commit 90fb3cc: [Fix] Fix embedded images linking logic

[Vsevolod Stakhov]

Author: Vsevolod Stakhov
Date: 2019-09-19 12:07:03 +0100
URL: https://github.com/rspamd/rspamd/commit/90fb3cc2868a3ee4d5995aefb71fb6d66b450edf (HEAD -> master)

[Fix] Fix embedded images linking logic

---
 rules/html.lua        |  1 -
 src/libmime/images.c  | 84 +++++++++++++++++++++++++++++++--------------------
 src/libmime/images.h  |  5 +++
 src/libmime/message.c |  1 +
 4 files changed, 58 insertions(+), 33 deletions(-)

diff --git a/rules/html.lua b/rules/html.lua
index fb32e9179..9e582a731 100644
--- a/rules/html.lua
+++ b/rules/html.lua
@@ -101,7 +101,6 @@ rspamd_config.R_EMPTY_IMAGE = {
       if p:is_html() then -- if the current part is html part
         local hc = p:get_html() -- we get HTML context
         local len = p:get_length() -- and part's length
-
         if hc and len < 50 then -- if we have a part that has less than 50 bytes of text
           local images = hc:get_images() -- then we check for HTML images
 
diff --git a/src/libmime/images.c b/src/libmime/images.c
index cb59bc88e..073776359 100644
--- a/src/libmime/images.c
+++ b/src/libmime/images.c
@@ -596,14 +596,8 @@ rspamd_maybe_process_image (rspamd_mempool_t *pool,
 static void
 process_image (struct rspamd_task *task, struct rspamd_mime_part *part)
 {
-	struct rspamd_mime_header *rh;
-	struct rspamd_mime_text_part *tp;
-	struct html_image *himg;
-	const gchar *cid, *html_cid;
-	guint cid_len, i, j;
 	struct rspamd_image *img;
 
-
 	img = rspamd_maybe_process_image (task->task_pool, &part->parsed_data);
 
 	if (img != NULL) {
@@ -619,7 +613,45 @@ process_image (struct rspamd_task *task, struct rspamd_mime_part *part)
 
 		part->flags |= RSPAMD_MIME_PART_IMAGE;
 		part->specific.img = img;
+	}
+}
+
+const gchar *
+rspamd_image_type_str (enum rspamd_image_type type)
+{
+	switch (type) {
+	case IMAGE_TYPE_PNG:
+		return "PNG";
+		break;
+	case IMAGE_TYPE_JPG:
+		return "JPEG";
+		break;
+	case IMAGE_TYPE_GIF:
+		return "GIF";
+		break;
+	case IMAGE_TYPE_BMP:
+		return "BMP";
+		break;
+	default:
+		break;
+	}
 
+	return "unknown";
+}
+
+static void
+rspamd_image_process_part (struct rspamd_task *task, struct rspamd_mime_part *part)
+{
+	struct rspamd_mime_header *rh;
+	struct rspamd_mime_text_part *tp;
+	struct html_image *himg;
+	const gchar *cid, *html_cid;
+	guint cid_len, i, j;
+	struct rspamd_image *img;
+
+	img = (struct rspamd_image *)part->specific.img;
+
+	if (img) {
 		/* Check Content-Id */
 		rh = rspamd_message_get_header_from_hash (part->raw_headers,
 				"Content-Id");
@@ -638,15 +670,14 @@ process_image (struct rspamd_task *task, struct rspamd_mime_part *part)
 					cid_len --;
 				}
 
-
 				PTR_ARRAY_FOREACH (MESSAGE_FIELD (task, text_parts), i, tp) {
 					if (IS_PART_HTML (tp) && tp->html != NULL &&
-							tp->html->images != NULL) {
+						tp->html->images != NULL) {
 						for (j = 0; j < tp->html->images->len; j ++) {
 							himg = g_ptr_array_index (tp->html->images, j);
 
 							if ((himg->flags & RSPAMD_HTML_FLAG_IMAGE_EMBEDDED) &&
-									himg->src) {
+								himg->src) {
 								html_cid = himg->src;
 
 								if (strncmp (html_cid, "cid:", 4) == 0) {
@@ -654,7 +685,7 @@ process_image (struct rspamd_task *task, struct rspamd_mime_part *part)
 								}
 
 								if (strlen (html_cid) == cid_len &&
-										memcmp (html_cid, cid, cid_len) == 0) {
+									memcmp (html_cid, cid, cid_len) == 0) {
 									img->html_image = himg;
 									himg->embedded_image = img;
 
@@ -664,6 +695,7 @@ process_image (struct rspamd_task *task, struct rspamd_mime_part *part)
 									if (himg->height == 0) {
 										himg->height = img->height;
 									}
+
 									if (himg->width == 0) {
 										himg->width = img->width;
 									}
@@ -672,32 +704,20 @@ process_image (struct rspamd_task *task, struct rspamd_mime_part *part)
 						}
 					}
 				}
-
 			}
-
 		}
 	}
 }
 
-const gchar *
-rspamd_image_type_str (enum rspamd_image_type type)
+void
+rspamd_images_link (struct rspamd_task *task)
 {
-	switch (type) {
-	case IMAGE_TYPE_PNG:
-		return "PNG";
-		break;
-	case IMAGE_TYPE_JPG:
-		return "JPEG";
-		break;
-	case IMAGE_TYPE_GIF:
-		return "GIF";
-		break;
-	case IMAGE_TYPE_BMP:
-		return "BMP";
-		break;
-	default:
-		break;
-	}
+	struct rspamd_mime_part *part;
+	guint i;
 
-	return "unknown";
-}
+	PTR_ARRAY_FOREACH (MESSAGE_FIELD (task, parts), i, part) {
+		if (part->flags & RSPAMD_MIME_PART_IMAGE) {
+			rspamd_image_process_part (task, part);
+		}
+	}
+}
\ No newline at end of file
diff --git a/src/libmime/images.h b/src/libmime/images.h
index bf1175b2a..964a1b35a 100644
--- a/src/libmime/images.h
+++ b/src/libmime/images.h
@@ -39,6 +39,11 @@ struct rspamd_image {
  */
 void rspamd_images_process (struct rspamd_task *task);
 
+/*
+ * Link embedded images to the HTML parts
+ */
+void rspamd_images_link (struct rspamd_task *task);
+
 /**
  * Processes image in raw data
  * @param task
diff --git a/src/libmime/message.c b/src/libmime/message.c
index 0f260d726..cb95945fe 100644
--- a/src/libmime/message.c
+++ b/src/libmime/message.c
@@ -1570,6 +1570,7 @@ rspamd_message_process (struct rspamd_task *task)
 		}
 	}
 
+	rspamd_images_link (task);
 	rspamd_tokenize_meta_words (task);
 }