[FFmpeg-cvslog] screenpresso: Add extended pixel format support

Vittorio Giovara git at videolan.org
Mon May 9 00:47:26 CEST 2016


ffmpeg | branch: master | Vittorio Giovara <vittorio.giovara at gmail.com> | Tue Mar 29 20:30:47 2016 +0200| [95db8c757cb003a71b040b567f38be74151deb5c] | committer: Vittorio Giovara

screenpresso: Add extended pixel format support

Signed-off-by: Vittorio Giovara <vittorio.giovara at gmail.com>

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=95db8c757cb003a71b040b567f38be74151deb5c
---

 libavcodec/screenpresso.c |   46 ++++++++++++++++++++++++++++++---------------
 1 file changed, 31 insertions(+), 15 deletions(-)

diff --git a/libavcodec/screenpresso.c b/libavcodec/screenpresso.c
index 88a927b..41d1755 100644
--- a/libavcodec/screenpresso.c
+++ b/libavcodec/screenpresso.c
@@ -30,7 +30,7 @@
  * rebuilt frame (not the reference), and since there is no coordinate system
  * they contain exactly as many pixel as the keyframe.
  *
- * Supports: BGR24
+ * Supports: BGRA, BGR24, RGB555
  */
 
 #include <stdint.h>
@@ -79,10 +79,8 @@ static av_cold int screenpresso_init(AVCodecContext *avctx)
     if (!ctx->current)
         return AVERROR(ENOMEM);
 
-    avctx->pix_fmt = AV_PIX_FMT_BGR24;
-
-    /* Allocate maximum size possible, a full frame */
-    ctx->inflated_size = avctx->width * avctx->height * 3;
+    /* Allocate maximum size possible, a full RGBA frame */
+    ctx->inflated_size = avctx->width * avctx->height * 4;
     ctx->inflated_buf  = av_malloc(ctx->inflated_size);
     if (!ctx->inflated_buf)
         return AVERROR(ENOMEM);
@@ -108,7 +106,7 @@ static int screenpresso_decode_frame(AVCodecContext *avctx, void *data,
     ScreenpressoContext *ctx = avctx->priv_data;
     AVFrame *frame = data;
     uLongf length = ctx->inflated_size;
-    int keyframe;
+    int keyframe, component_size, src_linesize;
     int ret;
 
     /* Size check */
@@ -118,13 +116,28 @@ static int screenpresso_decode_frame(AVCodecContext *avctx, void *data,
     }
 
     /* Basic sanity check, but not really harmful */
-    if ((avpkt->data[0] != 0x73 && avpkt->data[0] != 0x72) ||
-        avpkt->data[1] != 8) { // bpp probably
-        av_log(avctx, AV_LOG_WARNING, "Unknown header 0x%02X%02X\n",
-               avpkt->data[0], avpkt->data[1]);
-    }
+    if (avpkt->data[0] != 0x73 && avpkt->data[0] != 0x72)
+        av_log(avctx, AV_LOG_WARNING, "Unknown header 0x%02X\n", avpkt->data[0]);
     keyframe = (avpkt->data[0] == 0x73);
 
+    /* Pixel size */
+    component_size = ((avpkt->data[1] >> 2) & 0x03) + 1;
+    switch (component_size) {
+    case 2:
+        avctx->pix_fmt = AV_PIX_FMT_RGB555LE;
+        break;
+    case 3:
+        avctx->pix_fmt = AV_PIX_FMT_BGR24;
+        break;
+    case 4:
+        avctx->pix_fmt = AV_PIX_FMT_BGRA;
+        break;
+    default:
+        av_log(avctx, AV_LOG_ERROR, "Invalid bits per pixel value (%d)\n",
+               component_size);
+        return AVERROR_INVALIDDATA;
+    }
+
     /* Inflate the frame after the 2 byte header */
     ret = uncompress(ctx->inflated_buf, &length,
                      avpkt->data + 2, avpkt->size - 2);
@@ -137,18 +150,21 @@ static int screenpresso_decode_frame(AVCodecContext *avctx, void *data,
     if (ret < 0)
         return ret;
 
+    /* Codec has aligned strides */
+    src_linesize = FFALIGN(avctx->width * component_size, 4);
+
     /* When a keyframe is found, copy it (flipped) */
     if (keyframe)
         av_image_copy_plane(ctx->current->data[0] +
                             ctx->current->linesize[0] * (avctx->height - 1),
                             -1 * ctx->current->linesize[0],
-                            ctx->inflated_buf, avctx->width * 3,
-                            avctx->width * 3, avctx->height);
+                            ctx->inflated_buf, src_linesize,
+                            avctx->width * component_size, avctx->height);
     /* Otherwise sum the delta on top of the current frame */
     else
         sum_delta_flipped(ctx->current->data[0], ctx->current->linesize[0],
-                          ctx->inflated_buf, avctx->width * 3,
-                          avctx->width * 3, avctx->height);
+                          ctx->inflated_buf, src_linesize,
+                          avctx->width * component_size, avctx->height);
 
     /* Frame is ready to be output */
     ret = av_frame_ref(frame, ctx->current);



More information about the ffmpeg-cvslog mailing list