On Fri, 18 Jul 2014, Luca Barbato wrote:

Certain servers accept only PUT as valid method.
---

Now w/out the stray enum.

libavformat/http.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/libavformat/http.c b/libavformat/http.c
index 948930a..cb800ce 100644
--- a/libavformat/http.c
+++ b/libavformat/http.c
@@ -85,6 +85,7 @@ typedef struct {
#endif
    AVDictionary *chained_options;
    int send_expect_100;
+    char *write_method;
} HTTPContext;

#define OFFSET(x) offsetof(HTTPContext, x)
@@ -110,6 +111,7 @@ static const AVOption options[] = {
{"location", "The actual location of the data received", OFFSET(location), 
AV_OPT_TYPE_STRING, { 0 }, 0, 0, D|E },
{"offset", "initial byte offset", OFFSET(off), AV_OPT_TYPE_INT64, {.i64 = 0}, 
0, INT64_MAX, D },
{"end_offset", "try to limit the request to bytes preceding this offset", 
OFFSET(end_off), AV_OPT_TYPE_INT64, {.i64 = 0}, 0, INT64_MAX, D },
+{"write_method", "Method to write to the remote server", OFFSET(write_method), 
AV_OPT_TYPE_STRING, { .str = "POST" }, 0, 0, E, },
{NULL}
};
#define HTTP_CLASS(flavor)\
@@ -555,7 +557,12 @@ static int http_connect(URLContext *h, const char *path, 
const char *local_path,
        s->chunked_post = 0;
    }

-    method = post ? "POST" : "GET";
+    if (post) {
+        method = s->write_method;
+    } else {
+        method = "GET";
+    }
+

I still think it would make even more sense to allow this even for the reading mode as well. I don't have any specific use case in mind, but it would feel stupid to have to add a separate read_method if we later come up with such a use case.

The risk of someone shooting themselves in the foot with it doesn't feel that big either - and in case they do, it should be pretty clear what happened.

In that case you can't set a default for this in the avoptions table, but check whether the method string is set and not empty, otherwise use the default POST/GET.

// Martin
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to