ffplay: Use sws_scale to scale subtitles

Fixes some files from Ticket679

This also changes subtitles to 4:2:0 matching the output format and thus
simplifying the blend code.
This restricts placement to the chroma sample resolution though, speak up
if you consider this a problem, say so, the code could be changed to use
YUV444 for subtitles and scaling them down while blending, this would be
slower though.
The current code only uses a single swscale context and reinitializes it
as needed, this could be changed as well if needed

Reviewed-by: Marton Balint <cus@passwd.hu>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

1 files changed, 71 insertions, 213 deletions

diff --git a/ffplay.c b/ffplay.c
index 15bc6f1e2d..cc61dde0b2 100644
--- a/ffplay.c
+++ b/ffplay.c
@@ -223,6 +223,9 @@ typedef struct VideoState {
     Decoder viddec;
     Decoder subdec;
 
+    int viddec_width;
+    int viddec_height;
+
     int audio_stream;
 
     int av_sync_type;
@@ -278,6 +281,7 @@ typedef struct VideoState {
 #if !CONFIG_AVFILTER
     struct SwsContext *img_convert_ctx;
 #endif
+    struct SwsContext *sub_convert_ctx;
     SDL_Rect last_display_rect;
     int eof;
 
@@ -829,229 +833,50 @@ static void fill_border(int xleft, int ytop, int width, int height, int x, int y
 #define ALPHA_BLEND(a, oldp, newp, s)\
 ((((oldp << s) * (255 - (a))) + (newp * (a))) / (255 << s))
 
-#define RGBA_IN(r, g, b, a, s)\
-{\
-    unsigned int v = ((const uint32_t *)(s))[0];\
-    a = (v >> 24) & 0xff;\
-    r = (v >> 16) & 0xff;\
-    g = (v >> 8) & 0xff;\
-    b = v & 0xff;\
-}
-
-#define YUVA_IN(y, u, v, a, s, pal)\
-{\
-    unsigned int val = ((const uint32_t *)(pal))[*(const uint8_t*)(s)];\
-    a = (val >> 24) & 0xff;\
-    y = (val >> 16) & 0xff;\
-    u = (val >> 8) & 0xff;\
-    v = val & 0xff;\
-}
-
-#define YUVA_OUT(d, y, u, v, a)\
-{\
-    ((uint32_t *)(d))[0] = (a << 24) | (y << 16) | (u << 8) | v;\
-}
 
 
 #define BPP 1
 
 static void blend_subrect(AVPicture *dst, const AVSubtitleRect *rect, int imgw, int imgh)
 {
-    int wrap, wrap3, width2, skip2;
-    int y, u, v, a, u1, v1, a1, w, h;
+    int x, y, Y, U, V, A;
     uint8_t *lum, *cb, *cr;
-    const uint8_t *p;
-    const uint32_t *pal;
     int dstx, dsty, dstw, dsth;
+    const AVPicture *src = &rect->pict;
 
     dstw = av_clip(rect->w, 0, imgw);
     dsth = av_clip(rect->h, 0, imgh);
     dstx = av_clip(rect->x, 0, imgw - dstw);
     dsty = av_clip(rect->y, 0, imgh - dsth);
-    lum = dst->data[0] + dsty * dst->linesize[0];
-    cb  = dst->data[1] + (dsty >> 1) * dst->linesize[1];
-    cr  = dst->data[2] + (dsty >> 1) * dst->linesize[2];
-
-    width2 = ((dstw + 1) >> 1) + (dstx & ~dstw & 1);
-    skip2 = dstx >> 1;
-    wrap = dst->linesize[0];
-    wrap3 = rect->pict.linesize[0];
-    p = rect->pict.data[0];
-    pal = (const uint32_t *)rect->pict.data[1];  /* Now in YCrCb! */
-
-    if (dsty & 1) {
-        lum += dstx;
-        cb += skip2;
-        cr += skip2;
-
-        if (dstx & 1) {
-            YUVA_IN(y, u, v, a, p, pal);
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-            cb[0] = ALPHA_BLEND(a >> 2, cb[0], u, 0);
-            cr[0] = ALPHA_BLEND(a >> 2, cr[0], v, 0);
-            cb++;
-            cr++;
+    lum = dst->data[0] + dstx + dsty * dst->linesize[0];
+    cb  = dst->data[1] + dstx/2 + (dsty >> 1) * dst->linesize[1];
+    cr  = dst->data[2] + dstx/2 + (dsty >> 1) * dst->linesize[2];
+
+    for (y = 0; y<dsth; y++) {
+        for (x = 0; x<dstw; x++) {
+            Y = src->data[0][x + y*src->linesize[0]];
+            A = src->data[3][x + y*src->linesize[3]];
+            lum[0] = ALPHA_BLEND(A, lum[0], Y, 0);
             lum++;
-            p += BPP;
-        }
-        for (w = dstw - (dstx & 1); w >= 2; w -= 2) {
-            YUVA_IN(y, u, v, a, p, pal);
-            u1 = u;
-            v1 = v;
-            a1 = a;
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-
-            YUVA_IN(y, u, v, a, p + BPP, pal);
-            u1 += u;
-            v1 += v;
-            a1 += a;
-            lum[1] = ALPHA_BLEND(a, lum[1], y, 0);
-            cb[0] = ALPHA_BLEND(a1 >> 2, cb[0], u1, 1);
-            cr[0] = ALPHA_BLEND(a1 >> 2, cr[0], v1, 1);
-            cb++;
-            cr++;
-            p += 2 * BPP;
-            lum += 2;
-        }
-        if (w) {
-            YUVA_IN(y, u, v, a, p, pal);
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-            cb[0] = ALPHA_BLEND(a >> 2, cb[0], u, 0);
-            cr[0] = ALPHA_BLEND(a >> 2, cr[0], v, 0);
-            p++;
-            lum++;
-        }
-        p += wrap3 - dstw * BPP;
-        lum += wrap - dstw - dstx;
-        cb += dst->linesize[1] - width2 - skip2;
-        cr += dst->linesize[2] - width2 - skip2;
-    }
-    for (h = dsth - (dsty & 1); h >= 2; h -= 2) {
-        lum += dstx;
-        cb += skip2;
-        cr += skip2;
-
-        if (dstx & 1) {
-            YUVA_IN(y, u, v, a, p, pal);
-            u1 = u;
-            v1 = v;
-            a1 = a;
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-            p += wrap3;
-            lum += wrap;
-            YUVA_IN(y, u, v, a, p, pal);
-            u1 += u;
-            v1 += v;
-            a1 += a;
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-            cb[0] = ALPHA_BLEND(a1 >> 2, cb[0], u1, 1);
-            cr[0] = ALPHA_BLEND(a1 >> 2, cr[0], v1, 1);
-            cb++;
-            cr++;
-            p += -wrap3 + BPP;
-            lum += -wrap + 1;
         }
-        for (w = dstw - (dstx & 1); w >= 2; w -= 2) {
-            YUVA_IN(y, u, v, a, p, pal);
-            u1 = u;
-            v1 = v;
-            a1 = a;
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-
-            YUVA_IN(y, u, v, a, p + BPP, pal);
-            u1 += u;
-            v1 += v;
-            a1 += a;
-            lum[1] = ALPHA_BLEND(a, lum[1], y, 0);
-            p += wrap3;
-            lum += wrap;
-
-            YUVA_IN(y, u, v, a, p, pal);
-            u1 += u;
-            v1 += v;
-            a1 += a;
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-
-            YUVA_IN(y, u, v, a, p + BPP, pal);
-            u1 += u;
-            v1 += v;
-            a1 += a;
-            lum[1] = ALPHA_BLEND(a, lum[1], y, 0);
-
-            cb[0] = ALPHA_BLEND(a1 >> 2, cb[0], u1, 2);
-            cr[0] = ALPHA_BLEND(a1 >> 2, cr[0], v1, 2);
+        lum += dst->linesize[0] - dstw;
+    }
 
+    for (y = 0; y<dsth/2; y++) {
+        for (x = 0; x<dstw/2; x++) {
+            U = src->data[1][x + y*src->linesize[1]];
+            V = src->data[2][x + y*src->linesize[2]];
+            A = src->data[3][2*x     +  2*y   *src->linesize[3]]
+              + src->data[3][2*x + 1 +  2*y   *src->linesize[3]]
+              + src->data[3][2*x + 1 + (2*y+1)*src->linesize[3]]
+              + src->data[3][2*x     + (2*y+1)*src->linesize[3]];
+            cb[0] = ALPHA_BLEND(A>>2, cb[0], U, 0);
+            cr[0] = ALPHA_BLEND(A>>2, cr[0], V, 0);
             cb++;
             cr++;
-            p += -wrap3 + 2 * BPP;
-            lum += -wrap + 2;
-        }
-        if (w) {
-            YUVA_IN(y, u, v, a, p, pal);
-            u1 = u;
-            v1 = v;
-            a1 = a;
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-            p += wrap3;
-            lum += wrap;
-            YUVA_IN(y, u, v, a, p, pal);
-            u1 += u;
-            v1 += v;
-            a1 += a;
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-            cb[0] = ALPHA_BLEND(a1 >> 2, cb[0], u1, 1);
-            cr[0] = ALPHA_BLEND(a1 >> 2, cr[0], v1, 1);
-            cb++;
-            cr++;
-            p += -wrap3 + BPP;
-            lum += -wrap + 1;
-        }
-        p += wrap3 + (wrap3 - dstw * BPP);
-        lum += wrap + (wrap - dstw - dstx);
-        cb += dst->linesize[1] - width2 - skip2;
-        cr += dst->linesize[2] - width2 - skip2;
-    }
-    /* handle odd height */
-    if (h) {
-        lum += dstx;
-        cb += skip2;
-        cr += skip2;
-
-        if (dstx & 1) {
-            YUVA_IN(y, u, v, a, p, pal);
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-            cb[0] = ALPHA_BLEND(a >> 2, cb[0], u, 0);
-            cr[0] = ALPHA_BLEND(a >> 2, cr[0], v, 0);
-            cb++;
-            cr++;
-            lum++;
-            p += BPP;
-        }
-        for (w = dstw - (dstx & 1); w >= 2; w -= 2) {
-            YUVA_IN(y, u, v, a, p, pal);
-            u1 = u;
-            v1 = v;
-            a1 = a;
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-
-            YUVA_IN(y, u, v, a, p + BPP, pal);
-            u1 += u;
-            v1 += v;
-            a1 += a;
-            lum[1] = ALPHA_BLEND(a, lum[1], y, 0);
-            cb[0] = ALPHA_BLEND(a1 >> 2, cb[0], u, 1);
-            cr[0] = ALPHA_BLEND(a1 >> 2, cr[0], v, 1);
-            cb++;
-            cr++;
-            p += 2 * BPP;
-            lum += 2;
-        }
-        if (w) {
-            YUVA_IN(y, u, v, a, p, pal);
-            lum[0] = ALPHA_BLEND(a, lum[0], y, 0);
-            cb[0] = ALPHA_BLEND(a >> 2, cb[0], u, 0);
-            cr[0] = ALPHA_BLEND(a >> 2, cr[0], v, 0);
         }
+        cb += dst->linesize[1] - dstw/2;
+        cr += dst->linesize[2] - dstw/2;
     }
 }
 
@@ -1306,6 +1131,7 @@ static void stream_close(VideoState *is)
 #if !CONFIG_AVFILTER
     sws_freeContext(is->img_convert_ctx);
 #endif
+    sws_freeContext(is->sub_convert_ctx);
     av_free(is);
 }
 
@@ -1893,6 +1719,9 @@ static int get_video_frame(VideoState *is, AVFrame *frame)
 
         frame->sample_aspect_ratio = av_guess_sample_aspect_ratio(is->ic, is->video_st, frame);
 
+        is->viddec_width  = frame->width;
+        is->viddec_height = frame->height;
+
         if (framedrop>0 || (framedrop && get_master_sync_type(is) != AV_SYNC_VIDEO_MASTER)) {
             if (frame->pts != AV_NOPTS_VALUE) {
                 double diff = dpts - get_master_clock(is);
@@ -2328,8 +2157,7 @@ static int subtitle_thread(void *arg)
     Frame *sp;
     int got_subtitle;
     double pts;
-    int i, j;
-    int r, g, b, y, u, v, a;
+    int i;
 
     for (;;) {
         if (!(sp = frame_queue_peek_writable(&is->subpq)))
@@ -2348,14 +2176,41 @@ static int subtitle_thread(void *arg)
 
             for (i = 0; i < sp->sub.num_rects; i++)
             {
-                for (j = 0; j < sp->sub.rects[i]->nb_colors; j++)
-                {
-                    RGBA_IN(r, g, b, a, (uint32_t*)sp->sub.rects[i]->pict.data[1] + j);
-                    y = RGB_TO_Y_CCIR(r, g, b);
-                    u = RGB_TO_U_CCIR(r, g, b, 0);
-                    v = RGB_TO_V_CCIR(r, g, b, 0);
-                    YUVA_OUT((uint32_t*)sp->sub.rects[i]->pict.data[1] + j, y, u, v, a);
+                int in_w = sp->sub.rects[i]->w;
+                int in_h = sp->sub.rects[i]->h;
+                int subw = is->subdec.avctx->width  ? is->subdec.avctx->width  : is->viddec_width;
+                int subh = is->subdec.avctx->height ? is->subdec.avctx->height : is->viddec_height;
+                int out_w = is->viddec_width  ? in_w * is->viddec_width  / subw : in_w;
+                int out_h = is->viddec_height ? in_h * is->viddec_height / subh : in_h;
+                AVPicture newpic;
+
+                //can not use avpicture_alloc as it is not compatible with avsubtitle_free()
+                av_image_fill_linesizes(newpic.linesize, AV_PIX_FMT_YUVA420P, out_w);
+                newpic.data[0] = av_malloc(newpic.linesize[0] * out_h);
+                newpic.data[3] = av_malloc(newpic.linesize[3] * out_h);
+                newpic.data[1] = av_malloc(newpic.linesize[1] * ((out_h+1)/2));
+                newpic.data[2] = av_malloc(newpic.linesize[2] * ((out_h+1)/2));
+
+                is->sub_convert_ctx = sws_getCachedContext(is->sub_convert_ctx,
+                    in_w, in_h, AV_PIX_FMT_PAL8, out_w, out_h,
+                    AV_PIX_FMT_YUVA420P, sws_flags, NULL, NULL, NULL);
+                if (!is->sub_convert_ctx || !newpic.data[0] || !newpic.data[3] ||
+                    !newpic.data[1] || !newpic.data[2]
+                ) {
+                    av_log(NULL, AV_LOG_FATAL, "Cannot initialize the sub conversion context\n");
+                    exit(1);
                 }
+                sws_scale(is->sub_convert_ctx,
+                          (void*)sp->sub.rects[i]->pict.data, sp->sub.rects[i]->pict.linesize,
+                          0, in_h, newpic.data, newpic.linesize);
+
+                av_free(sp->sub.rects[i]->pict.data[0]);
+                av_free(sp->sub.rects[i]->pict.data[1]);
+                sp->sub.rects[i]->pict = newpic;
+                sp->sub.rects[i]->w = out_w;
+                sp->sub.rects[i]->h = out_h;
+                sp->sub.rects[i]->x = sp->sub.rects[i]->x * out_w / in_w;
+                sp->sub.rects[i]->y = sp->sub.rects[i]->y * out_h / in_h;
             }
 
             /* now we can update the picture count */
@@ -2772,6 +2627,9 @@ static int stream_component_open(VideoState *is, int stream_index)
         is->video_stream = stream_index;
         is->video_st = ic->streams[stream_index];
 
+        is->viddec_width  = avctx->width;
+        is->viddec_height = avctx->height;
+
         decoder_init(&is->viddec, avctx, &is->videoq, is->continue_read_thread);
         decoder_start(&is->viddec, video_thread, is);
         is->queue_attachments_req = 1;