Message ID | 20220406120056.10537-2-markg85@gmail.com |
---|---|
State | New |
Headers | show |
Series | Add IPFS protocol support. | expand |
Context | Check | Description |
---|---|---|
andriy/make_x86 | success | Make finished |
andriy/make_fate_x86 | success | Make fate finished |
andriy/make_aarch64_jetson | success | Make finished |
andriy/make_fate_aarch64_jetson | success | Make fate finished |
On Wed, Apr 6, 2022 at 2:02 PM Mark Gaiser <markg85@gmail.com> wrote: > This patch adds support for: > - ffplay ipfs://<cid> > - ffplay ipns://<cid> > > IPFS data can be played from so called "ipfs gateways". > A gateway is essentially a webserver that gives access to the > distributed IPFS network. > > This protocol support (ipfs and ipns) therefore translates > ipfs:// and ipns:// to a http:// url. This resulting url is > then handled by the http protocol. It could also be https > depending on the gateway provided. > > To use this protocol, a gateway must be provided. > If you do nothing it will try to find it in your > $HOME/.ipfs/gateway file. The ways to set it manually are: > 1. Define a -gateway <url> to the gateway. > 2. Define $IPFS_GATEWAY with the full http link to the gateway. > 3. Define $IPFS_PATH and point it to the IPFS data path. > 4. Have IPFS running in your local user folder (under $HOME/.ipfs). > > Signed-off-by: Mark Gaiser <markg85@gmail.com> > --- > configure | 2 + > doc/protocols.texi | 30 ++++ > libavformat/Makefile | 2 + > libavformat/ipfsgateway.c | 339 ++++++++++++++++++++++++++++++++++++++ > libavformat/protocols.c | 2 + > 5 files changed, 375 insertions(+) > create mode 100644 libavformat/ipfsgateway.c > > diff --git a/configure b/configure > index e4d36aa639..55af90957a 100755 > --- a/configure > +++ b/configure > @@ -3579,6 +3579,8 @@ udp_protocol_select="network" > udplite_protocol_select="network" > unix_protocol_deps="sys_un_h" > unix_protocol_select="network" > +ipfs_protocol_select="https_protocol" > +ipns_protocol_select="https_protocol" > > # external library protocols > libamqp_protocol_deps="librabbitmq" > diff --git a/doc/protocols.texi b/doc/protocols.texi > index d207df0b52..90a9eefde0 100644 > --- a/doc/protocols.texi > +++ b/doc/protocols.texi > @@ -2025,5 +2025,35 @@ decoding errors. > > @end table > > +@section ipfs > + > +InterPlanetary File System (IPFS) protocol support. One can access files > stored > +on the IPFS network through so called gateways. Those are http(s) > endpoints. > +This protocol wraps the IPFS native protocols (ipfs:// and ipns://) to be > send > +to such a gateway. Users can (and should) host their own node which means > this > +protocol will use your local gateway to access files on the IPFS network. > + > +If a user doesn't have a node of their own then the public gateway > dweb.link is > +used by default. > + > +You can use this protocol in 2 ways. Using IPFS: > +@example > +ffplay ipfs://QmbGtJg23skhvFmu9mJiePVByhfzu5rwo74MEkVDYAmF5T > +@end example > + > +Or the IPNS protocol (IPNS is mutable IPFS): > +@example > +ffplay ipns://QmbGtJg23skhvFmu9mJiePVByhfzu5rwo74MEkVDYAmF5T > +@end example > + > +You can also change the gateway to be used: > + > +@table @option > + > +@item gateway > +Defines the gateway to use. When nothing is provided the protocol will > first try > +your local gateway. If that fails dweb.link will be used. > + > +@end table > > @c man end PROTOCOLS > diff --git a/libavformat/Makefile b/libavformat/Makefile > index d7182d6bd8..e3233fd7ac 100644 > --- a/libavformat/Makefile > +++ b/libavformat/Makefile > @@ -660,6 +660,8 @@ OBJS-$(CONFIG_SRTP_PROTOCOL) += > srtpproto.o srtp.o > OBJS-$(CONFIG_SUBFILE_PROTOCOL) += subfile.o > OBJS-$(CONFIG_TEE_PROTOCOL) += teeproto.o tee_common.o > OBJS-$(CONFIG_TCP_PROTOCOL) += tcp.o > +OBJS-$(CONFIG_IPFS_PROTOCOL) += ipfsgateway.o > +OBJS-$(CONFIG_IPNS_PROTOCOL) += ipfsgateway.o > TLS-OBJS-$(CONFIG_GNUTLS) += tls_gnutls.o > TLS-OBJS-$(CONFIG_LIBTLS) += tls_libtls.o > TLS-OBJS-$(CONFIG_MBEDTLS) += tls_mbedtls.o > diff --git a/libavformat/ipfsgateway.c b/libavformat/ipfsgateway.c > new file mode 100644 > index 0000000000..05fd37bbc5 > --- /dev/null > +++ b/libavformat/ipfsgateway.c > @@ -0,0 +1,339 @@ > +/* > + * IPFS and IPNS protocol support through IPFS Gateway. > + * Copyright (c) 2022 Mark Gaiser > + * > + * This file is part of FFmpeg. > + * > + * FFmpeg is free software; you can redistribute it and/or > + * modify it under the terms of the GNU Lesser General Public > + * License as published by the Free Software Foundation; either > + * version 2.1 of the License, or (at your option) any later version. > + * > + * FFmpeg is distributed in the hope that it will be useful, > + * but WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > + * Lesser General Public License for more details. > + * > + * You should have received a copy of the GNU Lesser General Public > + * License along with FFmpeg; if not, write to the Free Software > + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA > 02110-1301 USA > + */ > + > +#include "libavutil/avstring.h" > +#include "libavutil/opt.h" > +#include <sys/stat.h> > +#include "os_support.h" > +#include "url.h" > + > +typedef struct IPFSGatewayContext { > + AVClass *class; > + URLContext *inner; > + // Is filled by the -gateway argument and not changed after. > + char *gateway; > + // If the above gateway is non null, it will be copied into this > buffer. > + // Else this buffer will contain the auto detected gateway. > + // In either case, the gateway to use will be in this buffer. > + char gateway_buffer[PATH_MAX]; > +} IPFSGatewayContext; > + > +// A best-effort way to find the IPFS gateway. > +// Only the most appropiate gateway is set. It's not actually requested > +// (http call) to prevent a potential slowdown in startup. A potential > timeout > +// is handled by the HTTP protocol. > +static int populate_ipfs_gateway(URLContext *h) > +{ > + IPFSGatewayContext *c = h->priv_data; > + char ipfs_full_data_folder[PATH_MAX]; > + char ipfs_gateway_file[PATH_MAX]; > + struct stat st; > + int stat_ret = 0; > + int ret = AVERROR(EINVAL); > + FILE *gateway_file = NULL; > + > + // Test $IPFS_GATEWAY. > + if (getenv("IPFS_GATEWAY") != NULL) { > + if (snprintf(c->gateway_buffer, sizeof(c->gateway_buffer), "%s", > + getenv("IPFS_GATEWAY")) > + >= sizeof(c->gateway_buffer)) { > + av_log(h, AV_LOG_WARNING, > + "The IPFS_GATEWAY environment variable " > + "exceeds the maximum length. " > + "We allow a max of %zu characters\n", > + sizeof(c->gateway_buffer)); > + ret = AVERROR(EINVAL); > + goto err; > + } > + > + ret = 1; > + goto err; > + } else > + av_log(h, AV_LOG_DEBUG, "$IPFS_GATEWAY is empty.\n"); > + > + // We need to know the IPFS folder to - eventually - read the > contents of > + // the "gateway" file which would tell us the gateway to use. > + if (getenv("IPFS_PATH") == NULL) { > + av_log(h, AV_LOG_DEBUG, "$IPFS_PATH is empty.\n"); > + > + // Try via the home folder. > + if (getenv("HOME") == NULL) { > + av_log(h, AV_LOG_WARNING, "$HOME appears to be empty.\n"); > + ret = AVERROR(EINVAL); > + goto err; > + } > + > + // Verify the composed path fits. > + if (snprintf(ipfs_full_data_folder, sizeof(ipfs_full_data_folder), > + "%s/.ipfs/", getenv("HOME")) > + >= sizeof(ipfs_full_data_folder)) { > + av_log(h, AV_LOG_WARNING, > + "The IPFS data path exceeds the " > + "max path length (%zu)\n", > + sizeof(ipfs_full_data_folder)); > + ret = AVERROR(EINVAL); > + goto err; > + } > + > + // Stat the folder. > + // It should exist in a default IPFS setup when run as local user. > + stat_ret = stat(ipfs_full_data_folder, &st); > + > + if (stat_ret < 0) { > + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. We > tried:\n"); > + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was empty.\n"); > + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) which > doesn't exist.\n", ipfs_full_data_folder); > + ret = AVERROR(ENOENT); > + goto err; > + } > + } else { > + if (snprintf(ipfs_full_data_folder, > sizeof(ipfs_full_data_folder), "%s", > + getenv("IPFS_PATH")) > + >= sizeof(ipfs_full_data_folder)) { > + av_log(h, AV_LOG_WARNING, > + "The IPFS_PATH environment variable " > + "exceeds the maximum length. " > + "We allow a max of %zu characters\n", > + sizeof(c->gateway_buffer)); > + ret = AVERROR(EINVAL); > + goto err; > + } > + } > + > + // Copy the fully composed gateway path into ipfs_gateway_file. > + if (snprintf(ipfs_gateway_file, sizeof(ipfs_gateway_file), > "%sgateway", > + ipfs_full_data_folder) > + >= sizeof(ipfs_gateway_file)) { > + av_log(h, AV_LOG_WARNING, > + "The IPFS gateway file path exceeds " > + "the max path length (%zu)\n", > + sizeof(ipfs_gateway_file)); > + ret = AVERROR(ENOENT); > + goto err; > + } > + > + // Get the contents of the gateway file. > + gateway_file = av_fopen_utf8(ipfs_gateway_file, "r"); > + if (!gateway_file) { > + av_log(h, AV_LOG_WARNING, > + "The IPFS gateway file (full uri: %s) doesn't exist. " > + "Is the gateway enabled?\n", > + ipfs_gateway_file); > + ret = AVERROR(ENOENT); > + goto err; > + } > + > + // Read a single line (fgets stops at new line mark). > + if (!fgets(c->gateway_buffer, sizeof(c->gateway_buffer) - 1, > gateway_file)) { > + av_log(h, AV_LOG_WARNING, "Unable to read from file (full uri: > %s).\n", > + ipfs_gateway_file); > + ret = AVERROR(ENOENT); > + goto err; > + } > + > + // Replace first occurence of end of line with \0 > + c->gateway_buffer[strcspn(c->gateway_buffer, "\r\n")] = 0; > + > + // If strlen finds anything longer then 0 characters then we have a > + // potential gateway url. > + if (*c->gateway_buffer == '\0') { > + av_log(h, AV_LOG_WARNING, > + "The IPFS gateway file (full uri: %s) appears to be empty. > " > + "Is the gateway started?\n", > + ipfs_gateway_file); > + ret = AVERROR(EILSEQ); > + goto err; > + } else { > + // We're done, the c->gateway_buffer has something that looks > valid. > + ret = 1; > + goto err; > + } > + > +err: > + if (gateway_file) > + fclose(gateway_file); > + > + return ret; > +} > + > +static int translate_ipfs_to_http(URLContext *h, const char *uri, int > flags, AVDictionary **options) > +{ > + const char *ipfs_cid; > + char *fulluri = NULL; > + int ret; > + IPFSGatewayContext *c = h->priv_data; > + > + // Test for ipfs://, ipfs:, ipns:// and ipns:. This prefix is > stripped from > + // the string leaving just the CID in ipfs_cid. > + int is_ipfs = av_stristart(uri, "ipfs://", &ipfs_cid); > + int is_ipns = av_stristart(uri, "ipns://", &ipfs_cid); > + > + // We must have either ipns or ipfs. > + if (!is_ipfs && !is_ipns) { > + ret = AVERROR(EINVAL); > + av_log(h, AV_LOG_WARNING, "Unsupported url %s\n", uri); > + goto err; > + } > + > + // If the CID has a length greater then 0 then we assume we have a > proper working one. > + // It could still be wrong but in that case the gateway should save > us and > + // ruturn a 403 error. The http protocol handles this. > + if (strlen(ipfs_cid) < 1) { > + av_log(h, AV_LOG_WARNING, "A CID must be provided.\n"); > + ret = AVERROR(EILSEQ); > + goto err; > + } > + > + // Populate c->gateway_buffer with whatever is in c->gateway > + if (c->gateway != NULL) { > + if (snprintf(c->gateway_buffer, sizeof(c->gateway_buffer), "%s", > + c->gateway) > + >= sizeof(c->gateway_buffer)) { > + av_log(h, AV_LOG_WARNING, > + "The -gateway parameter is too long. " > + "We allow a max of %zu characters\n", > + sizeof(c->gateway_buffer)); > + ret = AVERROR(EINVAL); > + goto err; > + } > + } else { > + // Populate the IPFS gateway if we have any. > + // If not, inform the user how to properly set one. > + ret = populate_ipfs_gateway(h); > + > + if (ret < 1) { > + // We fallback on dweb.link (managed by Protocol Labs). > + snprintf(c->gateway_buffer, sizeof(c->gateway_buffer), " > https://dweb.link"); > + > + av_log(h, AV_LOG_WARNING, > + "IPFS does not appear to be running. " > + "You’re now using the public gateway at dweb.link.\n"); > + av_log(h, AV_LOG_INFO, > + "Installing IPFS locally is recommended to " > + "improve performance and reliability, " > + "and not share all your activity with a single IPFS > gateway.\n" > + "There are multiple options to define this gateway.\n" > + "1. Call ffmpeg with a gateway param, " > + "without a trailing slash: -gateway <url>.\n" > + "2. Define an $IPFS_GATEWAY environment variable with > the " > + "full HTTP URL to the gateway " > + "without trailing forward slash.\n" > + "3. Define an $IPFS_PATH environment variable " > + "and point it to the IPFS data path " > + "- this is typically ~/.ipfs\n"); > + } > + } > + > + // Test if the gateway starts with either http:// or https:// > + if (av_stristart(c->gateway_buffer, "http://", NULL) == 0 > + && av_stristart(c->gateway_buffer, "https://", NULL) == 0) { > + av_log(h, AV_LOG_WARNING, > + "The gateway URL didn't start with http:// or " > + "https:// and is therefore invalid.\n"); > + ret = AVERROR(EILSEQ); > + goto err; > + } > + > + // Concatenate the url. > + // This ends up with something like: http://localhost:8080/ipfs/Qm... > .. > + // The format of "%s%s%s%s" is the following: > + // 1st %s = The gateway. > + // 2nd %s = If the gateway didn't end in a slash, add a "/". > Otherwise it's an empty string > + // 3rd %s = Either ipns/ or ipfs/. > + // 4th %s = The IPFS CID (Qm..., bafy..., ...). > + fulluri = av_asprintf("%s%s%s%s", > + c->gateway_buffer, > + (c->gateway_buffer[strlen(c->gateway_buffer) - > 1] == '/') ? "" : "/", > + (is_ipns) ? "ipns/" : "ipfs/", > + ipfs_cid); > + > + if (!fulluri) { > + av_log(h, AV_LOG_ERROR, "Failed to compose the URL\n"); > + ret = AVERROR(ENOMEM); > + goto err; > + } > + > + // Pass the URL back to FFMpeg's protocol handler. > + ret = ffurl_open_whitelist(&c->inner, fulluri, flags, > + &h->interrupt_callback, options, > + h->protocol_whitelist, > + h->protocol_blacklist, h); > + if (ret < 0) { > + av_log(h, AV_LOG_WARNING, "Unable to open resource: %s\n", > fulluri); > + goto err; > + } > + > +err: > + av_free(fulluri); > + return ret; > +} > + > +static int ipfs_read(URLContext *h, unsigned char *buf, int size) > +{ > + IPFSGatewayContext *c = h->priv_data; > + return ffurl_read(c->inner, buf, size); > +} > + > +static int64_t ipfs_seek(URLContext *h, int64_t pos, int whence) > +{ > + IPFSGatewayContext *c = h->priv_data; > + return ffurl_seek(c->inner, pos, whence); > +} > + > +static int ipfs_close(URLContext *h) > +{ > + IPFSGatewayContext *c = h->priv_data; > + return ffurl_closep(&c->inner); > +} > + > +#define OFFSET(x) offsetof(IPFSGatewayContext, x) > + > +static const AVOption options[] = { > + {"gateway", "The gateway to ask for IPFS data.", OFFSET(gateway), > AV_OPT_TYPE_STRING, {.str = NULL}, 0, 0, AV_OPT_FLAG_DECODING_PARAM}, > + {NULL}, > +}; > + > +static const AVClass ipfs_context_class = { > + .class_name = "IPFS", > + .item_name = av_default_item_name, > + .option = options, > + .version = LIBAVUTIL_VERSION_INT, > +}; > + > +const URLProtocol ff_ipfs_protocol = { > + .name = "ipfs", > + .url_open2 = translate_ipfs_to_http, > + .url_read = ipfs_read, > + .url_seek = ipfs_seek, > + .url_close = ipfs_close, > + .priv_data_size = sizeof(IPFSGatewayContext), > + .priv_data_class = &ipfs_context_class, > +}; > + > +const URLProtocol ff_ipns_protocol = { > + .name = "ipns", > + .url_open2 = translate_ipfs_to_http, > + .url_read = ipfs_read, > + .url_seek = ipfs_seek, > + .url_close = ipfs_close, > + .priv_data_size = sizeof(IPFSGatewayContext), > + .priv_data_class = &ipfs_context_class, > +}; > diff --git a/libavformat/protocols.c b/libavformat/protocols.c > index d07563cd0c..6ee62a598a 100644 > --- a/libavformat/protocols.c > +++ b/libavformat/protocols.c > @@ -71,6 +71,8 @@ extern const URLProtocol ff_libsrt_protocol; > extern const URLProtocol ff_libssh_protocol; > extern const URLProtocol ff_libsmbclient_protocol; > extern const URLProtocol ff_libzmq_protocol; > +extern const URLProtocol ff_ipfs_protocol; > +extern const URLProtocol ff_ipns_protocol; > > #include "libavformat/protocol_list.c" > > -- > 2.35.1 > > I just keep pinging... :) Well.. This is the one to be merged, right?
On Wed, Apr 06, 2022 at 02:00:56PM +0200, Mark Gaiser wrote: [...] > + if (stat_ret < 0) { > + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. We tried:\n"); > + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was empty.\n"); > + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) which doesn't exist.\n", ipfs_full_data_folder); The 3 av_log() can also be combined If nothing else is found then ill change that myself and apply in a day or 2 thx [...]
On Sun, Apr 10, 2022 at 4:41 PM Michael Niedermayer <michael@niedermayer.cc> wrote: > On Wed, Apr 06, 2022 at 02:00:56PM +0200, Mark Gaiser wrote: > [...] > > > + if (stat_ret < 0) { > > > + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. We > tried:\n"); > > + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was empty.\n"); > > + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) which > doesn't exist.\n", ipfs_full_data_folder); > > The 3 av_log() can also be combined > > If nothing else is found then ill change that myself and apply in a day or > 2 > That would be awesome!! I'm looking forward to that very much. > > thx > > [...] > -- > Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB > > Awnsering whenever a program halts or runs forever is > On a turing machine, in general impossible (turings halting problem). > On any real computer, always possible as a real computer has a finite > number > of states N, and will either halt in less than N cycles or never halt. > _______________________________________________ > ffmpeg-devel mailing list > ffmpeg-devel@ffmpeg.org > https://ffmpeg.org/mailman/listinfo/ffmpeg-devel > > To unsubscribe, visit link above, or email > ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe". >
On Sun, Apr 10, 2022 at 04:41:27PM +0200, Michael Niedermayer wrote: > On Wed, Apr 06, 2022 at 02:00:56PM +0200, Mark Gaiser wrote: > [...] > > > + if (stat_ret < 0) { > > > + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. We tried:\n"); > > + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was empty.\n"); > > + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) which doesn't exist.\n", ipfs_full_data_folder); > > The 3 av_log() can also be combined > > If nothing else is found then ill change that myself and apply in a day or 2 applied thx [...]
On Tue, Apr 12, 2022 at 8:03 PM Michael Niedermayer <michael@niedermayer.cc> wrote: > On Sun, Apr 10, 2022 at 04:41:27PM +0200, Michael Niedermayer wrote: > > On Wed, Apr 06, 2022 at 02:00:56PM +0200, Mark Gaiser wrote: > > [...] > > > > > + if (stat_ret < 0) { > > > > > + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. We > tried:\n"); > > > + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was > empty.\n"); > > > + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) > which doesn't exist.\n", ipfs_full_data_folder); > > > > The 3 av_log() can also be combined > > > > If nothing else is found then ill change that myself and apply in a day > or 2 > > applied > Awesome! Thank you so much! :) Do you plan to add it to the changelog for 5.1 too? Or do I need to make a new patch for that? ;) > > thx > > [...] > > -- > Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB > > No great genius has ever existed without some touch of madness. -- > Aristotle > _______________________________________________ > ffmpeg-devel mailing list > ffmpeg-devel@ffmpeg.org > https://ffmpeg.org/mailman/listinfo/ffmpeg-devel > > To unsubscribe, visit link above, or email > ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe". >
On 4/12/2022 3:02 PM, Michael Niedermayer wrote: > On Sun, Apr 10, 2022 at 04:41:27PM +0200, Michael Niedermayer wrote: >> On Wed, Apr 06, 2022 at 02:00:56PM +0200, Mark Gaiser wrote: >> [...] >> >>> + if (stat_ret < 0) { >> >>> + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. We tried:\n"); >>> + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was empty.\n"); >>> + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) which doesn't exist.\n", ipfs_full_data_folder); >> >> The 3 av_log() can also be combined >> >> If nothing else is found then ill change that myself and apply in a day or 2 > > applied > > thx This broke msvc > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): error C2065: 'PATH_MAX': undeclared identifier > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): error C2057: expected constant expression > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): error C2065: 'PATH_MAX': undeclared identifier > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): error C2057: expected constant expression > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): error C2466: cannot allocate an array of constant size 0 > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): error C2133: 'ipfs_full_data_folder': unknown size > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): error C2065: 'PATH_MAX': undeclared identifier > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): error C2057: expected constant expression > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): error C2466: cannot allocate an array of constant size 0 > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): error C2133: 'ipfs_gateway_file': unknown size PATH_MAX seems to be POSIX only.
On Wed, Apr 13, 2022 at 1:10 AM James Almer <jamrial@gmail.com> wrote: > > > On 4/12/2022 3:02 PM, Michael Niedermayer wrote: > > On Sun, Apr 10, 2022 at 04:41:27PM +0200, Michael Niedermayer wrote: > >> On Wed, Apr 06, 2022 at 02:00:56PM +0200, Mark Gaiser wrote: > >> [...] > >> > >>> + if (stat_ret < 0) { > >> > >>> + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. We > tried:\n"); > >>> + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was > empty.\n"); > >>> + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) > which doesn't exist.\n", ipfs_full_data_folder); > >> > >> The 3 av_log() can also be combined > >> > >> If nothing else is found then ill change that myself and apply in a day > or 2 > > > > applied > > > > thx > > This broke msvc > > > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): > error C2065: 'PATH_MAX': undeclared identifier > > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): > error C2057: expected constant expression > > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > error C2065: 'PATH_MAX': undeclared identifier > > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > error C2057: expected constant expression > > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > error C2466: cannot allocate an array of constant size 0 > > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > error C2133: 'ipfs_full_data_folder': unknown size > > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > error C2065: 'PATH_MAX': undeclared identifier > > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > error C2057: expected constant expression > > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > error C2466: cannot allocate an array of constant size 0 > > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > error C2133: 'ipfs_gateway_file': unknown size > > PATH_MAX seems to be POSIX only. > Seriously? Shouldn't that define work on windows too? And why didn't patchwork catch this? What is the intended define to use for this? Note that it used to be a fixed number in the code itself (4096?) but I changed it to PATH_MAX as per review suggestions.... I grepped over the ffmpeg source. This code is indeed the first cross platform code in ffmpeg that uses it. The sole other user is v4l2_m2m.h but that's obviously for windows only. > _______________________________________________ > ffmpeg-devel mailing list > ffmpeg-devel@ffmpeg.org > https://ffmpeg.org/mailman/listinfo/ffmpeg-devel > > To unsubscribe, visit link above, or email > ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe". >
Mark Gaiser: > On Wed, Apr 13, 2022 at 1:10 AM James Almer <jamrial@gmail.com> wrote: > >> >> >> On 4/12/2022 3:02 PM, Michael Niedermayer wrote: >>> On Sun, Apr 10, 2022 at 04:41:27PM +0200, Michael Niedermayer wrote: >>>> On Wed, Apr 06, 2022 at 02:00:56PM +0200, Mark Gaiser wrote: >>>> [...] >>>> >>>>> + if (stat_ret < 0) { >>>> >>>>> + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. We >> tried:\n"); >>>>> + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was >> empty.\n"); >>>>> + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) >> which doesn't exist.\n", ipfs_full_data_folder); >>>> >>>> The 3 av_log() can also be combined >>>> >>>> If nothing else is found then ill change that myself and apply in a day >> or 2 >>> >>> applied >>> >>> thx >> >> This broke msvc >> >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): >> error C2065: 'PATH_MAX': undeclared identifier >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): >> error C2057: expected constant expression >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): >> error C2065: 'PATH_MAX': undeclared identifier >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): >> error C2057: expected constant expression >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): >> error C2466: cannot allocate an array of constant size 0 >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): >> error C2133: 'ipfs_full_data_folder': unknown size >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): >> error C2065: 'PATH_MAX': undeclared identifier >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): >> error C2057: expected constant expression >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): >> error C2466: cannot allocate an array of constant size 0 >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): >> error C2133: 'ipfs_gateway_file': unknown size >> >> PATH_MAX seems to be POSIX only. >> > > Seriously? > > Shouldn't that define work on windows too? For "that define" to work that define needs to exist. Windows has MAX_PATH, but it is pretty small (260). > And why didn't patchwork catch this? > Because there is no MSVC patchwork box. > What is the intended define to use for this? > Note that it used to be a fixed number in the code itself (4096?) but I > changed it to PATH_MAX as per review suggestions.... Then how about just adding #ifndef MAX_PATH #define MAX_PATH 4096 #endif or something like that. > I grepped over the ffmpeg source. This code is indeed the first cross > platform code in ffmpeg that uses it. The sole other user is v4l2_m2m.h but > that's obviously for windows only. You mean "not for Windows". - Andreas
On Wed, Apr 13, 2022 at 2:28 PM Andreas Rheinhardt < andreas.rheinhardt@outlook.com> wrote: > Mark Gaiser: > > On Wed, Apr 13, 2022 at 1:10 AM James Almer <jamrial@gmail.com> wrote: > > > >> > >> > >> On 4/12/2022 3:02 PM, Michael Niedermayer wrote: > >>> On Sun, Apr 10, 2022 at 04:41:27PM +0200, Michael Niedermayer wrote: > >>>> On Wed, Apr 06, 2022 at 02:00:56PM +0200, Mark Gaiser wrote: > >>>> [...] > >>>> > >>>>> + if (stat_ret < 0) { > >>>> > >>>>> + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. We > >> tried:\n"); > >>>>> + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was > >> empty.\n"); > >>>>> + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) > >> which doesn't exist.\n", ipfs_full_data_folder); > >>>> > >>>> The 3 av_log() can also be combined > >>>> > >>>> If nothing else is found then ill change that myself and apply in a > day > >> or 2 > >>> > >>> applied > >>> > >>> thx > >> > >> This broke msvc > >> > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): > >> error C2065: 'PATH_MAX': undeclared identifier > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): > >> error C2057: expected constant expression > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > >> error C2065: 'PATH_MAX': undeclared identifier > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > >> error C2057: expected constant expression > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > >> error C2466: cannot allocate an array of constant size 0 > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > >> error C2133: 'ipfs_full_data_folder': unknown size > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > >> error C2065: 'PATH_MAX': undeclared identifier > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > >> error C2057: expected constant expression > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > >> error C2466: cannot allocate an array of constant size 0 > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > >> error C2133: 'ipfs_gateway_file': unknown size > >> > >> PATH_MAX seems to be POSIX only. > >> > > > > Seriously? > > > > Shouldn't that define work on windows too? > > For "that define" to work that define needs to exist. Windows has > MAX_PATH, but it is pretty small (260). > > > And why didn't patchwork catch this? > > > > Because there is no MSVC patchwork box. > > > What is the intended define to use for this? > > Note that it used to be a fixed number in the code itself (4096?) but I > > changed it to PATH_MAX as per review suggestions.... > > Then how about just adding > #ifndef MAX_PATH > #define MAX_PATH 4096 > #endif > or something like that. > I'm fine with that but given past feedback I'll wait a little longer for a 2nd opinion too. > > > I grepped over the ffmpeg source. This code is indeed the first cross > > platform code in ffmpeg that uses it. The sole other user is v4l2_m2m.h > but > > that's obviously for windows only. > > You mean "not for Windows". > Oops. Yes, exactly! > - Andreas > _______________________________________________ > ffmpeg-devel mailing list > ffmpeg-devel@ffmpeg.org > https://ffmpeg.org/mailman/listinfo/ffmpeg-devel > > To unsubscribe, visit link above, or email > ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe". >
On Wed, Apr 13, 2022 at 3:14 PM Mark Gaiser <markg85@gmail.com> wrote: > > On Wed, Apr 13, 2022 at 2:28 PM Andreas Rheinhardt < > andreas.rheinhardt@outlook.com> wrote: > > > Mark Gaiser: > > > On Wed, Apr 13, 2022 at 1:10 AM James Almer <jamrial@gmail.com> wrote: > > > > > >> > > >> > > >> On 4/12/2022 3:02 PM, Michael Niedermayer wrote: > > >>> On Sun, Apr 10, 2022 at 04:41:27PM +0200, Michael Niedermayer wrote: > > >>>> On Wed, Apr 06, 2022 at 02:00:56PM +0200, Mark Gaiser wrote: > > >>>> [...] > > >>>> > > >>>>> + if (stat_ret < 0) { > > >>>> > > >>>>> + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. We > > >> tried:\n"); > > >>>>> + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was > > >> empty.\n"); > > >>>>> + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) > > >> which doesn't exist.\n", ipfs_full_data_folder); > > >>>> > > >>>> The 3 av_log() can also be combined > > >>>> > > >>>> If nothing else is found then ill change that myself and apply in a > > day > > >> or 2 > > >>> > > >>> applied > > >>> > > >>> thx > > >> > > >> This broke msvc > > >> > > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): > > >> error C2065: 'PATH_MAX': undeclared identifier > > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): > > >> error C2057: expected constant expression > > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > > >> error C2065: 'PATH_MAX': undeclared identifier > > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > > >> error C2057: expected constant expression > > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > > >> error C2466: cannot allocate an array of constant size 0 > > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > > >> error C2133: 'ipfs_full_data_folder': unknown size > > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > > >> error C2065: 'PATH_MAX': undeclared identifier > > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > > >> error C2057: expected constant expression > > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > > >> error C2466: cannot allocate an array of constant size 0 > > >>> C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > > >> error C2133: 'ipfs_gateway_file': unknown size > > >> > > >> PATH_MAX seems to be POSIX only. > > >> > > > > > > Seriously? > > > > > > Shouldn't that define work on windows too? > > > > For "that define" to work that define needs to exist. Windows has > > MAX_PATH, but it is pretty small (260). > > > > > And why didn't patchwork catch this? > > > > > > > Because there is no MSVC patchwork box. > > > > > What is the intended define to use for this? > > > Note that it used to be a fixed number in the code itself (4096?) but I > > > changed it to PATH_MAX as per review suggestions.... > > > > Then how about just adding > > #ifndef MAX_PATH > > #define MAX_PATH 4096 > > #endif > > or something like that. > > > > I'm fine with that but given past feedback I'll wait a little longer for a > 2nd opinion too. > master is currently broken, waiting is a luxury for another time. - Hendrik
On Wed, Apr 13, 2022 at 4:45 PM Hendrik Leppkes <h.leppkes@gmail.com> wrote: > On Wed, Apr 13, 2022 at 3:14 PM Mark Gaiser <markg85@gmail.com> wrote: > > > > On Wed, Apr 13, 2022 at 2:28 PM Andreas Rheinhardt < > > andreas.rheinhardt@outlook.com> wrote: > > > > > Mark Gaiser: > > > > On Wed, Apr 13, 2022 at 1:10 AM James Almer <jamrial@gmail.com> > wrote: > > > > > > > >> > > > >> > > > >> On 4/12/2022 3:02 PM, Michael Niedermayer wrote: > > > >>> On Sun, Apr 10, 2022 at 04:41:27PM +0200, Michael Niedermayer > wrote: > > > >>>> On Wed, Apr 06, 2022 at 02:00:56PM +0200, Mark Gaiser wrote: > > > >>>> [...] > > > >>>> > > > >>>>> + if (stat_ret < 0) { > > > >>>> > > > >>>>> + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. > We > > > >> tried:\n"); > > > >>>>> + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was > > > >> empty.\n"); > > > >>>>> + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) > > > >> which doesn't exist.\n", ipfs_full_data_folder); > > > >>>> > > > >>>> The 3 av_log() can also be combined > > > >>>> > > > >>>> If nothing else is found then ill change that myself and apply in > a > > > day > > > >> or 2 > > > >>> > > > >>> applied > > > >>> > > > >>> thx > > > >> > > > >> This broke msvc > > > >> > > > >>> > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): > > > >> error C2065: 'PATH_MAX': undeclared identifier > > > >>> > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(36): > > > >> error C2057: expected constant expression > > > >>> > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > > > >> error C2065: 'PATH_MAX': undeclared identifier > > > >>> > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > > > >> error C2057: expected constant expression > > > >>> > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > > > >> error C2466: cannot allocate an array of constant size 0 > > > >>> > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(46): > > > >> error C2133: 'ipfs_full_data_folder': unknown size > > > >>> > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > > > >> error C2065: 'PATH_MAX': undeclared identifier > > > >>> > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > > > >> error C2057: expected constant expression > > > >>> > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > > > >> error C2466: cannot allocate an array of constant size 0 > > > >>> > C:/Dev/ffmpeg/fate/x86_32-msvc-MD/src/libavformat/ipfsgateway.c(47): > > > >> error C2133: 'ipfs_gateway_file': unknown size > > > >> > > > >> PATH_MAX seems to be POSIX only. > > > >> > > > > > > > > Seriously? > > > > > > > > Shouldn't that define work on windows too? > > > > > > For "that define" to work that define needs to exist. Windows has > > > MAX_PATH, but it is pretty small (260). > > > > > > > And why didn't patchwork catch this? > > > > > > > > > > Because there is no MSVC patchwork box. > > > > > > > What is the intended define to use for this? > > > > Note that it used to be a fixed number in the code itself (4096?) > but I > > > > changed it to PATH_MAX as per review suggestions.... > > > > > > Then how about just adding > > > #ifndef MAX_PATH > > > #define MAX_PATH 4096 > > > #endif > > > or something like that. > > > > > > > I'm fine with that but given past feedback I'll wait a little longer for > a > > 2nd opinion too. > > > > master is currently broken, waiting is a luxury for another time. > Fine. Patch incoming. > - Hendrik > _______________________________________________ > ffmpeg-devel mailing list > ffmpeg-devel@ffmpeg.org > https://ffmpeg.org/mailman/listinfo/ffmpeg-devel > > To unsubscribe, visit link above, or email > ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe". >
diff --git a/configure b/configure index e4d36aa639..55af90957a 100755 --- a/configure +++ b/configure @@ -3579,6 +3579,8 @@ udp_protocol_select="network" udplite_protocol_select="network" unix_protocol_deps="sys_un_h" unix_protocol_select="network" +ipfs_protocol_select="https_protocol" +ipns_protocol_select="https_protocol" # external library protocols libamqp_protocol_deps="librabbitmq" diff --git a/doc/protocols.texi b/doc/protocols.texi index d207df0b52..90a9eefde0 100644 --- a/doc/protocols.texi +++ b/doc/protocols.texi @@ -2025,5 +2025,35 @@ decoding errors. @end table +@section ipfs + +InterPlanetary File System (IPFS) protocol support. One can access files stored +on the IPFS network through so called gateways. Those are http(s) endpoints. +This protocol wraps the IPFS native protocols (ipfs:// and ipns://) to be send +to such a gateway. Users can (and should) host their own node which means this +protocol will use your local gateway to access files on the IPFS network. + +If a user doesn't have a node of their own then the public gateway dweb.link is +used by default. + +You can use this protocol in 2 ways. Using IPFS: +@example +ffplay ipfs://QmbGtJg23skhvFmu9mJiePVByhfzu5rwo74MEkVDYAmF5T +@end example + +Or the IPNS protocol (IPNS is mutable IPFS): +@example +ffplay ipns://QmbGtJg23skhvFmu9mJiePVByhfzu5rwo74MEkVDYAmF5T +@end example + +You can also change the gateway to be used: + +@table @option + +@item gateway +Defines the gateway to use. When nothing is provided the protocol will first try +your local gateway. If that fails dweb.link will be used. + +@end table @c man end PROTOCOLS diff --git a/libavformat/Makefile b/libavformat/Makefile index d7182d6bd8..e3233fd7ac 100644 --- a/libavformat/Makefile +++ b/libavformat/Makefile @@ -660,6 +660,8 @@ OBJS-$(CONFIG_SRTP_PROTOCOL) += srtpproto.o srtp.o OBJS-$(CONFIG_SUBFILE_PROTOCOL) += subfile.o OBJS-$(CONFIG_TEE_PROTOCOL) += teeproto.o tee_common.o OBJS-$(CONFIG_TCP_PROTOCOL) += tcp.o +OBJS-$(CONFIG_IPFS_PROTOCOL) += ipfsgateway.o +OBJS-$(CONFIG_IPNS_PROTOCOL) += ipfsgateway.o TLS-OBJS-$(CONFIG_GNUTLS) += tls_gnutls.o TLS-OBJS-$(CONFIG_LIBTLS) += tls_libtls.o TLS-OBJS-$(CONFIG_MBEDTLS) += tls_mbedtls.o diff --git a/libavformat/ipfsgateway.c b/libavformat/ipfsgateway.c new file mode 100644 index 0000000000..05fd37bbc5 --- /dev/null +++ b/libavformat/ipfsgateway.c @@ -0,0 +1,339 @@ +/* + * IPFS and IPNS protocol support through IPFS Gateway. + * Copyright (c) 2022 Mark Gaiser + * + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * FFmpeg is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include "libavutil/avstring.h" +#include "libavutil/opt.h" +#include <sys/stat.h> +#include "os_support.h" +#include "url.h" + +typedef struct IPFSGatewayContext { + AVClass *class; + URLContext *inner; + // Is filled by the -gateway argument and not changed after. + char *gateway; + // If the above gateway is non null, it will be copied into this buffer. + // Else this buffer will contain the auto detected gateway. + // In either case, the gateway to use will be in this buffer. + char gateway_buffer[PATH_MAX]; +} IPFSGatewayContext; + +// A best-effort way to find the IPFS gateway. +// Only the most appropiate gateway is set. It's not actually requested +// (http call) to prevent a potential slowdown in startup. A potential timeout +// is handled by the HTTP protocol. +static int populate_ipfs_gateway(URLContext *h) +{ + IPFSGatewayContext *c = h->priv_data; + char ipfs_full_data_folder[PATH_MAX]; + char ipfs_gateway_file[PATH_MAX]; + struct stat st; + int stat_ret = 0; + int ret = AVERROR(EINVAL); + FILE *gateway_file = NULL; + + // Test $IPFS_GATEWAY. + if (getenv("IPFS_GATEWAY") != NULL) { + if (snprintf(c->gateway_buffer, sizeof(c->gateway_buffer), "%s", + getenv("IPFS_GATEWAY")) + >= sizeof(c->gateway_buffer)) { + av_log(h, AV_LOG_WARNING, + "The IPFS_GATEWAY environment variable " + "exceeds the maximum length. " + "We allow a max of %zu characters\n", + sizeof(c->gateway_buffer)); + ret = AVERROR(EINVAL); + goto err; + } + + ret = 1; + goto err; + } else + av_log(h, AV_LOG_DEBUG, "$IPFS_GATEWAY is empty.\n"); + + // We need to know the IPFS folder to - eventually - read the contents of + // the "gateway" file which would tell us the gateway to use. + if (getenv("IPFS_PATH") == NULL) { + av_log(h, AV_LOG_DEBUG, "$IPFS_PATH is empty.\n"); + + // Try via the home folder. + if (getenv("HOME") == NULL) { + av_log(h, AV_LOG_WARNING, "$HOME appears to be empty.\n"); + ret = AVERROR(EINVAL); + goto err; + } + + // Verify the composed path fits. + if (snprintf(ipfs_full_data_folder, sizeof(ipfs_full_data_folder), + "%s/.ipfs/", getenv("HOME")) + >= sizeof(ipfs_full_data_folder)) { + av_log(h, AV_LOG_WARNING, + "The IPFS data path exceeds the " + "max path length (%zu)\n", + sizeof(ipfs_full_data_folder)); + ret = AVERROR(EINVAL); + goto err; + } + + // Stat the folder. + // It should exist in a default IPFS setup when run as local user. + stat_ret = stat(ipfs_full_data_folder, &st); + + if (stat_ret < 0) { + av_log(h, AV_LOG_INFO, "Unable to find IPFS folder. We tried:\n"); + av_log(h, AV_LOG_INFO, "- $IPFS_PATH, which was empty.\n"); + av_log(h, AV_LOG_INFO, "- $HOME/.ipfs (full uri: %s) which doesn't exist.\n", ipfs_full_data_folder); + ret = AVERROR(ENOENT); + goto err; + } + } else { + if (snprintf(ipfs_full_data_folder, sizeof(ipfs_full_data_folder), "%s", + getenv("IPFS_PATH")) + >= sizeof(ipfs_full_data_folder)) { + av_log(h, AV_LOG_WARNING, + "The IPFS_PATH environment variable " + "exceeds the maximum length. " + "We allow a max of %zu characters\n", + sizeof(c->gateway_buffer)); + ret = AVERROR(EINVAL); + goto err; + } + } + + // Copy the fully composed gateway path into ipfs_gateway_file. + if (snprintf(ipfs_gateway_file, sizeof(ipfs_gateway_file), "%sgateway", + ipfs_full_data_folder) + >= sizeof(ipfs_gateway_file)) { + av_log(h, AV_LOG_WARNING, + "The IPFS gateway file path exceeds " + "the max path length (%zu)\n", + sizeof(ipfs_gateway_file)); + ret = AVERROR(ENOENT); + goto err; + } + + // Get the contents of the gateway file. + gateway_file = av_fopen_utf8(ipfs_gateway_file, "r"); + if (!gateway_file) { + av_log(h, AV_LOG_WARNING, + "The IPFS gateway file (full uri: %s) doesn't exist. " + "Is the gateway enabled?\n", + ipfs_gateway_file); + ret = AVERROR(ENOENT); + goto err; + } + + // Read a single line (fgets stops at new line mark). + if (!fgets(c->gateway_buffer, sizeof(c->gateway_buffer) - 1, gateway_file)) { + av_log(h, AV_LOG_WARNING, "Unable to read from file (full uri: %s).\n", + ipfs_gateway_file); + ret = AVERROR(ENOENT); + goto err; + } + + // Replace first occurence of end of line with \0 + c->gateway_buffer[strcspn(c->gateway_buffer, "\r\n")] = 0; + + // If strlen finds anything longer then 0 characters then we have a + // potential gateway url. + if (*c->gateway_buffer == '\0') { + av_log(h, AV_LOG_WARNING, + "The IPFS gateway file (full uri: %s) appears to be empty. " + "Is the gateway started?\n", + ipfs_gateway_file); + ret = AVERROR(EILSEQ); + goto err; + } else { + // We're done, the c->gateway_buffer has something that looks valid. + ret = 1; + goto err; + } + +err: + if (gateway_file) + fclose(gateway_file); + + return ret; +} + +static int translate_ipfs_to_http(URLContext *h, const char *uri, int flags, AVDictionary **options) +{ + const char *ipfs_cid; + char *fulluri = NULL; + int ret; + IPFSGatewayContext *c = h->priv_data; + + // Test for ipfs://, ipfs:, ipns:// and ipns:. This prefix is stripped from + // the string leaving just the CID in ipfs_cid. + int is_ipfs = av_stristart(uri, "ipfs://", &ipfs_cid); + int is_ipns = av_stristart(uri, "ipns://", &ipfs_cid); + + // We must have either ipns or ipfs. + if (!is_ipfs && !is_ipns) { + ret = AVERROR(EINVAL); + av_log(h, AV_LOG_WARNING, "Unsupported url %s\n", uri); + goto err; + } + + // If the CID has a length greater then 0 then we assume we have a proper working one. + // It could still be wrong but in that case the gateway should save us and + // ruturn a 403 error. The http protocol handles this. + if (strlen(ipfs_cid) < 1) { + av_log(h, AV_LOG_WARNING, "A CID must be provided.\n"); + ret = AVERROR(EILSEQ); + goto err; + } + + // Populate c->gateway_buffer with whatever is in c->gateway + if (c->gateway != NULL) { + if (snprintf(c->gateway_buffer, sizeof(c->gateway_buffer), "%s", + c->gateway) + >= sizeof(c->gateway_buffer)) { + av_log(h, AV_LOG_WARNING, + "The -gateway parameter is too long. " + "We allow a max of %zu characters\n", + sizeof(c->gateway_buffer)); + ret = AVERROR(EINVAL); + goto err; + } + } else { + // Populate the IPFS gateway if we have any. + // If not, inform the user how to properly set one. + ret = populate_ipfs_gateway(h); + + if (ret < 1) { + // We fallback on dweb.link (managed by Protocol Labs). + snprintf(c->gateway_buffer, sizeof(c->gateway_buffer), "https://dweb.link"); + + av_log(h, AV_LOG_WARNING, + "IPFS does not appear to be running. " + "You’re now using the public gateway at dweb.link.\n"); + av_log(h, AV_LOG_INFO, + "Installing IPFS locally is recommended to " + "improve performance and reliability, " + "and not share all your activity with a single IPFS gateway.\n" + "There are multiple options to define this gateway.\n" + "1. Call ffmpeg with a gateway param, " + "without a trailing slash: -gateway <url>.\n" + "2. Define an $IPFS_GATEWAY environment variable with the " + "full HTTP URL to the gateway " + "without trailing forward slash.\n" + "3. Define an $IPFS_PATH environment variable " + "and point it to the IPFS data path " + "- this is typically ~/.ipfs\n"); + } + } + + // Test if the gateway starts with either http:// or https:// + if (av_stristart(c->gateway_buffer, "http://", NULL) == 0 + && av_stristart(c->gateway_buffer, "https://", NULL) == 0) { + av_log(h, AV_LOG_WARNING, + "The gateway URL didn't start with http:// or " + "https:// and is therefore invalid.\n"); + ret = AVERROR(EILSEQ); + goto err; + } + + // Concatenate the url. + // This ends up with something like: http://localhost:8080/ipfs/Qm..... + // The format of "%s%s%s%s" is the following: + // 1st %s = The gateway. + // 2nd %s = If the gateway didn't end in a slash, add a "/". Otherwise it's an empty string + // 3rd %s = Either ipns/ or ipfs/. + // 4th %s = The IPFS CID (Qm..., bafy..., ...). + fulluri = av_asprintf("%s%s%s%s", + c->gateway_buffer, + (c->gateway_buffer[strlen(c->gateway_buffer) - 1] == '/') ? "" : "/", + (is_ipns) ? "ipns/" : "ipfs/", + ipfs_cid); + + if (!fulluri) { + av_log(h, AV_LOG_ERROR, "Failed to compose the URL\n"); + ret = AVERROR(ENOMEM); + goto err; + } + + // Pass the URL back to FFMpeg's protocol handler. + ret = ffurl_open_whitelist(&c->inner, fulluri, flags, + &h->interrupt_callback, options, + h->protocol_whitelist, + h->protocol_blacklist, h); + if (ret < 0) { + av_log(h, AV_LOG_WARNING, "Unable to open resource: %s\n", fulluri); + goto err; + } + +err: + av_free(fulluri); + return ret; +} + +static int ipfs_read(URLContext *h, unsigned char *buf, int size) +{ + IPFSGatewayContext *c = h->priv_data; + return ffurl_read(c->inner, buf, size); +} + +static int64_t ipfs_seek(URLContext *h, int64_t pos, int whence) +{ + IPFSGatewayContext *c = h->priv_data; + return ffurl_seek(c->inner, pos, whence); +} + +static int ipfs_close(URLContext *h) +{ + IPFSGatewayContext *c = h->priv_data; + return ffurl_closep(&c->inner); +} + +#define OFFSET(x) offsetof(IPFSGatewayContext, x) + +static const AVOption options[] = { + {"gateway", "The gateway to ask for IPFS data.", OFFSET(gateway), AV_OPT_TYPE_STRING, {.str = NULL}, 0, 0, AV_OPT_FLAG_DECODING_PARAM}, + {NULL}, +}; + +static const AVClass ipfs_context_class = { + .class_name = "IPFS", + .item_name = av_default_item_name, + .option = options, + .version = LIBAVUTIL_VERSION_INT, +}; + +const URLProtocol ff_ipfs_protocol = { + .name = "ipfs", + .url_open2 = translate_ipfs_to_http, + .url_read = ipfs_read, + .url_seek = ipfs_seek, + .url_close = ipfs_close, + .priv_data_size = sizeof(IPFSGatewayContext), + .priv_data_class = &ipfs_context_class, +}; + +const URLProtocol ff_ipns_protocol = { + .name = "ipns", + .url_open2 = translate_ipfs_to_http, + .url_read = ipfs_read, + .url_seek = ipfs_seek, + .url_close = ipfs_close, + .priv_data_size = sizeof(IPFSGatewayContext), + .priv_data_class = &ipfs_context_class, +}; diff --git a/libavformat/protocols.c b/libavformat/protocols.c index d07563cd0c..6ee62a598a 100644 --- a/libavformat/protocols.c +++ b/libavformat/protocols.c @@ -71,6 +71,8 @@ extern const URLProtocol ff_libsrt_protocol; extern const URLProtocol ff_libssh_protocol; extern const URLProtocol ff_libsmbclient_protocol; extern const URLProtocol ff_libzmq_protocol; +extern const URLProtocol ff_ipfs_protocol; +extern const URLProtocol ff_ipns_protocol; #include "libavformat/protocol_list.c"
This patch adds support for: - ffplay ipfs://<cid> - ffplay ipns://<cid> IPFS data can be played from so called "ipfs gateways". A gateway is essentially a webserver that gives access to the distributed IPFS network. This protocol support (ipfs and ipns) therefore translates ipfs:// and ipns:// to a http:// url. This resulting url is then handled by the http protocol. It could also be https depending on the gateway provided. To use this protocol, a gateway must be provided. If you do nothing it will try to find it in your $HOME/.ipfs/gateway file. The ways to set it manually are: 1. Define a -gateway <url> to the gateway. 2. Define $IPFS_GATEWAY with the full http link to the gateway. 3. Define $IPFS_PATH and point it to the IPFS data path. 4. Have IPFS running in your local user folder (under $HOME/.ipfs). Signed-off-by: Mark Gaiser <markg85@gmail.com> --- configure | 2 + doc/protocols.texi | 30 ++++ libavformat/Makefile | 2 + libavformat/ipfsgateway.c | 339 ++++++++++++++++++++++++++++++++++++++ libavformat/protocols.c | 2 + 5 files changed, 375 insertions(+) create mode 100644 libavformat/ipfsgateway.c