Skip to content
Snippets Groups Projects
polib.py 60.6 KiB
Newer Older
# no-check-code
#
# License: MIT (see LICENSE file provided)
# vim: set expandtab tabstop=4 shiftwidth=4 softtabstop=4:

"""
**polib** allows you to manipulate, create, modify gettext files (pot, po and
mo files).  You can load existing files, iterate through it's entries, add,
modify entries, comments or metadata, etc. or create new po files from scratch.
**polib** provides a simple and pythonic API via the :func:`~polib.pofile` and
:func:`~polib.mofile` convenience functions.
__author__ = 'David Jean Louis <izimobil@gmail.com>'
__version__ = '1.0.7'
__all__ = [
    'pofile',
    'POFile',
    'POEntry',
    'mofile',
    'MOFile',
    'MOEntry',
    'default_encoding',
    'escape',
    'unescape',
    'detect_encoding',
]
import os
import re
import textwrap

try:
    import io
except ImportError:
    # replacement of io.open() for python < 2.6
    # we use codecs instead
        @staticmethod
        def open(fpath, mode='r', encoding=None):
            return codecs.open(fpath, mode, encoding)
# the default encoding to use when encoding cannot be detected
default_encoding = 'utf-8'

# python 2/3 compatibility helpers {{{


if sys.version_info[:2] < (3, 0):
    PY3 = False
    text_type = unicode

    def b(s):
        return s

    def u(s):
        return unicode(s, "unicode_escape")

else:
    PY3 = True
    text_type = str

    def b(s):
        return s.encode("latin-1")

    def u(s):
        return s
# _pofile_or_mofile {{{
def _pofile_or_mofile(f, type, **kwargs):
    """
    Internal function used by :func:`polib.pofile` and :func:`polib.mofile` to
    honor the DRY concept.
    # get the file encoding
    enc = kwargs.get('encoding')
    if enc is None:
        enc = detect_encoding(f, type == 'mofile')

    # parse the file
    kls = type == 'pofile' and _POFileParser or _MOFileParser
    parser = kls(
        f,
        check_for_duplicates=kwargs.get('check_for_duplicates', False),
        klass=kwargs.get('klass'),
    )
    instance = parser.parse()
    instance.wrapwidth = kwargs.get('wrapwidth', 78)
    return instance

def _is_file(filename_or_contents):
    """
    Safely returns the value of os.path.exists(filename_or_contents).

    Arguments:

    ``filename_or_contents``
        either a filename, or a string holding the contents of some file.
        In the latter case, this function will always return False.
    """
    try:
        return os.path.exists(filename_or_contents)
    except (ValueError, UnicodeEncodeError):
        return False
# function pofile() {{{

def pofile(pofile, **kwargs):
    """
    Convenience function that parses the po or pot file ``pofile`` and returns
    a :class:`~polib.POFile` instance.

    Arguments:

    ``pofile``
        string, full or relative path to the po/pot file or its content (data).

    ``wrapwidth``
        integer, the wrap width, only useful when the ``-w`` option was passed
        to xgettext (optional, default: ``78``).

    ``encoding``
        string, the encoding to use (e.g. "utf-8") (default: ``None``, the
        encoding will be auto-detected).

    ``check_for_duplicates``
        whether to check for duplicate entries when adding entries to the
        file (optional, default: ``False``).

    ``klass``
        class which is used to instantiate the return value (optional,
        default: ``None``, the return value with be a :class:`~polib.POFile`
        instance).
    """
    return _pofile_or_mofile(pofile, 'pofile', **kwargs)
# function mofile() {{{

def mofile(mofile, **kwargs):
    Convenience function that parses the mo file ``mofile`` and returns a
    :class:`~polib.MOFile` instance.
    ``mofile``
        string, full or relative path to the mo file or its content (data).
    ``wrapwidth``
        integer, the wrap width, only useful when the ``-w`` option was passed
        to xgettext to generate the po file that was used to format the mo file
        (optional, default: ``78``).

    ``encoding``
        string, the encoding to use (e.g. "utf-8") (default: ``None``, the
        encoding will be auto-detected).

    ``check_for_duplicates``
        whether to check for duplicate entries when adding entries to the
        file (optional, default: ``False``).

    ``klass``
        class which is used to instantiate the return value (optional,
        default: ``None``, the return value with be a :class:`~polib.POFile`
        instance).
    return _pofile_or_mofile(mofile, 'mofile', **kwargs)
# }}}
# function detect_encoding() {{{

def detect_encoding(file, binary_mode=False):
    Try to detect the encoding used by the ``file``. The ``file`` argument can
    be a PO or MO file path or a string containing the contents of the file.
    If the encoding cannot be detected, the function will return the value of
    ``default_encoding``.
    Arguments:

    ``file``
        string, full or relative path to the po/mo file or its content.
    ``binary_mode``
        boolean, set this to True if ``file`` is a mo file.
    """
    PATTERN = r'"?Content-Type:.+? charset=([\w_\-:\.]+)'
    rxt = re.compile(u(PATTERN))
    rxb = re.compile(b(PATTERN))
    def charset_exists(charset):
        """Check whether ``charset`` is valid or not."""
        try:
            codecs.lookup(charset)
        except LookupError:
            return False
        return True

    if not _is_file(file):
        match = rxt.search(file)
        if match:
            enc = match.group(1).strip()
            if charset_exists(enc):
                return enc
        # For PY3, always treat as binary
        if binary_mode or PY3:
        f = open(file, mode)
        for l in f.readlines():
            match = rx.search(l)
            if match:
                f.close()
                enc = match.group(1).strip()
                if not isinstance(enc, text_type):
                    enc = enc.decode('utf-8')
                if charset_exists(enc):
                    return enc
        f.close()
    return default_encoding
# }}}
# function escape() {{{

def escape(st):
    """
    Escapes the characters ``\\\\``, ``\\t``, ``\\n``, ``\\r`` and ``"`` in
    the given string ``st`` and returns it.
    return (
        st.replace('\\', r'\\')
        .replace('\t', r'\t')
        .replace('\r', r'\r')
        .replace('\n', r'\n')
        .replace('\"', r'\"')
    )


# }}}
# function unescape() {{{

def unescape(st):
    """
    Unescapes the characters ``\\\\``, ``\\t``, ``\\n``, ``\\r`` and ``"`` in
    the given string ``st`` and returns it.
    def unescape_repl(m):
        m = m.group(1)
        if m == 'n':
            return '\n'
        if m == 't':
            return '\t'
        if m == 'r':
            return '\r'
        if m == '\\':
            return '\\'
        return m  # handles escaped double quote
    return re.sub(r'\\(\\|n|t|r|")', unescape_repl, st)
# }}}
# class _BaseFile {{{

class _BaseFile(list):
    """
    Common base class for the :class:`~polib.POFile` and :class:`~polib.MOFile`
    classes. This class should **not** be instanciated directly.
    """

    def __init__(self, *args, **kwargs):
        """
        Constructor, accepts the following keyword arguments:

        ``pofile``
            string, the path to the po or mo file, or its content as a string.
        ``wrapwidth``
            integer, the wrap width, only useful when the ``-w`` option was
            passed to xgettext (optional, default: ``78``).

        ``encoding``
            string, the encoding to use, defaults to ``default_encoding``
            global variable (optional).

        ``check_for_duplicates``
            whether to check for duplicate entries when adding entries to the
            file, (optional, default: ``False``).
        """
        list.__init__(self)
        # the opened file handle
        pofile = kwargs.get('pofile', None)
        if pofile and _is_file(pofile):
            self.fpath = pofile
        else:
            self.fpath = kwargs.get('fpath')
        # the width at which lines should be wrapped
        self.wrapwidth = kwargs.get('wrapwidth', 78)
        # the file encoding
        self.encoding = kwargs.get('encoding', default_encoding)
        # whether to check for duplicate entries or not
        self.check_for_duplicates = kwargs.get('check_for_duplicates', False)
        # header
        self.header = ''
        # both po and mo files have metadata
        self.metadata = {}
        self.metadata_is_fuzzy = 0

    def __unicode__(self):
        Returns the unicode representation of the file.
        entries = [self.metadata_as_entry()] + [
            e for e in self if not e.obsolete
        ]
        for entry in entries:
            ret.append(entry.__unicode__(self.wrapwidth))
        for entry in self.obsolete_entries():
            ret.append(entry.__unicode__(self.wrapwidth))
        assert isinstance(ret, text_type)
        # if type(ret) != text_type:
        #    return unicode(ret, self.encoding)
        def __str__(self):
            return self.__unicode__()
        def __str__(self):
            """
            Returns the string representation of the file.
            """
            return unicode(self).encode(self.encoding)

    def __contains__(self, entry):
        """
        Overridden ``list`` method to implement the membership test (in and
        not in).
        The method considers that an entry is in the file if it finds an entry
        that has the same msgid (the test is **case sensitive**) and the same
        msgctxt (or none for both entries).
        Argument:

        ``entry``
            an instance of :class:`~polib._BaseEntry`.
        return (
            self.find(entry.msgid, by='msgid', msgctxt=entry.msgctxt)
    def __eq__(self, other):

    def append(self, entry):
        """
        Overridden method to check for duplicates entries, if a user tries to
        add an entry that is already in the file, the method will raise a
        ``ValueError`` exception.
        Argument:

        ``entry``
            an instance of :class:`~polib._BaseEntry`.
        """
        if self.check_for_duplicates and entry in self:
            raise ValueError('Entry "%s" already exists' % entry.msgid)
        super(_BaseFile, self).append(entry)

    def insert(self, index, entry):
        """
        Overridden method to check for duplicates entries, if a user tries to
        add an entry that is already in the file, the method will raise a
        ``ValueError`` exception.
        ``index``
            index at which the entry should be inserted.

        ``entry``
            an instance of :class:`~polib._BaseEntry`.
        """
        if self.check_for_duplicates and entry in self:
            raise ValueError('Entry "%s" already exists' % entry.msgid)
        super(_BaseFile, self).insert(index, entry)

    def metadata_as_entry(self):
        """
        Returns the file metadata as a :class:`~polib.POFile` instance.
        """
        e = POEntry(msgid='')
        mdata = self.ordered_metadata()
        if mdata:
            strs = []
            for name, value in mdata:
                # Strip whitespace off each line in a multi-line entry
                strs.append('%s: %s' % (name, value))
            e.msgstr = '\n'.join(strs) + '\n'
        if self.metadata_is_fuzzy:
            e.flags.append('fuzzy')
        return e

    def save(self, fpath=None, repr_method='__unicode__'):
        Saves the po file to ``fpath``.
        If it is an existing file and no ``fpath`` is provided, then the
        existing file is rewritten with the modified data.

        Keyword arguments:
        ``fpath``
            string, full or relative path to the file.

        ``repr_method``
            string, the method to use for output.
        """
        if self.fpath is None and fpath is None:
            raise IOError('You must provide a file path to save() method')
        contents = getattr(self, repr_method)()
        if fpath is None:
            fpath = self.fpath
        if repr_method == 'to_binary':
            fhandle = open(fpath, 'wb')
        else:
            fhandle = io.open(fpath, 'w', encoding=self.encoding)
            if not isinstance(contents, text_type):
                contents = contents.decode(self.encoding)
        fhandle.write(contents)
        fhandle.close()
        # set the file path if not set
        if self.fpath is None and fpath:
            self.fpath = fpath
    def find(
        self, st, by='msgid', include_obsolete_entries=False, msgctxt=False
    ):
        """
        Find the entry which msgid (or property identified by the ``by``
        argument) matches the string ``st``.
        Keyword arguments:

        ``st``
            string, the string to search for.
        ``by``
            string, the property to use for comparison (default: ``msgid``).

        ``include_obsolete_entries``
            boolean, whether to also search in entries that are obsolete.

        ``msgctxt``
            string, allows specifying a specific message context for the
        if include_obsolete_entries:
            entries = self[:]
        else:
            entries = [e for e in self if not e.obsolete]
        for e in entries:
            if getattr(e, by) == st:
                if msgctxt is not False and e.msgctxt != msgctxt:
                return e
        return None

    def ordered_metadata(self):
        """
        Convenience method that returns an ordered version of the metadata
        dictionary. The return value is list of tuples (metadata name,
        metadata_value).
        """
        # copy the dict first
        metadata = self.metadata.copy()
        data_order = [
            'Project-Id-Version',
            'Report-Msgid-Bugs-To',
            'POT-Creation-Date',
            'PO-Revision-Date',
            'Last-Translator',
            'Language-Team',
            'MIME-Version',
            'Content-Type',
            'Content-Transfer-Encoding',
            'Language',
            'Plural-Forms',
        ]
        ordered_data = []
        for data in data_order:
            try:
                value = metadata.pop(data)
                ordered_data.append((data, value))
            except KeyError:
                pass
        # the rest of the metadata will be alphabetically ordered since there
        # are no specs for this AFAIK
        for data in sorted(metadata.keys()):
            value = metadata[data]
            ordered_data.append((data, value))
        return ordered_data

    def to_binary(self):
        """
        Return the binary representation of the file.
        """
        offsets = []
        entries = self.translated_entries()
        # add metadata entry
        entries.sort(key=lambda o: o.msgctxt or o.msgid)
        mentry = self.metadata_as_entry()
        # mentry.msgstr = mentry.msgstr.replace('\\n', '').lstrip()
        entries = [mentry] + entries
        entries_len = len(entries)
        for e in entries:
            # For each string, we need size and file offset.  Each string is
            # NUL terminated; the NUL does not count into the size.
            if e.msgctxt:
                # Contexts are stored by storing the concatenation of the
                # context, a <EOT> byte, and the original string
                msgid = self._encode(e.msgctxt + '\4')
            if e.msgid_plural:
                msgstr = []
                for index in sorted(e.msgstr_plural.keys()):
                    msgstr.append(e.msgstr_plural[index])
                msgid += self._encode(e.msgid + '\0' + e.msgid_plural)
                msgstr = self._encode('\0'.join(msgstr))
            else:
                msgid += self._encode(e.msgid)
                msgstr = self._encode(e.msgstr)
            offsets.append((len(ids), len(msgid), len(strs), len(msgstr)))
            ids += msgid + b('\0')
            strs += msgstr + b('\0')
        # The header is 7 32-bit unsigned integers.
        keystart = 7 * 4 + 16 * entries_len
        # and the values start after the keys
        valuestart = keystart + len(ids)
        koffsets = []
        voffsets = []
        # The string table first has the list of keys, then the list of values.
        # Each entry has first the size of the string, then the file offset.
        for o1, l1, o2, l2 in offsets:
            koffsets += [l1, o1 + keystart]
            voffsets += [l2, o2 + valuestart]
        offsets = koffsets + voffsets

        output = struct.pack(
            "Iiiiiii",
            # Magic number
            MOFile.MAGIC,
            # Version
            0,
            # number of entries
            entries_len,
            # start of key index
            7 * 4,
            # start of value index
            7 * 4 + entries_len * 8,
            # size and offset of hash table, we don't use hash tables
        if PY3 and sys.version_info.minor > 1:  # python 3.2 or superior
            output += array.array("i", offsets).tobytes()
        else:
            output += array.array("i", offsets).tostring()
        output += ids
        output += strs
        return output

    def _encode(self, mixed):
        """
        Encodes the given ``mixed`` argument with the file encoding if and
        only if it's an unicode string and returns the encoded string.
        if isinstance(mixed, text_type):
            mixed = mixed.encode(self.encoding)
# }}}
# class POFile {{{

class POFile(_BaseFile):
    Po (or Pot) file reader/writer.
    This class inherits the :class:`~polib._BaseFile` class and, by extension,
    the python ``list`` type.
    """
    def __unicode__(self):
        """
        Returns the unicode representation of the po file.
        """
        ret, headers = '', self.header.split('\n')
        for header in headers:
            if not len(header):
                ret += "#\n"
            elif header[:1] in [',', ':']:
                ret += '#%s\n' % header
            else:
                ret += '# %s\n' % header
        if not isinstance(ret, text_type):
            ret = ret.decode(self.encoding)

        return ret + _BaseFile.__unicode__(self)

    def save_as_mofile(self, fpath):
        """
        Saves the binary representation of the file to given ``fpath``.
        Keyword argument:

        ``fpath``
            string, full or relative path to the mo file.
        """
        _BaseFile.save(self, fpath, 'to_binary')

    def percent_translated(self):
        """
        Convenience method that returns the percentage of translated
        messages.
        """
        total = len([e for e in self if not e.obsolete])
        if total == 0:
            return 100
        translated = len(self.translated_entries())
        return int(translated * 100 / float(total))

    def translated_entries(self):
        """
        Convenience method that returns the list of translated entries.
        """
        return [e for e in self if e.translated()]

    def untranslated_entries(self):
        """
        Convenience method that returns the list of untranslated entries.
        return [
            e
            for e in self
            if not e.translated() and not e.obsolete and not 'fuzzy' in e.flags
        ]

    def fuzzy_entries(self):
        """
        Convenience method that returns the list of fuzzy entries.
        """
        return [e for e in self if 'fuzzy' in e.flags]

    def obsolete_entries(self):
        """
        Convenience method that returns the list of obsolete entries.
        """
        return [e for e in self if e.obsolete]

    def merge(self, refpot):
        """
        Convenience method that merges the current pofile with the pot file
        provided. It behaves exactly as the gettext msgmerge utility:

        * comments of this file will be preserved, but extracted comments and
          occurrences will be discarded;
        * any translations or comments in the file will be discarded, however,
          dot comments and file positions will be preserved;
        * the fuzzy flags are preserved.
        Keyword argument:
        ``refpot``
            object POFile, the reference catalog.
        # Store entries in dict/set for faster access
        self_entries = {entry.msgid: entry for entry in self}
        refpot_msgids = {entry.msgid for entry in refpot}
        # Merge entries that are in the refpot
        for entry in refpot:
            e = self_entries.get(entry.msgid)
            if e is None:
                e = POEntry()
                self.append(e)
            e.merge(entry)
        # ok, now we must "obsolete" entries that are not in the refpot anymore
        for entry in self:
            if entry.msgid not in refpot_msgids:
                entry.obsolete = True
# }}}
# class MOFile {{{

class MOFile(_BaseFile):
    Mo file reader/writer.
    This class inherits the :class:`~polib._BaseFile` class and, by
    extension, the python ``list`` type.
    """

    MAGIC = 0x950412DE
    MAGIC_SWAPPED = 0xDE120495

    def __init__(self, *args, **kwargs):
        """
        Constructor, accepts all keywords arguments accepted by
        :class:`~polib._BaseFile` class.
        """
        _BaseFile.__init__(self, *args, **kwargs)
        self.magic_number = None
        self.version = 0

    def save_as_pofile(self, fpath):
        """
        Saves the mofile as a pofile to ``fpath``.
        Keyword argument:

        ``fpath``
            string, full or relative path to the file.
        """
        _BaseFile.save(self, fpath)

    def save(self, fpath=None):
        Saves the mofile to ``fpath``.
        Keyword argument:

        ``fpath``
            string, full or relative path to the file.
        """
        _BaseFile.save(self, fpath, 'to_binary')

    def percent_translated(self):
        """
        Convenience method to keep the same interface with POFile instances.
        """
        return 100

    def translated_entries(self):
        """
        Convenience method to keep the same interface with POFile instances.
        """
        return self

    def untranslated_entries(self):
        """
        Convenience method to keep the same interface with POFile instances.
        """
        return []

    def fuzzy_entries(self):
        """
        Convenience method to keep the same interface with POFile instances.
        """
        return []

    def obsolete_entries(self):
        """
        Convenience method to keep the same interface with POFile instances.
        """
        return []
# }}}
# class _BaseEntry {{{

class _BaseEntry:
    Base class for :class:`~polib.POEntry` and :class:`~polib.MOEntry` classes.
    This class should **not** be instanciated directly.
    """

    def __init__(self, *args, **kwargs):
        """
        Constructor, accepts the following keyword arguments:

        ``msgid``
            string, the entry msgid.

        ``msgstr``
            string, the entry msgstr.

        ``msgid_plural``
            string, the entry msgid_plural.

        ``msgstr_plural``
            list, the entry msgstr_plural lines.

        ``msgctxt``
            string, the entry context (msgctxt).

        ``obsolete``
            bool, whether the entry is "obsolete" or not.

        ``encoding``
            string, the encoding to use, defaults to ``default_encoding``
            global variable (optional).
        """
        self.msgid = kwargs.get('msgid', '')
        self.msgstr = kwargs.get('msgstr', '')
        self.msgid_plural = kwargs.get('msgid_plural', '')
        self.msgstr_plural = kwargs.get('msgstr_plural', {})
        self.msgctxt = kwargs.get('msgctxt', None)
        self.obsolete = kwargs.get('obsolete', False)
        self.encoding = kwargs.get('encoding', default_encoding)

    def __unicode__(self, wrapwidth=78):
        Returns the unicode representation of the entry.
        """
        if self.obsolete:
            delflag = '#~ '
        else:
            delflag = ''
        ret = []
        # write the msgctxt if any
        if self.msgctxt is not None:
            ret += self._str_field(
                "msgctxt", delflag, "", self.msgctxt, wrapwidth
            )
        # write the msgid
        ret += self._str_field("msgid", delflag, "", self.msgid, wrapwidth)
        # write the msgid_plural if any
        if self.msgid_plural:
            ret += self._str_field(
                "msgid_plural", delflag, "", self.msgid_plural, wrapwidth
            )
        if self.msgstr_plural:
            # write the msgstr_plural if any
            msgstrs = self.msgstr_plural
            keys = list(msgstrs)
            keys.sort()
            for index in keys:
                msgstr = msgstrs[index]
                plural_index = '[%s]' % index
                ret += self._str_field(
                    "msgstr", delflag, plural_index, msgstr, wrapwidth
                )
        else:
            # otherwise write the msgstr
            ret += self._str_field(
                "msgstr", delflag, "", self.msgstr, wrapwidth
            )
        ret.append('')
        def __str__(self):
            return self.__unicode__()
        def __str__(self):
            """
            Returns the string representation of the entry.
            """
            return unicode(self).encode(self.encoding)
    def __eq__(self, other):
    def _str_field(self, fieldname, delflag, plural_index, field, wrapwidth=78):
        lines = field.splitlines(True)
        if len(lines) > 1:
            lines = [''] + lines  # start with initial empty line
            escaped_field = escape(field)
            specialchars_count = 0
            for c in ['\\', '\n', '\r', '\t', '"']:
                specialchars_count += field.count(c)
            # comparison must take into account fieldname length + one space
            # + 2 quotes (eg. msgid "<string>")
            flength = len(fieldname) + 3
            if plural_index:
                flength += len(plural_index)
            real_wrapwidth = wrapwidth - flength + specialchars_count
            if wrapwidth > 0 and len(field) > real_wrapwidth:
                # Wrap the line but take field name into account
                lines = [''] + [
                    unescape(item)
                    for item in wrap(
                        escaped_field,
                        wrapwidth - 2,  # 2 for quotes ""
                        drop_whitespace=False,
                        break_long_words=False,
                    )
                ]
                lines = [field]
        if fieldname.startswith('previous_'):
            # quick and dirty trick to get the real field name
            fieldname = fieldname[9:]

        ret = [
            '%s%s%s "%s"'
            % (delflag, fieldname, plural_index, escape(lines.pop(0)))
        ]
        for line in lines:
            ret.append('%s"%s"' % (delflag, escape(line)))
# }}}
# class POEntry {{{

class POEntry(_BaseEntry):
    """
    Represents a po file entry.
    """

    def __init__(self, *args, **kwargs):
        """
        Constructor, accepts the following keyword arguments:

        ``comment``
            string, the entry comment.

        ``tcomment``
            string, the entry translator comment.

        ``occurrences``
            list, the entry occurrences.

        ``flags``
            list, the entry flags.

        ``previous_msgctxt``
            string, the entry previous context.

        ``previous_msgid``
            string, the entry previous msgid.

        ``previous_msgid_plural``
            string, the entry previous msgid_plural.

        ``linenum``
            integer, the line number of the entry
        _BaseEntry.__init__(self, *args, **kwargs)
        self.comment = kwargs.get('comment', '')
        self.tcomment = kwargs.get('tcomment', '')
        self.occurrences = kwargs.get('occurrences', [])
        self.flags = kwargs.get('flags', [])
        self.previous_msgctxt = kwargs.get('previous_msgctxt', None)
        self.previous_msgid = kwargs.get('previous_msgid', None)