    (
      CompanyName|FileDescription|LegalCopyright|
      ProductName|Comments|LegalTrademarks
    )",\s+"(?P<text1>.*?([^"]|""))"\s
    ''', re.MULTILINE | re.VERBOSE)

  def Parse(self):
    '''Knows how to parse VERSIONINFO resource sections.'''
    self.ReadSection()
    self._RegExpParse(self.version_re_, self.text_)

  # TODO(joi) May need to override the Translate() method to change the
  # "Translation" VALUE block to indicate the correct language code.


class RCData(Section):
  '''A resource section that contains some data .'''

  # A typical rcdataresource section looks like this:
  #
  # IDR_BLAH        RCDATA      { 1, 2, 3, 4 }

  dialog_re_ = lazy_re.compile('''
    ^(?P<id1>[A-Z0-9_]+)\s+RCDATA\s+(DISCARDABLE)?\s+\{.*?\}
    ''', re.MULTILINE | re.VERBOSE | re.DOTALL)

  def Parse(self):
    '''Implementation for resource types w/braces (not BEGIN/END)
    '''
    rc_text = self._LoadInputFile()

    out = ''
    begin_count = 0
    openbrace_count = 0
    assert self.extkey
    first_line_re = re.compile(r'\s*' + self.extkey + r'\b')
    for line in rc_text.splitlines(True):
      if out or first_line_re.match(line):
        out += line

      # We stop once the braces balance (could happen in one line).
      begin_count_was = begin_count
      if len(out) > 0:
        openbrace_count += line.count('{')
        begin_count += line.count('{')
        begin_count -= line.count('}')
      if ((begin_count_was == 1 and begin_count == 0) or
         (openbrace_count > 0 and begin_count == 0)):
        break

    if len(out) == 0:
      raise exception.SectionNotFound('%s in file %s' % (self.extkey, self.rc_file))

    self.text_ = out

    self._RegExpParse(self.dialog_re_, out)


class Accelerators(Section):
  '''An ACCELERATORS table.
  '''

  # A typical ACCELERATORS section looks like this:
  #
  # IDR_ACCELERATOR1 ACCELERATORS
  # BEGIN
  #   "^C",           ID_ACCELERATOR32770,    ASCII,  NOINVERT
  #   "^V",           ID_ACCELERATOR32771,    ASCII,  NOINVERT
  #   VK_INSERT,      ID_ACCELERATOR32772,    VIRTKEY, CONTROL, NOINVERT
  # END

  accelerators_re_ = lazy_re.compile('''
    # Match the ID on the first line
    ^(?P<id1>[A-Z0-9_]+)\s+ACCELERATORS\s+
    |
    # Match accelerators specified as VK_XXX
    \s+VK_[A-Z0-9_]+,\s*(?P<id2>[A-Z0-9_]+)\s*,
    |
    # Match accelerators specified as e.g. "^C"
    \s+"[^"]*",\s+(?P<id3>[A-Z0-9_]+)\s*,
    ''', re.MULTILINE | re.VERBOSE)

  def Parse(self):
    '''Knows how to parse ACCELERATORS resource sections.'''
    self.ReadSection()
    self._RegExpParse(self.accelerators_re_, self.text_)