 |
Expand extended_entities URLs in retweets - tscrape - twitter scraper |
 |
git clone git://git.codemadness.org/tscrape |
 |
Log |
|
Files |
|
Refs |
|
README |
|
LICENSE |
|
--- |
|
commit 9cc1b7e985affd8764385d3d6cf6476804230cdd |
|
parent d270b9dc10bc3d05b096f2dd34256dc9b962b951 |
|
Author: Leonardo Taccari <[email protected]> |
|
Date: Sat, 6 Jun 2020 01:50:22 +0200 |
|
|
|
Expand extended_entities URLs in retweets |
|
|
|
Diffstat: |
|
M tscrape.c | 23 +++++++++++++++++++++++ |
|
|
|
1 file changed, 23 insertions(+), 0 deletions(-) |
|
--- |
|
diff --git a/tscrape.c b/tscrape.c |
|
@@ -415,6 +415,29 @@ processnodes(struct json_node *nodes, size_t depth, const … |
|
url[0] = '\0'; |
|
} |
|
} |
|
+ |
|
+ /* [].retweeted_status.extended_entities.media[].url */ |
|
+ if (depth == 7 && |
|
+ nodes[0].type == JSON_TYPE_ARRAY && |
|
+ nodes[1].type == JSON_TYPE_OBJECT && |
|
+ nodes[2].type == JSON_TYPE_OBJECT && |
|
+ nodes[3].type == JSON_TYPE_OBJECT && |
|
+ nodes[4].type == JSON_TYPE_ARRAY && |
|
+ nodes[5].type == JSON_TYPE_OBJECT && |
|
+ nodes[6].type == JSON_TYPE_STRING && |
|
+ !strcmp(nodes[2].name, "retweeted_status") && |
|
+ !strcmp(nodes[3].name, "extended_entities") && |
|
+ !strcmp(nodes[4].name, "media")) { |
|
+ if (!strcmp(nodes[6].name, "url")) { |
|
+// printf("DEBUG: url: %s\n", str); |
|
+ strlcpy(url, str, sizeof(url)); |
|
+ } else if (!strcmp(nodes[6].name, "expanded_url")) { |
|
+// printf("DEBUG: expanded_url: %s\n", str); |
|
+ /* assumes "expanded_url" is specified after "url" */ |
|
+ addurl(url, str); |
|
+ url[0] = '\0'; |
|
+ } |
|
+ } |
|
} |
|
|
|
int |
