moonfire-nvr/src/main.rs

247 lines
9.3 KiB
Rust
Raw Normal View History

Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
// This file is part of Moonfire NVR, a security camera digital video recorder.
// Copyright (C) 2016 Scott Lamb <slamb@slamb.org>
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// In addition, as a special exception, the copyright holders give
// permission to link the code of portions of this program with the
// OpenSSL library under certain conditions as described in each
// individual source file, and distribute linked combinations including
// the two.
//
// You must obey the GNU General Public License in all respects for all
// of the code used other than OpenSSL. If you modify file(s) with this
// exception, you may extend this exception to your version of the
// file(s), but you are not obligated to do so. If you do not wish to do
// so, delete this exception statement from your version. If you delete
// this exception statement from all source files in the program, then
// also delete it here.
//
// This program 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 General Public License for more details.
//
// You should have received a copy of the GNU General Public License
// along with this program. If not, see <http://www.gnu.org/licenses/>.
#![cfg_attr(all(feature="nightly", test), feature(test))]
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
extern crate byteorder;
extern crate core;
#[macro_use] extern crate chan;
extern crate chan_signal;
extern crate docopt;
#[macro_use] extern crate ffmpeg;
extern crate ffmpeg_sys;
extern crate fnv;
extern crate http_entity;
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
extern crate hyper;
#[macro_use] extern crate lazy_static;
extern crate libc;
#[macro_use] extern crate log;
extern crate lru_cache;
extern crate rusqlite;
extern crate memmap;
#[macro_use] extern crate mime;
extern crate openssl;
extern crate regex;
extern crate rustc_serialize;
extern crate serde;
extern crate serde_json;
extern crate slog;
extern crate slog_envlogger;
extern crate slog_stdlog;
extern crate slog_term;
extern crate smallvec;
extern crate time;
extern crate url;
extern crate uuid;
use hyper::server::Server;
use slog::DrainExt;
use std::path::Path;
use std::sync::Arc;
use std::sync::atomic::{AtomicBool, Ordering};
use std::thread;
2017-01-07 01:54:19 -05:00
mod check;
mod clock;
mod coding;
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
mod db;
mod dir;
mod error;
mod h264;
mod mmapfile;
mod mp4;
mod pieces;
mod recording;
mod stream;
mod streamer;
mod strutil;
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
#[cfg(test)] mod testutil;
mod upgrade;
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
mod web;
/// Commandline usage string. This is in the particular format expected by the `docopt` crate.
/// Besides being printed on --help or argument parsing error, it's actually parsed to define the
/// allowed commandline arguments and their defaults.
const USAGE: &'static str = "
Usage: moonfire-nvr [options]
moonfire-nvr --upgrade [options]
2017-01-07 01:54:19 -05:00
moonfire-nvr --check [options]
moonfire-nvr --ts <ts>...
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
moonfire-nvr (--help | --version)
Options:
-h, --help Show this message.
--version Show the version of moonfire-nvr.
--db-dir=DIR Set the directory holding the SQLite3 index database.
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
This is typically on a flash device.
[default: /var/lib/moonfire-nvr/db]
--sample-file-dir=DIR Set the directory holding video data.
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
This is typically on a hard drive.
[default: /var/lib/moonfire-nvr/sample]
--http-addr=ADDR Set the bind address for the unencrypted HTTP server.
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
[default: 0.0.0.0:8080]
--read-only Forces read-only mode / disables recording.
--preset-journal=MODE With --upgrade, resets the SQLite journal_mode to
the specified mode prior to the upgrade. The default,
delete, is recommended. off is very dangerous but
may be desirable in some circumstances. See
guide/schema.md for more information. The journal
mode will be reset to wal after the upgrade.
[default: delete]
--no-vacuum With --upgrade, skips the normal post-upgrade vacuum
operation.
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
";
/// Commandline arguments corresponding to `USAGE`; automatically filled by the `docopt` crate.
#[derive(RustcDecodable)]
struct Args {
flag_db_dir: String,
flag_sample_file_dir: String,
flag_http_addr: String,
flag_read_only: bool,
2017-01-07 01:54:19 -05:00
flag_check: bool,
flag_upgrade: bool,
flag_ts: bool,
flag_no_vacuum: bool,
flag_preset_journal: String,
arg_ts: Vec<String>,
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
}
fn main() {
// Parse commandline arguments.
let version = "Moonfire NVR 0.1.0".to_owned();
let args: Args = docopt::Docopt::new(USAGE)
.and_then(|d| d.version(Some(version)).decode())
.unwrap_or_else(|e| e.exit());
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
// Watch for termination signals.
// This must be started before any threads are spawned (such as the async logger thread) so
// that signals will be blocked in all threads.
let signal = chan_signal::notify(&[chan_signal::Signal::INT, chan_signal::Signal::TERM]);
// Initialize logging.
2017-01-07 01:54:19 -05:00
// Use async logging for serving because otherwise it blocks useful work.
// Use sync logging for other modes because async apparently is never flushed before the
// program exits, and partial output from these tools is very confusing.
let drain = slog_term::StreamerBuilder::new();
let drain = slog_envlogger::new(if args.flag_upgrade || args.flag_check { drain }
else { drain.async() }.full().build());
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
slog_stdlog::set_logger(slog::Logger::root(drain.ignore_err(), None)).unwrap();
// Open the database and populate cached state.
let db_dir = dir::Fd::open(&args.flag_db_dir).unwrap();
db_dir.lock(if args.flag_read_only { libc::LOCK_SH } else { libc::LOCK_EX } | libc::LOCK_NB)
.unwrap();
let conn = rusqlite::Connection::open_with_flags(
Path::new(&args.flag_db_dir).join("db"),
if args.flag_read_only {
rusqlite::SQLITE_OPEN_READ_ONLY
} else {
rusqlite::SQLITE_OPEN_READ_WRITE
} |
// rusqlite::Connection is not Sync, so there's no reason to tell SQLite3 to use the
// serialized threading mode.
rusqlite::SQLITE_OPEN_NO_MUTEX).unwrap();
if args.flag_upgrade {
upgrade::run(conn, &args.flag_preset_journal, args.flag_no_vacuum).unwrap();
2017-01-07 01:54:19 -05:00
} else if args.flag_check {
check::run(conn, &args.flag_sample_file_dir).unwrap();
} else if args.flag_ts {
run_ts(args.arg_ts).unwrap();
} else {
run(args, conn, &signal);
}
}
fn run_ts(timestamps: Vec<String>) -> Result<(), error::Error> {
for timestamp in &timestamps {
let t = recording::Time::parse(timestamp)?;
println!("{} == {}", t, t.0);
}
Ok(())
}
fn run(args: Args, conn: rusqlite::Connection, signal: &chan::Receiver<chan_signal::Signal>) {
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
let db = Arc::new(db::Database::new(conn).unwrap());
let dir = dir::SampleFileDir::new(&args.flag_sample_file_dir, db.clone()).unwrap();
info!("Database is loaded.");
// Start a streamer for each camera.
let shutdown = Arc::new(AtomicBool::new(false));
let mut streamers = Vec::new();
let syncer = if !args.flag_read_only {
let (syncer_channel, syncer_join) = dir::start_syncer(dir.clone()).unwrap();
let l = db.lock();
let cameras = l.cameras_by_id().len();
let env = streamer::Environment{
db: &db,
dir: &dir,
clocks: &clock::REAL,
opener: &*stream::FFMPEG,
shutdown: &shutdown,
};
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
for (i, (id, camera)) in l.cameras_by_id().iter().enumerate() {
let rotate_offset_sec = streamer::ROTATE_INTERVAL_SEC * i as i64 / cameras as i64;
let mut streamer = streamer::Streamer::new(&env, syncer_channel.clone(), *id, camera,
rotate_offset_sec,
streamer::ROTATE_INTERVAL_SEC);
Rust rewrite I should have submitted/pushed more incrementally but just played with it on my computer as I was learning the language. The new Rust version more or less matches the functionality of the current C++ version, although there are many caveats listed below. Upgrade notes: when moving from the C++ version, I recommend dropping and recreating the "recording_cover" index in SQLite3 to pick up the addition of the "video_sync_samples" column: $ sudo systemctl stop moonfire-nvr $ sudo -u moonfire-nvr sqlite3 /var/lib/moonfire-nvr/db/db sqlite> drop index recording_cover; sqlite3> create index ...rest of command as in schema.sql...; sqlite3> ^D Some known visible differences from the C++ version: * .mp4 generation queries SQLite3 differently. Before it would just get all video indexes in a single query. Now it leads with a query that should be satisfiable by the covering index (assuming the index has been recreated as noted above), then queries individual recording's indexes as needed to fill a LRU cache. I believe this is roughly similar speed for the initial hit (which generates the moov part of the file) and significantly faster when seeking. I would have done it a while ago with the C++ version but didn't want to track down a lru cache library. It was easier to find with Rust. * On startup, the Rust version cleans up old reserved files. This is as in the design; the C++ version was just missing this code. * The .html recording list output is a little different. It's in ascending order, with the most current segment shorten than an hour rather than the oldest. This is less ergonomic, but it was easy. I could fix it or just wait to obsolete it with some fancier JavaScript UI. * commandline argument parsing and logging have changed formats due to different underlying libraries. * The JSON output isn't quite right (matching the spec / C++ implementation) yet. Additional caveats: * I haven't done any proof-reading of prep.sh + install instructions. * There's a lot of code quality work to do: adding (back) comments and test coverage, developing a good Rust style. * The ffmpeg foreign function interface is particularly sketchy. I'd eventually like to switch to something based on autogenerated bindings. I'd also like to use pure Rust code where practical, but once I do on-NVR motion detection I'll need to existing C/C++ libraries for speed (H.264 decoding + OpenCL-based analysis).
2016-11-25 17:34:00 -05:00
let name = format!("stream-{}", streamer.short_name());
streamers.push(thread::Builder::new().name(name).spawn(move|| {
streamer.run();
}).expect("can't create thread"));
}
Some((syncer_channel, syncer_join))
} else { None };
// Start the web interface.
let server = Server::http(args.flag_http_addr.as_str()).unwrap();
let h = web::Handler::new(db.clone(), dir.clone());
let _guard = server.handle(h);
info!("Ready to serve HTTP requests");
// Wait for a signal and shut down.
chan_select! {
signal.recv() -> signal => info!("Received signal {:?}; shutting down streamers.", signal),
}
shutdown.store(true, Ordering::SeqCst);
for streamer in streamers.drain(..) {
streamer.join().unwrap();
}
if let Some((syncer_channel, syncer_join)) = syncer {
info!("Shutting down syncer.");
drop(syncer_channel);
syncer_join.join().unwrap();
}
info!("Exiting.");
// TODO: drain the logger.
std::process::exit(0);
}