}
static gboolean
-rspamd_html_process_tag(rspamd_mempool_t *pool,
+html_process_tag(rspamd_mempool_t *pool,
struct html_content *hc,
struct html_tag *tag,
GNode **cur_level,
struct html_content *hc,
struct html_tag *tag,
const char *in,
- struct tag_content_parser_state parser_env)
+ struct tag_content_parser_state &parser_env)
{
enum tag_parser_state {
parse_start = 0,
tag->extra = bl;
}
-
-}
-
-/* Unconverted C part */
-
-
-static gboolean
-rspamd_html_propagate_lengths(GNode *node, gpointer _unused) {
+static auto
+html_propagate_lengths(GNode *node, gpointer _unused) -> gboolean
+{
GNode *child;
struct html_tag *tag = static_cast<html_tag *>(node->data), *cld_tag;
return FALSE;
}
-static void
-rspamd_html_propagate_style(struct html_content *hc,
+static auto
+html_propagate_style(struct html_content *hc,
struct html_tag *tag,
struct html_block *bl,
- GQueue *blocks) {
+ GQueue *blocks) -> void
+{
struct html_block *bl_parent;
gboolean push_block = FALSE;
-
/* Propagate from the parent if needed */
bl_parent = static_cast<html_block *>(g_queue_peek_tail(blocks));
}
}
+using tags_vector = std::vector<std::unique_ptr<struct html_tag>>;
-GByteArray*
-rspamd_html_process_part_full (rspamd_mempool_t *pool,
- struct html_content *hc,
- GByteArray *in,
- GList **exceptions,
- khash_t (rspamd_url_hash) *url_set,
- GPtrArray *part_urls,
- bool allow_css)
+static auto
+tags_vector_ptr_dtor(void *ptr)
{
- const guchar *p, *c, *end, *savep = NULL;
+ auto *ptags = (tags_vector *)ptr;
+
+ delete ptags;
+}
+
+static auto
+html_process_part_full (rspamd_mempool_t *pool,
+ struct html_content *hc,
+ GByteArray *in,
+ GList **exceptions,
+ khash_t (rspamd_url_hash) *url_set,
+ GPtrArray *part_urls,
+ bool allow_css) -> GByteArray*
+{
+ const guchar *p, *c, *end;
guchar t;
gboolean closing = FALSE, need_decode = FALSE, save_space = FALSE,
balanced;
GByteArray *dest;
guint obrace = 0, ebrace = 0;
GNode *cur_level = NULL;
- gint substate = 0, len, href_offset = -1;
- struct html_tag *cur_tag = NULL, *content_tag = NULL;
struct rspamd_url *url = NULL;
+ gint len, href_offset = -1;
+ struct html_tag *cur_tag = NULL, *content_tag = NULL;
GQueue *styles_blocks;
+ struct tag_content_parser_state content_parser_env;
+ tags_vector *all_tags;
enum {
parse_start = 0,
g_assert (hc != NULL);
g_assert (pool != NULL);
+ all_tags = new tags_vector(128);
+ rspamd_mempool_add_destructor(pool, tags_vector_ptr_dtor, all_tags);
+
hc->tags_seen = (guchar *)rspamd_mempool_alloc0 (pool, NBYTES (N_TAGS));
/* Set white background color by default */
continue;
default:
state = tag_content;
- substate = 0;
- savep = NULL;
- cur_tag = rspamd_mempool_alloc0_type (pool, struct html_tag);
- cur_tag->params = g_queue_new ();
- rspamd_mempool_add_destructor (pool,
- (rspamd_mempool_destruct_t)g_queue_free, cur_tag->params);
+ content_parser_env.reset();
+
+ all_tags->emplace_back(std::make_unique<html_tag>());
+ cur_tag = all_tags->back().get();
break;
}
g_byte_array_append (dest, c, (p - c));
- len = rspamd_html_decode_entitles_inplace (
+ len = decode_html_entitles_inplace(
reinterpret_cast<gchar *>(dest->data + old_offset),
- p - c);
+ (std::size_t)(p - c));
dest->len = dest->len + len - (p - c);
if (content_tag) {
if (save_space) {
/* Append one space if needed */
if (dest->len > 0 &&
- !g_ascii_isspace (dest->data[dest->len - 1])) {
- g_byte_array_append (dest, reinterpret_cast<const guint8 *>(" "), 1);
+ !g_ascii_isspace (dest->data[dest->len - 1])) {
+ g_byte_array_append (dest,
+ reinterpret_cast<const guint8 *>(" "), 1);
if (content_tag) {
if (content_tag->content_length == 0) {
/*
}
g_byte_array_append (dest, c, (p - c));
- len = rspamd_html_decode_entitles_inplace (
+ len = decode_html_entitles_inplace (
reinterpret_cast<gchar *>(dest->data + old_offset),
p - c);
dest->len = dest->len + len - (p - c);
if (allow_css) {
GError *err = NULL;
- hc->css_style = rspamd_css_parse_style (pool, p, end_style, hc->css_style,
+ hc->css_style = rspamd_css_parse_style(pool, p, end_style, hc->css_style,
&err);
if (err) {
break;
case tag_content:
- rspamd_html_parse_tag_content (pool, hc, cur_tag,
- p, &substate, &savep);
+ parse_tag_content(pool, hc, cur_tag,
+ reinterpret_cast<const char *>(p), content_parser_env);
if (t == '>') {
if (closing) {
cur_tag->flags |= FL_CLOSING;
break;
case tag_end:
- substate = 0;
- savep = NULL;
+ content_parser_env.reset();
if (cur_tag != NULL) {
balanced = TRUE;
- if (rspamd_html_process_tag (pool, hc, cur_tag, &cur_level,
+ if (html_process_tag (pool, hc, cur_tag, &cur_level,
&balanced)) {
state = content_write;
need_decode = FALSE;
/* Handle newlines */
if (cur_tag->id == Tag_BR || cur_tag->id == Tag_HR) {
if (dest->len > 0 && dest->data[dest->len - 1] != '\n') {
- g_byte_array_append (dest, reinterpret_cast<const guint8 *>("\r\n"), 2);
+ g_byte_array_append (dest,
+ reinterpret_cast<const guint8 *>("\r\n"), 2);
if (content_tag) {
if (content_tag->content_length == 0) {
}
if ((cur_tag->id == Tag_P ||
- cur_tag->id == Tag_TR ||
- cur_tag->id == Tag_DIV)) {
+ cur_tag->id == Tag_TR ||
+ cur_tag->id == Tag_DIV)) {
if (dest->len > 0 && dest->data[dest->len - 1] != '\n') {
g_byte_array_append (dest, reinterpret_cast<const guint8 *>("\r\n"), 2);
/* XXX: uncomment when styles parsing is not so broken */
if (cur_tag->flags & FL_HREF /* && !(cur_tag->flags & FL_IGNORE) */) {
if (!(cur_tag->flags & (FL_CLOSING))) {
- url = rspamd_html_process_url_tag (pool, cur_tag, hc);
+ auto maybe_url = html_process_url_tag(pool, cur_tag, hc);
- if (url != NULL) {
+ if (maybe_url) {
+ url = maybe_url.value();
if (url_set != NULL) {
struct rspamd_url *maybe_existing =
- rspamd_url_set_add_or_return (url_set, url);
- if (maybe_existing == url) {
- rspamd_process_html_url (pool, url, url_set,
+ rspamd_url_set_add_or_return (url_set, maybe_url.value());
+ if (maybe_existing == maybe_url.value()) {
+ process_html_query_url(pool, url, url_set,
part_urls);
}
else {
prev_tag = static_cast<html_tag *>(cur_level->prev->data);
if (prev_tag->id == Tag_A &&
- !(prev_tag->flags & (FL_CLOSING)) &&
- prev_tag->extra) {
+ !(prev_tag->flags & (FL_CLOSING)) &&
+ prev_tag->extra) {
prev_url = static_cast<rspamd_url *>(prev_tag->extra);
- rspamd_html_check_displayed_url (pool,
+ std::string_view disp_part{
+ reinterpret_cast<const gchar *>(dest->data + href_offset),
+ dest->len - href_offset};
+ html_check_displayed_url (pool,
exceptions, url_set,
- dest, href_offset,
+ disp_part,
+ href_offset,
prev_url);
}
}
/* Insert exception */
if (url != NULL && (gint) dest->len > href_offset) {
- rspamd_html_check_displayed_url (pool,
+ std::string_view disp_part{
+ reinterpret_cast<const gchar *>(dest->data + href_offset),
+ dest->len - href_offset};
+ html_check_displayed_url (pool,
exceptions, url_set,
- dest, href_offset,
+ disp_part,
+ href_offset,
url);
}
* Base is allowed only within head tag but HTML is retarded
*/
if (hc->base_url == NULL) {
- url = rspamd_html_process_url_tag (pool, cur_tag, hc);
+ auto maybe_url = html_process_url_tag(pool, cur_tag, hc);
- if (url != NULL) {
+ if (maybe_url) {
msg_debug_html ("got valid base tag");
hc->base_url = url;
cur_tag->extra = url;
}
if (cur_tag->id == Tag_IMG && !(cur_tag->flags & FL_CLOSING)) {
- rspamd_html_process_img_tag (pool, cur_tag, hc, url_set,
+ html_process_img_tag(pool, cur_tag, hc, url_set,
part_urls, dest);
}
else if (cur_tag->id == Tag_LINK && !(cur_tag->flags & FL_CLOSING)) {
- rspamd_html_process_link_tag (pool, cur_tag, hc, url_set,
+ html_process_link_tag(pool, cur_tag, hc, url_set,
part_urls);
}
else if (cur_tag->flags & FL_BLOCK) {
}
}
else {
- rspamd_html_process_block_tag (pool, cur_tag, hc);
+ html_process_block_tag(pool, cur_tag, hc);
bl = static_cast<html_block *>(cur_tag->extra);
if (bl) {
- rspamd_html_propagate_style (hc, cur_tag,
+ html_propagate_style(hc, cur_tag,
bl, styles_blocks);
/* Check visibility */
if (hc->html_tags) {
g_node_traverse (hc->html_tags, G_POST_ORDER, G_TRAVERSE_ALL, -1,
- rspamd_html_propagate_lengths, NULL);
+ html_propagate_lengths, NULL);
}
g_queue_free (styles_blocks);
return dest;
}
+}
+
+GByteArray*
+rspamd_html_process_part_full (rspamd_mempool_t *pool,
+ struct html_content *hc,
+ GByteArray *in,
+ GList **exceptions,
+ khash_t (rspamd_url_hash) *url_set,
+ GPtrArray *part_urls,
+ bool allow_css)
+{
+ return rspamd::html::html_process_part_full(pool, hc, in, exceptions, url_set,
+ part_urls, allow_css);
+}
+
GByteArray*
rspamd_html_process_part (rspamd_mempool_t *pool,
struct html_content *hc,
projects / rspamd.git / commitdiff