Le 9 octobre 2024 13:16:37 GMT+03:00, Radoslav Gerganov <rgerga...@gmail.com> a
écrit :
>Fix regression introduced with commit 69fcc09 which doesn't honor
>reconnect_at_eof and breaks support for live/endless streams. For such
>streams we should reconnect at EOF and clients set the reconnect_at_eof
>option to signal this.
>
>Signed-off-by: Radoslav Gerganov <rgerga...@gmail.com>
>---
> libavformat/http.c | 13 +++++++------
> 1 file changed, 7 insertions(+), 6 deletions(-)
>
>diff --git a/libavformat/http.c b/libavformat/http.c
>index ec60bc0b17..27f8b98744 100644
>--- a/libavformat/http.c
>+++ b/libavformat/http.c
>@@ -1988,12 +1988,13 @@ static int64_t http_seek_internal(URLContext *h,
>int64_t off, int whence, int fo
> if (s->off && h->is_streamed)
> return AVERROR(ENOSYS);
>
>- /* do not try to make a new connection if seeking past the end of the
>file */
>- if (s->end_off || s->filesize != UINT64_MAX) {
>- uint64_t end_pos = s->end_off ? s->end_off : s->filesize;
>- if (s->off >= end_pos)
>- return s->off;
>- }
>+ if (!s->reconnect_at_eof)
>+ /* do not try to make a new connection if seeking past the end of the
>file */
>+ if (s->end_off || s->filesize != UINT64_MAX) {
>+ uint64_t end_pos = s->end_off ? s->end_off : s->filesize;
>+ if (s->off >= end_pos)
>+ return s->off;
>+ }
AFAICT, reconnecting blindly makes no sense if the file size is known, flag or
no flag.
With that said the current code seems odd. Normal seek semantics allow for
seeking to or even past the end of a file. The offset should not be capped like
this. The server will return 416 which should be treated as "successfully"
seeking. Or if the file has changed size in the mean time, it will return 206
and we can continue streaming.
But either way, this patch doesn't look right.
>
> /* if the location changed (redirect), revert to the original uri */
> if (strcmp(s->uri, s->location)) {
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
