static void test_mail_html2text(void)
{
string_t *str = t_str_new(128);
struct mail_html2text *ht;
unsigned int i, j;
test_begin("mail_html2text()");
for (i = 0; i < N_ELEMENTS(tests); i++) {
ht = mail_html2text_init(MAIL_HTML2TEXT_FLAG_SKIP_QUOTED);
for (j = 0; tests[i].input[j] != '\0'; j++) {
unsigned char c = tests[i].input[j];
mail_html2text_more(ht, &c, 1, str);
}
test_assert_idx(strcmp(str_c(str), tests[i].output) == 0, i);
mail_html2text_deinit(&ht);
str_truncate(str, 0);
}
/* test without skipping quoted */
ht = mail_html2text_init(0);
mail_html2text_more(ht, (const void *)test_blockquote_input,
strlen(test_blockquote_input), str);
test_assert(strcmp(str_c(str), test_blockquote_output) == 0);
mail_html2text_deinit(&ht);
test_end();
}
static void test_mail_html2text_random(void)
{
string_t *str = t_str_new(128);
struct mail_html2text *ht;
test_begin("mail_html2text() random");
for (unsigned int i = 0; i < 1000; i++) {
char valid_chars[] = { '0', 'a', '<', '>', '&', ';', '\\', '\'', '"', '/' };
unsigned char s[2];
ht = mail_html2text_init(0);
for (unsigned int i = 0; i < 100; i++) {
s[0] = valid_chars[rand() % N_ELEMENTS(valid_chars)];
s[1] = valid_chars[rand() % N_ELEMENTS(valid_chars)];
mail_html2text_more(ht, s, rand()%2+1, str);
}
mail_html2text_deinit(&ht);
str_truncate(str, 0);
}
test_end();
}
static bool snippet_generate(struct snippet_context *ctx,
const unsigned char *data, size_t size)
{
size_t i, count;
if (ctx->html2text != NULL) {
buffer_set_used_size(ctx->plain_output, 0);
mail_html2text_more(ctx->html2text, data, size,
ctx->plain_output);
data = ctx->plain_output->data;
size = ctx->plain_output->used;
}
/* message-decoder should feed us only valid and complete
UTF-8 input */
for (i = 0; i < size; i += count) {
count = 1;
switch (ctx->state) {
case SNIPPET_STATE_NEWLINE:
if (data[i] == '>' && ctx->html2text == NULL) {
ctx->state = SNIPPET_STATE_QUOTED;
break;
}
ctx->state = SNIPPET_STATE_NORMAL;
/* fallthrough */
case SNIPPET_STATE_NORMAL:
if (size-i >= 3 &&
((data[i] == 0xEF && data[i+1] == 0xBB && data[i+2] == 0xBF) ||
(data[i] == 0xBF && data[i+1] == 0xBB && data[i+2] == 0xEF))) {
count += 2; /* because we skip +1 next */
break;
}
if (data[i] == '\r' || data[i] == '\n' ||
data[i] == '\t' || data[i] == ' ') {
/* skip any leading whitespace */
if (str_len(ctx->snippet) > 1)
ctx->add_whitespace = TRUE;
if (data[i] == '\n')
ctx->state = SNIPPET_STATE_NEWLINE;
break;
}
if (ctx->add_whitespace) {
str_append_c(ctx->snippet, ' ');
ctx->add_whitespace = FALSE;
if (ctx->chars_left-- == 0)
return FALSE;
}
if (ctx->chars_left-- == 0)
return FALSE;
count = uni_utf8_char_bytes(data[i]);
i_assert(i + count <= size);
str_append_n(ctx->snippet, data + i, count);
break;
case SNIPPET_STATE_QUOTED:
if (data[i] == '\n')
ctx->state = SNIPPET_STATE_NEWLINE;
break;
}
}
return TRUE;
}
