| 1 | /* |
| 2 | * RTP Depacketization of QCELP/PureVoice, RFC 2658 |
| 3 | * Copyright (c) 2010 Martin Storsjo |
| 4 | * |
| 5 | * This file is part of FFmpeg. |
| 6 | * |
| 7 | * FFmpeg is free software; you can redistribute it and/or |
| 8 | * modify it under the terms of the GNU Lesser General Public |
| 9 | * License as published by the Free Software Foundation; either |
| 10 | * version 2.1 of the License, or (at your option) any later version. |
| 11 | * |
| 12 | * FFmpeg is distributed in the hope that it will be useful, |
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 15 | * Lesser General Public License for more details. |
| 16 | * |
| 17 | * You should have received a copy of the GNU Lesser General Public |
| 18 | * License along with FFmpeg; if not, write to the Free Software |
| 19 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
| 20 | */ |
| 21 | |
| 22 | #include "rtpdec_formats.h" |
| 23 | |
| 24 | static const uint8_t frame_sizes[] = { |
| 25 | 1, 4, 8, 17, 35 |
| 26 | }; |
| 27 | |
| 28 | typedef struct { |
| 29 | int pos; |
| 30 | int size; |
| 31 | /* The largest frame is 35 bytes, only 10 frames are allowed per |
| 32 | * packet, and we return the first one immediately, so allocate |
| 33 | * space for 9 frames */ |
| 34 | uint8_t data[35*9]; |
| 35 | } InterleavePacket; |
| 36 | |
| 37 | struct PayloadContext { |
| 38 | int interleave_size; |
| 39 | int interleave_index; |
| 40 | InterleavePacket group[6]; |
| 41 | int group_finished; |
| 42 | |
| 43 | /* The maximum packet size, 10 frames of 35 bytes each, and one |
| 44 | * packet header byte. */ |
| 45 | uint8_t next_data[1 + 35*10]; |
| 46 | int next_size; |
| 47 | uint32_t next_timestamp; |
| 48 | }; |
| 49 | |
| 50 | static PayloadContext *qcelp_new_context(void) |
| 51 | { |
| 52 | return av_mallocz(sizeof(PayloadContext)); |
| 53 | } |
| 54 | |
| 55 | static void qcelp_free_context(PayloadContext *data) |
| 56 | { |
| 57 | av_free(data); |
| 58 | } |
| 59 | |
| 60 | static int return_stored_frame(AVFormatContext *ctx, PayloadContext *data, |
| 61 | AVStream *st, AVPacket *pkt, uint32_t *timestamp, |
| 62 | const uint8_t *buf, int len); |
| 63 | |
| 64 | static int store_packet(AVFormatContext *ctx, PayloadContext *data, |
| 65 | AVStream *st, AVPacket *pkt, uint32_t *timestamp, |
| 66 | const uint8_t *buf, int len) |
| 67 | { |
| 68 | int interleave_size, interleave_index; |
| 69 | int frame_size, ret; |
| 70 | InterleavePacket* ip; |
| 71 | |
| 72 | if (len < 2) |
| 73 | return AVERROR_INVALIDDATA; |
| 74 | |
| 75 | interleave_size = buf[0] >> 3 & 7; |
| 76 | interleave_index = buf[0] & 7; |
| 77 | |
| 78 | if (interleave_size > 5) { |
| 79 | av_log(ctx, AV_LOG_ERROR, "Invalid interleave size %d\n", |
| 80 | interleave_size); |
| 81 | return AVERROR_INVALIDDATA; |
| 82 | } |
| 83 | if (interleave_index > interleave_size) { |
| 84 | av_log(ctx, AV_LOG_ERROR, "Invalid interleave index %d/%d\n", |
| 85 | interleave_index, interleave_size); |
| 86 | return AVERROR_INVALIDDATA; |
| 87 | } |
| 88 | if (interleave_size != data->interleave_size) { |
| 89 | int i; |
| 90 | /* First packet, or changed interleave size */ |
| 91 | data->interleave_size = interleave_size; |
| 92 | data->interleave_index = 0; |
| 93 | for (i = 0; i < 6; i++) |
| 94 | data->group[i].size = 0; |
| 95 | } |
| 96 | |
| 97 | if (interleave_index < data->interleave_index) { |
| 98 | /* Wrapped around - missed the last packet of the previous group. */ |
| 99 | if (data->group_finished) { |
| 100 | /* No more data in the packets in this interleaving group, just |
| 101 | * start processing the next one */ |
| 102 | data->interleave_index = 0; |
| 103 | } else { |
| 104 | /* Stash away the current packet, emit everything we have of the |
| 105 | * previous group. */ |
| 106 | for (; data->interleave_index <= interleave_size; |
| 107 | data->interleave_index++) |
| 108 | data->group[data->interleave_index].size = 0; |
| 109 | |
| 110 | if (len > sizeof(data->next_data)) |
| 111 | return AVERROR_INVALIDDATA; |
| 112 | memcpy(data->next_data, buf, len); |
| 113 | data->next_size = len; |
| 114 | data->next_timestamp = *timestamp; |
| 115 | *timestamp = RTP_NOTS_VALUE; |
| 116 | |
| 117 | data->interleave_index = 0; |
| 118 | return return_stored_frame(ctx, data, st, pkt, timestamp, buf, len); |
| 119 | } |
| 120 | } |
| 121 | if (interleave_index > data->interleave_index) { |
| 122 | /* We missed a packet */ |
| 123 | for (; data->interleave_index < interleave_index; |
| 124 | data->interleave_index++) |
| 125 | data->group[data->interleave_index].size = 0; |
| 126 | } |
| 127 | data->interleave_index = interleave_index; |
| 128 | |
| 129 | if (buf[1] >= FF_ARRAY_ELEMS(frame_sizes)) |
| 130 | return AVERROR_INVALIDDATA; |
| 131 | frame_size = frame_sizes[buf[1]]; |
| 132 | if (1 + frame_size > len) |
| 133 | return AVERROR_INVALIDDATA; |
| 134 | |
| 135 | if (len - 1 - frame_size > sizeof(data->group[0].data)) |
| 136 | return AVERROR_INVALIDDATA; |
| 137 | |
| 138 | if ((ret = av_new_packet(pkt, frame_size)) < 0) |
| 139 | return ret; |
| 140 | memcpy(pkt->data, &buf[1], frame_size); |
| 141 | pkt->stream_index = st->index; |
| 142 | |
| 143 | ip = &data->group[data->interleave_index]; |
| 144 | ip->size = len - 1 - frame_size; |
| 145 | ip->pos = 0; |
| 146 | memcpy(ip->data, &buf[1 + frame_size], ip->size); |
| 147 | /* Each packet must contain the same number of frames according to the |
| 148 | * RFC. If there's no data left in this packet, there shouldn't be any |
| 149 | * in any of the other frames in the interleaving group either. */ |
| 150 | data->group_finished = ip->size == 0; |
| 151 | |
| 152 | if (interleave_index == interleave_size) { |
| 153 | data->interleave_index = 0; |
| 154 | return !data->group_finished; |
| 155 | } else { |
| 156 | data->interleave_index++; |
| 157 | return 0; |
| 158 | } |
| 159 | } |
| 160 | |
| 161 | static int return_stored_frame(AVFormatContext *ctx, PayloadContext *data, |
| 162 | AVStream *st, AVPacket *pkt, uint32_t *timestamp, |
| 163 | const uint8_t *buf, int len) |
| 164 | { |
| 165 | InterleavePacket* ip = &data->group[data->interleave_index]; |
| 166 | int frame_size, ret; |
| 167 | |
| 168 | if (data->group_finished && data->interleave_index == 0) { |
| 169 | *timestamp = data->next_timestamp; |
| 170 | ret = store_packet(ctx, data, st, pkt, timestamp, data->next_data, |
| 171 | data->next_size); |
| 172 | data->next_size = 0; |
| 173 | return ret; |
| 174 | } |
| 175 | |
| 176 | if (ip->size == 0) { |
| 177 | /* No stored data for this interleave block, output an empty packet */ |
| 178 | if ((ret = av_new_packet(pkt, 1)) < 0) |
| 179 | return ret; |
| 180 | pkt->data[0] = 0; // Blank - could also be 14, Erasure |
| 181 | } else { |
| 182 | if (ip->pos >= ip->size) |
| 183 | return AVERROR_INVALIDDATA; |
| 184 | if (ip->data[ip->pos] >= FF_ARRAY_ELEMS(frame_sizes)) |
| 185 | return AVERROR_INVALIDDATA; |
| 186 | frame_size = frame_sizes[ip->data[ip->pos]]; |
| 187 | if (ip->pos + frame_size > ip->size) |
| 188 | return AVERROR_INVALIDDATA; |
| 189 | |
| 190 | if ((ret = av_new_packet(pkt, frame_size)) < 0) |
| 191 | return ret; |
| 192 | memcpy(pkt->data, &ip->data[ip->pos], frame_size); |
| 193 | |
| 194 | ip->pos += frame_size; |
| 195 | data->group_finished = ip->pos >= ip->size; |
| 196 | } |
| 197 | pkt->stream_index = st->index; |
| 198 | |
| 199 | if (data->interleave_index == data->interleave_size) { |
| 200 | data->interleave_index = 0; |
| 201 | if (!data->group_finished) |
| 202 | return 1; |
| 203 | else |
| 204 | return data->next_size > 0; |
| 205 | } else { |
| 206 | data->interleave_index++; |
| 207 | return 1; |
| 208 | } |
| 209 | } |
| 210 | |
| 211 | static int qcelp_parse_packet(AVFormatContext *ctx, PayloadContext *data, |
| 212 | AVStream *st, AVPacket *pkt, uint32_t *timestamp, |
| 213 | const uint8_t *buf, int len, uint16_t seq, |
| 214 | int flags) |
| 215 | { |
| 216 | if (buf) |
| 217 | return store_packet(ctx, data, st, pkt, timestamp, buf, len); |
| 218 | else |
| 219 | return return_stored_frame(ctx, data, st, pkt, timestamp, buf, len); |
| 220 | } |
| 221 | |
| 222 | RTPDynamicProtocolHandler ff_qcelp_dynamic_handler = { |
| 223 | .enc_name = "x-Purevoice", |
| 224 | .codec_type = AVMEDIA_TYPE_AUDIO, |
| 225 | .codec_id = AV_CODEC_ID_QCELP, |
| 226 | .static_payload_id = 12, |
| 227 | .alloc = qcelp_new_context, |
| 228 | .free = qcelp_free_context, |
| 229 | .parse_packet = qcelp_parse_packet |
| 230 | }; |