Beispiel #1
0
def strcoll(space, w_s1, w_s2):
    "string,string -> int. Compares two strings according to the locale."

    if (space.isinstance_w(w_s1, space.w_bytes) and
        space.isinstance_w(w_s2, space.w_bytes)):

        s1, s2 = space.bytes_w(w_s1), space.bytes_w(w_s2)
        s1_c = rffi.str2charp(s1)
        s2_c = rffi.str2charp(s2)
        try:
            return space.newint(_strcoll(s1_c, s2_c))
        finally:
            rffi.free_charp(s1_c)
            rffi.free_charp(s2_c)

    s1, l1 = space.utf8_len_w(w_s1)
    s2, l2 = space.utf8_len_w(w_s2)

    s1_c = rffi.utf82wcharp(s1, l1)
    s2_c = rffi.utf82wcharp(s2, l2)
    try:
        result = _wcscoll(s1_c, s2_c)
    finally:
        rffi.free_wcharp(s1_c)
        rffi.free_wcharp(s2_c)

    return space.newint(result)
Beispiel #2
0
def encodeex(encodebuf,
             utf8data,
             length,
             errors="strict",
             errorcb=None,
             namecb=None,
             ignore_error=0):
    inleft = length
    inbuf = rffi.utf82wcharp(utf8data, length)
    try:
        if pypy_cjk_enc_init(encodebuf, inbuf, inleft) < 0:
            raise MemoryError
        if ignore_error == 0:
            flags = MBENC_FLUSH | MBENC_RESET
        else:
            flags = 0
        while True:
            r = pypy_cjk_enc_chunk(encodebuf, flags)
            if r == 0 or r == ignore_error:
                break
            multibytecodec_encerror(encodebuf, r, errors, errorcb, namecb,
                                    utf8data)
        while flags & MBENC_RESET:
            r = pypy_cjk_enc_reset(encodebuf)
            if r == 0:
                break
            multibytecodec_encerror(encodebuf, r, errors, errorcb, namecb,
                                    utf8data)
        src = pypy_cjk_enc_outbuf(encodebuf)
        length = pypy_cjk_enc_outlen(encodebuf)
        return rffi.charpsize2str(src, length)
    finally:
        lltype.free(inbuf, flavor='raw')
Beispiel #3
0
def multibytecodec_decerror(decodebuf, e, errors, errorcb, namecb, stringdata):
    if e > 0:
        reason = "illegal multibyte sequence"
        esize = e
    elif e == MBERR_TOOFEW:
        reason = "incomplete multibyte sequence"
        esize = pypy_cjk_dec_inbuf_remaining(decodebuf)
    elif e == MBERR_NOMEMORY:
        raise MemoryError
    else:
        raise RuntimeError
    #
    # compute the unicode to use as a replacement -> 'replace', and
    # the current position in the input 'unicodedata' -> 'end'
    start = pypy_cjk_dec_inbuf_consumed(decodebuf)
    end = start + esize
    if errors == "strict":
        raise EncodeDecodeError(start, end, reason)
    elif errors == "ignore":
        replace = ""
    elif errors == "replace":
        replace = UNICODE_REPLACEMENT_CHARACTER
    else:
        assert errorcb
        replace, end, rettype, obj = errorcb(errors, namecb, reason,
                                             stringdata, start, end)
        # 'replace' is UTF8 encoded unicode, rettype is 'u'
    lgt = rutf8.codepoints_in_utf8(replace)
    inbuf = rffi.utf82wcharp(replace, lgt)
    try:
        r = pypy_cjk_dec_replace_on_error(decodebuf, inbuf, lgt, end)
    finally:
        lltype.free(inbuf, flavor='raw')
    if r == MBERR_NOMEMORY:
        raise MemoryError
Beispiel #4
0
def strcoll(space, w_s1, w_s2):
    "string,string -> int. Compares two strings according to the locale."

    s1, l1 = space.utf8_len_w(w_s1)
    s2, l2 = space.utf8_len_w(w_s2)
    if '\x00' in s1 or '\x00' in s2:
        raise oefmt(space.w_ValueError, "embedded null character")

    s1_c = rffi.utf82wcharp(s1, l1)
    s2_c = rffi.utf82wcharp(s2, l2)
    try:
        result = _wcscoll(s1_c, s2_c)
    finally:
        rffi.free_wcharp(s1_c)
        rffi.free_wcharp(s2_c)

    return space.newint(result)
Beispiel #5
0
def PyUnicode_AS_UNICODE(space, ref):
    """Return a pointer to the internal Py_UNICODE buffer of the object.  ref
    has to be a PyUnicodeObject (not checked)."""
    ref_unicode = rffi.cast(PyUnicodeObject, ref)
    if not ref_unicode.c_str:
        # Copy unicode buffer
        w_unicode = from_ref(space, rffi.cast(PyObject, ref))
        u, length = space.utf8_len_w(w_unicode)
        ref_unicode.c_str = rffi.utf82wcharp(u, length)
    return ref_unicode.c_str
Beispiel #6
0
        def stop_error_capture(self, w_done):
            if w_done is None:
                return

            w_text = self.space.call_function(w_done)
            p = rffi.utf82wcharp(self.space.utf8_w(w_text),
                                 self.space.len_w(w_text),
                                 track_allocation=False)
            if self.text_p:
                rffi.free_wcharp(self.text_p, track_allocation=False)
            self.text_p = p      # keepalive

            cffi_errorbox(p)
Beispiel #7
0
 def get_programname(self):
     if not self.programname:
         space = self.space
         argv = space.sys.get('argv')
         if space.len_w(argv):
             argv0 = space.getitem(argv, space.newint(0))
             progname = space.utf8_w(argv0)
             lgt = space.len_w(argv0)
         else:
             progname = "pypy3"
             lgt = len(progname)
         self.programname = rffi.utf82wcharp(progname, lgt)
         lltype.render_immortal(self.programname)
     return self.programname
Beispiel #8
0
    def descr_init(self,
                   space,
                   w_nameobj,
                   mode='r',
                   closefd=True,
                   w_opener=None):
        name = rffi.cast(rffi.CWCHARP, 0)
        self.fd = -1
        self.handle = rwin32.INVALID_HANDLE_VALUE
        self.readable = False
        self.writable = False
        self.blksize = 0
        rwa = False
        console_type = '\0'
        self.buf = ''

        if space.isinstance_w(w_nameobj, space.w_int):
            self.fd = space.int_w(w_nameobj)
            if self.fd < 0:
                raise oefmt(space.w_ValueError, "negative file descriptor")

        # make the flow analysis happy,otherwise it thinks w_path
        # is undefined later
        w_path = w_nameobj
        if self.fd < 0:
            from pypy.module.posix.interp_posix import fspath
            w_path = fspath(space, w_nameobj)
            console_type = _pyio_get_console_type(space, w_path)
            if not console_type:
                raise oefmt(space.w_ValueError, "Invalid console type")
            if console_type == '\0':
                raise oefmt(space.w_ValueError, "Cannot open non-console file")
        self.mode = 'u'
        for char in mode:
            if char in "+ax":
                # OK do nothing
                pass
            elif char == "b":
                self.mode = 'b'
            elif char == "r":
                if rwa:
                    raise oefmt(space.w_ValueError, "invalid mode: %s", mode)
                rwa = True
                self.readable = True
                if console_type == "x":
                    console_type = "r"
            elif char == "w":
                if rwa:
                    raise oefmt(space.w_ValueError, "invalid mode: %s", mode)
                rwa = True
                self.writable = True
                if console_type == 'x':
                    console_type = 'w'
            else:
                raise oefmt(space.w_ValueError, "invalid mode: %s", mode)
        if not rwa:
            raise oefmt(space.w_ValueError,
                        "Must have exactly one of read or write mode")

        if self.fd >= 0:
            self.handle = rwin32.get_osfhandle(self.fd)
            self.closehandle = False
        else:
            access = rwin32.GENERIC_READ
            self.closehandle = True
            if not closefd:
                raise oefmt(space.w_ValueError,
                            "Cannot use closefd=False with a file name")
            if self.writable:
                access = rwin32.GENERIC_WRITE

            traits = _preferred_traits(space.realunicode_w(w_path))
            if not (traits.str is unicode):
                raise oefmt(space.w_ValueError, "Non-unicode string name %s",
                            traits.str)
            win32traits = make_win32_traits(traits)

            pathlen = space.len_w(w_path)
            name = rffi.utf82wcharp(space.utf8_w(w_path), pathlen)
            self.handle = win32traits.CreateFile(
                name, rwin32.GENERIC_READ | rwin32.GENERIC_WRITE,
                rwin32.FILE_SHARE_READ | rwin32.FILE_SHARE_WRITE, rffi.NULL,
                win32traits.OPEN_EXISTING, 0, rffi.cast(rwin32.HANDLE, 0))
            if self.handle == rwin32.INVALID_HANDLE_VALUE:
                self.handle = win32traits.CreateFile(
                    name, access,
                    rwin32.FILE_SHARE_READ | rwin32.FILE_SHARE_WRITE,
                    rffi.NULL, win32traits.OPEN_EXISTING, 0,
                    rffi.cast(rwin32.HANDLE, 0))
            lltype.free(name, flavor='raw')

            if self.handle == rwin32.INVALID_HANDLE_VALUE:
                raise WindowsError(rwin32.GetLastError_saved(),
                                   "Failed to open handle")

        if console_type == '\0':
            console_type = _get_console_type(self.handle)

        if console_type == '\0':
            raise oefmt(space.w_ValueError, "Cannot open non-console file")

        if self.writable and console_type != 'w':
            raise oefmt(space.w_ValueError,
                        "Cannot open input buffer for writing")

        if self.readable and console_type != 'r':
            raise oefmt(space.w_ValueError,
                        "Cannot open output buffer for reading")

        self.blksize = DEFAULT_BUFFER_SIZE
Beispiel #9
0
 def handle_unichar_p(self, w_ffitype, w_obj, utf8val, utf8len):
     buf = rffi.utf82wcharp(utf8val, utf8len)
     self.w_func.to_free.append(rffi.cast(rffi.VOIDP, buf))
     addr = rffi.cast(rffi.ULONG, buf)
     self.argchain.arg(addr)