Function sz_py_export_string_like

python/stringzilla.c:444–511

* @brief Helper function to export a Python string-like object into a `sz_string_view_t`. * On failure, sets a Python exception and returns 0. */

 *          On failure, sets a Python exception and returns 0.
 */
SZ_DYNAMIC sz_bool_t sz_py_export_string_like(PyObject *object, sz_cptr_t *start, sz_size_t *length) {
    if (PyUnicode_Check(object)) {
        // Handle Python `str` object
        Py_ssize_t signed_length;
        *start = PyUnicode_AsUTF8AndSize(object, &signed_length);
        *length = (sz_size_t)signed_length;
        return 1;
    }
    else if (PyBytes_Check(object)) {
        // Handle Python `bytes` object
        // https://docs.python.org/3/c-api/bytes.html
        Py_ssize_t signed_length;
        if (PyBytes_AsStringAndSize(object, (sz_ptr_t *)start, &signed_length) == -1) {
            PyErr_SetString(PyExc_ValueError, "Couldn't access `bytes` buffer internals");
            return 0;
        }
        *length = (sz_size_t)signed_length;
        return 1;
    }
    else if (PyByteArray_Check(object)) {
        // Handle Python mutable `bytearray` object
        // https://docs.python.org/3/c-api/bytearray.html
        *start = PyByteArray_AS_STRING(object);
        *length = PyByteArray_GET_SIZE(object);
        return 1;
    }
    else if (PyObject_TypeCheck(object, &StrType)) {
        Str *str = (Str *)object;
        *start = str->memory.start;
        *length = str->memory.length;
        return 1;
    }
    else if (PyObject_TypeCheck(object, &FileType)) {
        File *file = (File *)object;
        *start = file->memory.start;
        *length = file->memory.length;
        return 1;
    }
    else if (PyMemoryView_Check(object)) {
        // Handle Python `memoryview` object
        // https://docs.python.org/3/c-api/memoryview.html
        // https://docs.python.org/3/c-api/buffer.html#c.Py_buffer
        Py_buffer *view = PyMemoryView_GET_BUFFER(object);
        // Make sure we are dealing with single-byte integral representations
        if (view->itemsize != 1) {
            PyErr_SetString(PyExc_ValueError, "Only single-byte integral types are supported");
            return 0;
        }
        // Let's make sure the data is contiguous.
        // This can be a bit trickier for high-dimensional arrays, but CPython has a built-in function for that.
        // The flag 'C' stands for C-style-contiguous, which means that the last dimension is contiguous.
        // The flag 'F' stands for Fortran-style-contiguous, which means that the first dimension is contiguous.
        // The flag 'A' stands for any-contiguous, which only means there are no gaps between elements.
        // For byte-level processing that's all we need.
        if (!PyBuffer_IsContiguous(view, 'A')) {
            PyErr_SetString(PyExc_ValueError, "The array must be contiguous");
            return 0;
        }