aboutsummaryrefslogtreecommitdiffstats
path: root/src/libmime
diff options
context:
space:
mode:
Diffstat (limited to 'src/libmime')
-rw-r--r--src/libmime/images.c2
-rw-r--r--src/libmime/message.c38
-rw-r--r--src/libmime/message.h9
-rw-r--r--src/libmime/mime_expressions.c23
4 files changed, 41 insertions, 31 deletions
diff --git a/src/libmime/images.c b/src/libmime/images.c
index c54f5845a..b3baa8e4c 100644
--- a/src/libmime/images.c
+++ b/src/libmime/images.c
@@ -684,7 +684,7 @@ rspamd_image_process_part (struct rspamd_task *task, struct rspamd_mime_part *pa
}
PTR_ARRAY_FOREACH (MESSAGE_FIELD (task, text_parts), i, tp) {
- if (IS_PART_HTML (tp) && tp->html != NULL &&
+ if (IS_TEXT_PART_HTML (tp) && tp->html != NULL &&
tp->html->images != NULL) {
for (j = 0; j < tp->html->images->len; j ++) {
himg = g_ptr_array_index (tp->html->images, j);
diff --git a/src/libmime/message.c b/src/libmime/message.c
index 2702d0f51..411b872c9 100644
--- a/src/libmime/message.c
+++ b/src/libmime/message.c
@@ -147,7 +147,7 @@ rspamd_mime_part_create_words (struct rspamd_task *task,
{
enum rspamd_tokenize_type tok_type;
- if (IS_PART_UTF (part)) {
+ if (IS_TEXT_PART_UTF (part)) {
#if U_ICU_VERSION_MAJOR_NUM < 50
/* Hack to prevent hang with Thai in old libicu */
@@ -209,8 +209,8 @@ rspamd_mime_part_detect_language (struct rspamd_task *task,
{
struct rspamd_lang_detector_res *lang;
- if (!IS_PART_EMPTY (part) && part->utf_words && part->utf_words->len > 0 &&
- task->lang_det) {
+ if (!IS_TEXT_PART_EMPTY (part) && part->utf_words && part->utf_words->len > 0 &&
+ task->lang_det) {
if (rspamd_language_detector_detect (task, task->lang_det, part)) {
lang = g_ptr_array_index (part->languages, 0);
part->language = lang->lang;
@@ -240,7 +240,7 @@ rspamd_strip_newlines_parse (struct rspamd_task *task,
} state = normal_char;
while (p < pe) {
- if (IS_PART_UTF (part)) {
+ if (IS_TEXT_PART_UTF (part)) {
gint32 off = p - begin;
U8_NEXT (begin, off, pe - begin, uc);
@@ -324,7 +324,7 @@ rspamd_strip_newlines_parse (struct rspamd_task *task,
c = p + 1;
- if (IS_PART_HTML (part) || !url_open_bracket) {
+ if (IS_TEXT_PART_HTML (part) || !url_open_bracket) {
g_byte_array_append (part->utf_stripped_content,
(const guint8 *)" ", 1);
g_ptr_array_add (part->newlines,
@@ -339,7 +339,7 @@ rspamd_strip_newlines_parse (struct rspamd_task *task,
case seen_cr:
/* \r\n */
if (!crlf_added) {
- if (IS_PART_HTML (part) || !url_open_bracket) {
+ if (IS_TEXT_PART_HTML (part) || !url_open_bracket) {
g_byte_array_append (part->utf_stripped_content,
(const guint8 *) " ", 1);
crlf_added = TRUE;
@@ -509,7 +509,7 @@ rspamd_normalize_text_part (struct rspamd_task *task,
part->newlines = g_ptr_array_sized_new (128);
- if (IS_PART_EMPTY (part)) {
+ if (IS_TEXT_PART_EMPTY (part)) {
part->utf_stripped_content = g_byte_array_new ();
}
else {
@@ -532,7 +532,7 @@ rspamd_normalize_text_part (struct rspamd_task *task,
}
}
- if (IS_PART_UTF (part)) {
+ if (IS_TEXT_PART_UTF (part)) {
utext_openUTF8 (&part->utf_stripped_text,
part->utf_stripped_content->data,
part->utf_stripped_content->len,
@@ -780,6 +780,7 @@ rspamd_message_process_text_part_maybe (struct rspamd_task *task,
struct rspamd_mime_text_part *text_part;
rspamd_ftok_t html_tok, xhtml_tok;
gboolean found_html = FALSE, found_txt = FALSE;
+ guint flags = 0;
enum rspamd_action_type act;
if ((mime_part->ct && (mime_part->ct->flags & RSPAMD_CONTENT_TYPE_TEXT)) ||
@@ -802,10 +803,14 @@ rspamd_message_process_text_part_maybe (struct rspamd_task *task,
/* Skip attachments */
if ((found_txt || found_html) &&
- (mime_part->cd && mime_part->cd->type == RSPAMD_CT_ATTACHMENT) &&
- (!task->cfg->check_text_attachements)) {
- debug_task ("skip attachments for checking as text parts");
- return FALSE;
+ (mime_part->cd && mime_part->cd->type == RSPAMD_CT_ATTACHMENT)) {
+ if (!task->cfg->check_text_attachements) {
+ debug_task ("skip attachments for checking as text parts");
+ return FALSE;
+ }
+ else {
+ flags |= RSPAMD_MIME_TEXT_PART_ATTACHMENT;
+ }
}
else if (!(found_txt || found_html)) {
/* Not a text part */
@@ -820,6 +825,7 @@ rspamd_message_process_text_part_maybe (struct rspamd_task *task,
text_part->parsed.begin = mime_part->parsed_data.begin;
text_part->parsed.len = mime_part->parsed_data.len;
text_part->utf_stripped_text = (UText)UTEXT_INITIALIZER;
+ text_part->flags |= flags;
if (found_html) {
if (!rspamd_message_process_html_text_part (task, text_part)) {
@@ -859,7 +865,7 @@ rspamd_message_process_text_part_maybe (struct rspamd_task *task,
/* Post process part */
rspamd_normalize_text_part (task, text_part);
- if (!IS_PART_HTML (text_part)) {
+ if (!IS_TEXT_PART_HTML (text_part)) {
if (mime_part->parent_part) {
struct rspamd_mime_part *parent = mime_part->parent_part;
@@ -1509,7 +1515,7 @@ rspamd_message_process (struct rspamd_task *task)
srch.len = 11;
if (rspamd_ftok_cmp (&p1->mime_part->parent_part->ct->subtype, &srch) == 0) {
- if (!IS_PART_EMPTY (p1) && !IS_PART_EMPTY (p2) &&
+ if (!IS_TEXT_PART_EMPTY (p1) && !IS_TEXT_PART_EMPTY (p2) &&
p1->normalized_hashes && p2->normalized_hashes) {
/*
* We also detect language on one part and propagate it to
@@ -1518,10 +1524,10 @@ rspamd_message_process (struct rspamd_task *task)
struct rspamd_mime_text_part *sel;
/* Prefer HTML as text part is not displayed normally */
- if (IS_PART_HTML (p1)) {
+ if (IS_TEXT_PART_HTML (p1)) {
sel = p1;
}
- else if (IS_PART_HTML (p2)) {
+ else if (IS_TEXT_PART_HTML (p2)) {
sel = p2;
}
else {
diff --git a/src/libmime/message.h b/src/libmime/message.h
index d6f1b76c0..13e40e2ef 100644
--- a/src/libmime/message.h
+++ b/src/libmime/message.h
@@ -120,11 +120,12 @@ struct rspamd_mime_part {
#define RSPAMD_MIME_TEXT_PART_FLAG_8BIT_ENCODED (1 << 5)
#define RSPAMD_MIME_TEXT_PART_HAS_SUBNORMAL (1 << 6)
#define RSPAMD_MIME_TEXT_PART_NORMALISED (1 << 7)
+#define RSPAMD_MIME_TEXT_PART_ATTACHMENT (1 << 8)
-#define IS_PART_EMPTY(part) ((part)->flags & RSPAMD_MIME_TEXT_PART_FLAG_EMPTY)
-#define IS_PART_UTF(part) ((part)->flags & RSPAMD_MIME_TEXT_PART_FLAG_UTF)
-#define IS_PART_RAW(part) (!((part)->flags & RSPAMD_MIME_TEXT_PART_FLAG_UTF))
-#define IS_PART_HTML(part) ((part)->flags & RSPAMD_MIME_TEXT_PART_FLAG_HTML)
+#define IS_TEXT_PART_EMPTY(part) ((part)->flags & RSPAMD_MIME_TEXT_PART_FLAG_EMPTY)
+#define IS_TEXT_PART_UTF(part) ((part)->flags & RSPAMD_MIME_TEXT_PART_FLAG_UTF)
+#define IS_TEXT_PART_HTML(part) ((part)->flags & RSPAMD_MIME_TEXT_PART_FLAG_HTML)
+#define IS_TEXT_PART_ATTACHMENT(part) ((part)->flags & RSPAMD_MIME_TEXT_PART_ATTACHMENT)
struct rspamd_mime_text_part {
diff --git a/src/libmime/mime_expressions.c b/src/libmime/mime_expressions.c
index d35bc136f..a06baffaf 100644
--- a/src/libmime/mime_expressions.c
+++ b/src/libmime/mime_expressions.c
@@ -1455,20 +1455,23 @@ rspamd_has_only_html_part (struct rspamd_task * task, GArray * args,
void *unused)
{
struct rspamd_mime_text_part *p;
+ guint i, cnt_html = 0, cnt_txt = 0;
gboolean res = FALSE;
- if (MESSAGE_FIELD (task, text_parts)->len == 1) {
+ PTR_ARRAY_FOREACH (MESSAGE_FIELD (task, text_parts), i, p) {
p = g_ptr_array_index (MESSAGE_FIELD (task, text_parts), 0);
- if (IS_PART_HTML (p)) {
- res = TRUE;
- }
- else {
- res = FALSE;
+ if (!IS_TEXT_PART_ATTACHMENT (p)) {
+ if (IS_TEXT_PART_HTML (p)) {
+ cnt_html++;
+ }
+ else {
+ cnt_txt++;
+ }
}
}
- return res;
+ return (cnt_html > 0 && cnt_txt == 0);
}
static gboolean
@@ -1565,7 +1568,7 @@ rspamd_is_html_balanced (struct rspamd_task * task, GArray * args, void *unused)
gboolean res = TRUE;
PTR_ARRAY_FOREACH (MESSAGE_FIELD (task, text_parts), i, p) {
- if (IS_PART_HTML (p)) {
+ if (IS_TEXT_PART_HTML (p)) {
if (p->flags & RSPAMD_MIME_TEXT_PART_FLAG_BALANCED) {
res = TRUE;
}
@@ -1600,7 +1603,7 @@ rspamd_has_html_tag (struct rspamd_task * task, GArray * args, void *unused)
}
PTR_ARRAY_FOREACH (MESSAGE_FIELD (task, text_parts), i, p) {
- if (IS_PART_HTML (p) && p->html) {
+ if (IS_TEXT_PART_HTML (p) && p->html) {
res = rspamd_html_tag_seen (p->html, arg->data);
}
@@ -1621,7 +1624,7 @@ rspamd_has_fake_html (struct rspamd_task * task, GArray * args, void *unused)
gboolean res = FALSE;
PTR_ARRAY_FOREACH (MESSAGE_FIELD (task, text_parts), i, p) {
- if (IS_PART_HTML (p) && (p->html == NULL || p->html->html_tags == NULL)) {
+ if (IS_TEXT_PART_HTML (p) && (p->html == NULL || p->html->html_tags == NULL)) {
res = TRUE;
}
generated by cgit v1.2.3 (git 2.39.1) at 2025-01-01 18:41:52 +0000