Relintai/wp_saver_rcpp_fw
1
0
Fork 0
mirror of https://github.com/Relintai/wp_saver_rcpp_fw.git synced 2024-11-14 04:57:23 +01:00
Code Issues Packages Projects Releases Wiki Activity

Extract next ling using the rel="next" attribute.

Browse Source
This commit is contained in:
Relintai 2021-11-21 10:44:08 +01:00
parent 44a5c167ad
commit 3a0640801b
1 changed files with 8 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
main.cpp
View File

@ -167,12 +167,10 @@ void download_posts(Database *db, const String &site) {
if (last_url == "") {
should_skip = false;
last_url = first_url;
}
RLOG_MSG("Post downloading started for " + site + " | last url: " + last_url);
bool done = false;
@ -197,22 +195,17 @@ void download_posts(Database *db, const String &site) {
RLOG_WARN("Couldn't extract data!\n");
}
HTMLParserTag *next_div = p.root->get_first("div", "class", "nav-next");
HTMLParserTag *n_link_tag = p.root->get_first("a", "rel", "next");
if (next_div) {
if (next_div->tags.size() == 1) {
HTMLParserTag *link = next_div->tags[0];
if (n_link_tag) {
next_link = n_link_tag->get_attribute_value("href");
next_link = link->get_attribute_value("href");
if (next_link == "") {
RLOG_WARN("Couldn't extract link!\n");
}
} else {
RLOG_WARN("Couldn't extract next_div! (tags.size() != 1)!\n");
if (next_link == "") {
RLOG_WARN("Couldn't extract link!\n");
}
} else {
RLOG_WARN("Couldn't extract next_div!\n");
next_link = "";
RLOG_WARN("Couldn't extract link tag!\n");
}
if (should_skip) {