Interview Code Reading Questions

is_binary.c

static void gather_stats(const char *buf, unsigned long size, struct text_stat *stats)
{
	unsigned long i;

	memset(stats, 0, sizeof(*stats));

	for (i = 0; i < size; i++) {
		unsigned char c = buf[i];
		if (c == '\r') {
			stats->cr++;
			if (i+1 < size && buf[i+1] == '\n')
				stats->crlf++;
			continue;
		}
		if (c == '\n') {
			stats->lf++;
			continue;
		}
		if (c == 127)
			/* DEL */
			stats->nonprintable++;
		else if (c < 32) {
			switch (c) {
				/* BS, HT, ESC and FF */
			case '\b': case '\t': case '\033': case '\014':
				stats->printable++;
				break;
			case 0:
				stats->nul++;
				/* fall through */
			default:
				stats->nonprintable++;
			}
		}
		else
			stats->printable++;
	}

	if (size >= 1 && buf[size-1] == '\032')
		stats->nonprintable--;
}


static int is_binary(unsigned long size, struct text_stat *stats)
{

	if (stats->nul)
		return 1;
	if ((stats->printable >> 7) < stats->nonprintable)
		return 1;
	/*
	 * Other heuristics? Average line length might be relevant,
	 * as might LF vs CR vs CRLF counts..
	 *
	 * NOTE! It might be normal to have a low ratio of CRLF to LF
	 * (somebody starts with a LF-only file and edits it with an editor
	 * that adds CRLF only to lines that are added..). But do  we
	 * want to support CR-only? Probably not.
	 */
	return 0;
}

ring_buffer.c

struct ring_buffer *rb_alloc(int nr_pages, long watermark, int cpu, int flags)
{
	struct ring_buffer *rb;
	unsigned long size;
	int i;

	size = sizeof(struct ring_buffer);
	size += nr_pages * sizeof(void *);

	rb = kzalloc(size, GFP_KERNEL);
	if (!rb)
		goto fail;

	rb->user_page = perf_mmap_alloc_page(cpu);
	if (!rb->user_page)
		goto fail_user_page;

	for (i = 0; i < nr_pages; i++) {
		rb->data_pages[i] = perf_mmap_alloc_page(cpu);
		if (!rb->data_pages[i])
			goto fail_data_pages;
	}

	rb->nr_pages = nr_pages;

	ring_buffer_init(rb, watermark, flags);

	return rb;

fail_data_pages:
	for (i--; i >= 0; i--)
		free_page((unsigned long)rb->data_pages[i]);

	free_page((unsigned long)rb->user_page);

fail_user_page:
	kfree(rb);

fail:
	return NULL;
}

splitargs.py

def split_args(args):
    '''
    Splits args on whitespace, but intelligently reassembles
    those that may have been split over a jinja2 block or quotes.

    Basically this is a variation shlex that has some more intelligence for
    how Ansible needs to use it.
    '''

    params = []
    args = args.strip()
    try:
        args = args.encode('utf-8')
        do_decode = True
    except UnicodeDecodeError:
        do_decode = False
    items = args.split('\n')

    quote_char = None
    inside_quotes = False
    print_depth   = 0 # used to count nested jinja2 {{ }} blocks
    block_depth   = 0 # used to count nested jinja2 {% %} blocks
    comment_depth = 0 # used to count nested jinja2 {# #} blocks

    for itemidx,item in enumerate(items):

        # we split on spaces and newlines separately, so that we
        # can tell which character we split on for reassembly
        # inside quotation characters
        tokens = item.strip().split(' ')

        line_continuation = False
        for idx,token in enumerate(tokens):
            if token == '\\' and not inside_quotes:
                line_continuation = True
                continue

            # store the previous quoting state for checking later
            was_inside_quotes = inside_quotes
            quote_char = _get_quote_state(token, quote_char)
            inside_quotes = quote_char is not None

            appended = False

            if inside_quotes and not was_inside_quotes:
                params.append(token)
                appended = True
            elif print_depth or block_depth or comment_depth or inside_quotes or was_inside_quotes:
                if idx == 0 and not inside_quotes and was_inside_quotes:
                    params[-1] = "%s%s" % (params[-1], token)
                elif len(tokens) > 1:
                    spacer = ''
                    if idx > 0:
                        spacer = ' '
                    params[-1] = "%s%s%s" % (params[-1], spacer, token)
                else:
                    spacer = ''
                    if not params[-1].endswith('\n') and idx == 0:
                        spacer = '\n'
                    params[-1] = "%s%s%s" % (params[-1], spacer, token)
                appended = True

            # if the number of paired block tags is not the same, the depth has changed, so we calculate that here
            # and may append the current token to the params (if we haven't previously done so)
            prev_print_depth = print_depth
            print_depth = _count_jinja2_blocks(token, print_depth, "{{", "}}")
            if print_depth != prev_print_depth and not appended:
                params.append(token)
                appended = True

            prev_block_depth = block_depth
            block_depth = _count_jinja2_blocks(token, block_depth, "{%", "%}")
            if block_depth != prev_block_depth and not appended:
                params.append(token)
                appended = True

            prev_comment_depth = comment_depth
            comment_depth = _count_jinja2_blocks(token, comment_depth, "{#", "#}")
            if comment_depth != prev_comment_depth and not appended:
                params.append(token)
                appended = True

            if not (print_depth or block_depth or comment_depth) and not inside_quotes and not appended and token != '':
                params.append(token)

        if len(items) > 1 and itemidx != len(items) - 1 and not line_continuation:
            if not params[-1].endswith('\n') or item == '':
                params[-1] += '\n'

        line_continuation = False

    if print_depth or block_depth or comment_depth or inside_quotes:
        raise Exception("error while splitting arguments, either an unbalanced jinja2 block or quotes")

    if do_decode:
        params = [x.decode('utf-8') for x in params]

    return params