moonfire-nvr/src/web.rs

476 lines
18 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/>.
extern crate hyper;
use core::borrow::Borrow;
use core::str::FromStr;
use db;
use dir::SampleFileDir;
use error::{Error, Result};
use 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
use hyper::{header,server,status};
use hyper::uri::RequestUri;
use mime;
use mp4;
use recording;
use serde_json;
use std::fmt;
use std::io::Write;
use std::ops::Range;
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
use std::sync::{Arc,MutexGuard};
use 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
use time;
use url::form_urlencoded;
use uuid::Uuid;
const BINARY_PREFIXES: &'static [&'static str] = &[" ", " Ki", " Mi", " Gi", " Ti", " Pi", " Ei"];
const DECIMAL_PREFIXES: &'static [&'static str] =&[" ", " k", " M", " G", " T", " P", " E"];
lazy_static! {
static ref JSON: mime::Mime = mime!(Application/Json);
static ref HTML: mime::Mime = mime!(Text/Html);
}
mod json { include!(concat!(env!("OUT_DIR"), "/serde_types.rs")); }
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
enum Path {
CamerasList, // "/" or "/cameras/"
Camera(Uuid), // "/cameras/<uuid>/"
CameraRecordings(Uuid), // "/cameras/<uuid>/recordings"
CameraViewMp4(Uuid), // "/cameras/<uuid>/view.mp4"
NotFound,
}
fn get_path_and_query(uri: &RequestUri) -> (&str, &str) {
match *uri {
RequestUri::AbsolutePath(ref both) => match both.find('?') {
Some(split) => (&both[..split], &both[split+1..]),
None => (both, ""),
},
RequestUri::AbsoluteUri(ref u) => (u.path(), u.query().unwrap_or("")),
_ => ("", ""),
}
}
fn decode_path(path: &str) -> Path {
if path == "/" {
return Path::CamerasList;
}
if !path.starts_with("/cameras/") {
return Path::NotFound;
}
let path = &path["/cameras/".len()..];
if path == "" {
return Path::CamerasList;
}
let slash = match path.find('/') {
None => { return Path::NotFound; },
Some(s) => s,
};
let (uuid, path) = path.split_at(slash);
// TODO(slamb): require uuid to be in canonical format.
let uuid = match Uuid::parse_str(uuid) {
Ok(u) => u,
Err(_) => { return Path::NotFound },
};
match path {
"/" => Path::Camera(uuid),
"/recordings" => Path::CameraRecordings(uuid),
"/view.mp4" => Path::CameraViewMp4(uuid),
_ => Path::NotFound,
}
}
fn is_json(req: &server::Request) -> bool {
if let Some(accept) = req.headers.get::<header::Accept>() {
return accept.len() == 1 && accept[0].item == *JSON &&
accept[0].quality == header::Quality(1000);
}
false
}
pub struct HtmlEscaped<'a>(&'a str);
impl<'a> fmt::Display for HtmlEscaped<'a> {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
let mut last_end = 0;
for (start, part) in self.0.match_indices(|c| c == '<' || c == '&') {
f.write_str(&self.0[last_end..start])?;
f.write_str(if part == "<" { "&lt;" } else { "&amp;" })?;
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
last_end = start + 1;
}
f.write_str(&self.0[last_end..])
}
}
pub struct Humanized(i64);
impl Humanized {
fn do_fmt(&self, base: f32, prefixes: &[&str], f: &mut fmt::Formatter) -> fmt::Result {
let mut n = self.0 as f32;
let mut i = 0;
loop {
if n < base || i >= prefixes.len() - 1 {
break;
}
n /= base;
i += 1;
}
write!(f, "{:.1}{}", n, prefixes[i])
}
}
impl fmt::Display for Humanized {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
self.do_fmt(1000., DECIMAL_PREFIXES, f)
}
}
impl fmt::Binary for Humanized {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
self.do_fmt(1024., BINARY_PREFIXES, f)
}
}
pub struct HumanizedTimestamp(Option<recording::Time>);
impl fmt::Display for HumanizedTimestamp {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
match self.0 {
None => f.write_str("n/a"),
Some(t) => {
let tm = time::at(time::Timespec{sec: t.unix_seconds(), nsec: 0});
write!(f, "{}",
tm.strftime("%a, %d %b %Y %H:%M:%S %Z").or_else(|_| Err(fmt::Error))?)
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
}
}
}
}
pub struct Handler {
db: Arc<db::Database>,
dir: Arc<SampleFileDir>,
}
impl Handler {
pub fn new(db: Arc<db::Database>, dir: Arc<SampleFileDir>) -> Self {
Handler{db: db, dir: dir}
}
fn not_found(&self, mut res: server::Response) -> Result<()> {
*res.status_mut() = status::StatusCode::NotFound;
res.send(b"not found")?;
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
Ok(())
}
fn list_cameras(&self, req: &server::Request, mut res: server::Response) -> Result<()> {
let json = is_json(req);
let buf = {
let db = self.db.lock();
if json {
serde_json::to_vec(&json::ListCameras{cameras: db.cameras_by_id()})?
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
} else {
self.list_cameras_html(db)?
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
}
};
res.headers_mut().set(header::ContentType(if json { JSON.clone() } else { HTML.clone() }));
res.send(&buf)?;
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
Ok(())
}
fn list_cameras_html(&self, db: MutexGuard<db::LockedDatabase>) -> Result<Vec<u8>> {
let mut buf = Vec::new();
buf.extend_from_slice(b"\
<!DOCTYPE html>\n\
<html>\n\
<head>\n\
<title>Camera list</title>\n\
<meta http-equiv=\"Content-Language\" content=\"en\">\n\
<style type=\"text/css\">\n\
.header { background-color: #ddd; }\n\
td { padding-right: 3em; }\n\
</style>\n\
</head>\n\
<body>\n\
<table>\n");
for row in db.cameras_by_id().values() {
write!(&mut buf, "\
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
<tr class=header><td colspan=2><a href=\"/cameras/{}/\">{}</a></td></tr>\n\
<tr><td>description</td><td>{}</td></tr>\n\
<tr><td>space</td><td>{:b}B / {:b}B ({:.1}%)</td></tr>\n\
<tr><td>uuid</td><td>{}</td></tr>\n\
<tr><td>oldest recording</td><td>{}</td></tr>\n\
<tr><td>newest recording</td><td>{}</td></tr>\n\
<tr><td>total duration</td><td>{}</td></tr>\n",
row.uuid, HtmlEscaped(&row.short_name), HtmlEscaped(&row.description),
Humanized(row.sample_file_bytes), Humanized(row.retain_bytes),
100. * row.sample_file_bytes as f32 / row.retain_bytes as f32,
row.uuid, HumanizedTimestamp(row.range.as_ref().map(|r| r.start)),
HumanizedTimestamp(row.range.as_ref().map(|r| r.end)),
row.duration)?;
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
}
Ok(buf)
}
fn camera(&self, uuid: Uuid, query: &str, req: &server::Request, mut res: server::Response)
-> Result<()> {
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 json = is_json(req);
let buf = {
let db = self.db.lock();
if json {
let camera = db.get_camera(uuid)
.ok_or_else(|| Error::new("no such camera".to_owned()))?;
serde_json::to_vec(&json::Camera::new(camera, true))?
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
} else {
self.camera_html(db, query, uuid)?
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
}
};
res.headers_mut().set(header::ContentType(if json { JSON.clone() } else { HTML.clone() }));
res.send(&buf)?;
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
Ok(())
}
fn camera_html(&self, db: MutexGuard<db::LockedDatabase>, query: &str,
uuid: Uuid) -> Result<Vec<u8>> {
let r = Handler::get_optional_range(query)?;
let camera = db.get_camera(uuid)
.ok_or_else(|| Error::new("no such camera".to_owned()))?;
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 mut buf = Vec::new();
write!(&mut buf, "\
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
<!DOCTYPE html>\n\
<html>\n\
<head>\n\
<title>{0} recordings</title>\n\
<meta http-equiv=\"Content-Language\" content=\"en\">\n\
<style type=\"text/css\">\n\
tr:not(:first-child):hover {{ background-color: #ddd; }}\n\
th, td {{ padding: 0.5ex 1.5em; text-align: right; }}\n\
</style>\n\
</head>\n\
<body>\n\
<h1>{0}</h1>\n\
<p>{1}</p>\n\
<table>\n\
<tr><th>start</th><th>end</th><th>resolution</th>\
<th>fps</th><th>size</th><th>bitrate</th>\
</tr>\n",
HtmlEscaped(&camera.short_name), HtmlEscaped(&camera.description))?;
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
// Rather than listing each 60-second recording, generate a HTML row for aggregated .mp4
// files of up to FORCE_SPLIT_DURATION each, provided there is no gap or change in video
// parameters between recordings.
static FORCE_SPLIT_DURATION: recording::Duration =
recording::Duration(60 * 60 * recording::TIME_UNITS_PER_SEC);
db.list_aggregated_recordings(camera.id, &r, FORCE_SPLIT_DURATION, |row| {
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 seconds = (row.range.end.0 - row.range.start.0) / recording::TIME_UNITS_PER_SEC;
write!(&mut buf, "\
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
<tr><td><a href=\"view.mp4?start_time_90k={}&end_time_90k={}\">{}</a></td>\
<td>{}</td><td>{}x{}</td><td>{:.0}</td><td>{:b}B</td><td>{}bps</td></tr>\n",
row.range.start.0, row.range.end.0,
HumanizedTimestamp(Some(row.range.start)),
HumanizedTimestamp(Some(row.range.end)), row.video_sample_entry.width,
row.video_sample_entry.height,
if seconds == 0 { 0. } else { row.video_samples as f32 / seconds as f32 },
Humanized(row.sample_file_bytes),
Humanized(if seconds == 0 { 0 } else { row.sample_file_bytes * 8 / seconds }))?;
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
Ok(())
})?;
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
buf.extend_from_slice(b"</table>\n</html>\n");
Ok(buf)
}
fn camera_recordings(&self, uuid: Uuid, query: &str, req: &server::Request,
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
mut res: server::Response) -> Result<()> {
let r = Handler::get_optional_range(query)?;
if !is_json(req) {
*res.status_mut() = status::StatusCode::NotAcceptable;
res.send(b"only available for JSON requests")?;
return Ok(());
}
let mut out = json::ListRecordings{recordings: Vec::new()};
{
let db = self.db.lock();
let camera = db.get_camera(uuid)
.ok_or_else(|| Error::new("no such camera".to_owned()))?;
db.list_aggregated_recordings(camera.id, &r, recording::Duration(i64::max_value()),
|row| {
out.recordings.push(json::Recording{
start_time_90k: row.range.start.0,
end_time_90k: row.range.end.0,
sample_file_bytes: row.sample_file_bytes,
video_samples: row.video_samples,
video_sample_entry_width: row.video_sample_entry.width,
video_sample_entry_height: row.video_sample_entry.height,
video_sample_entry_sha1: strutil::hex(&row.video_sample_entry.sha1),
});
Ok(())
})?;
}
let buf = serde_json::to_vec(&out)?;
res.headers_mut().set(header::ContentType(JSON.clone()));
res.send(&buf)?;
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
Ok(())
}
fn camera_view_mp4(&self, uuid: Uuid, query: &str, req: &server::Request,
res: server::Response) -> Result<()> {
let camera_id = {
let db = self.db.lock();
let camera = db.get_camera(uuid)
.ok_or_else(|| Error::new("no such camera".to_owned()))?;
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
camera.id
};
let mut start = None;
let mut end = None;
let mut include_ts = false;
for (key, value) in form_urlencoded::parse(query.as_bytes()) {
let (key, value) = (key.borrow(), value.borrow());
match key {
"start_time_90k" => start = Some(recording::Time(i64::from_str(value)?)),
"end_time_90k" => end = Some(recording::Time(i64::from_str(value)?)),
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
"ts" => { include_ts = value == "true"; },
_ => {},
}
};
let start = start.ok_or_else(|| Error::new("start_time_90k missing".to_owned()))?;
let end = end.ok_or_else(|| Error::new("end_time_90k missing".to_owned()))?;
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 desired_range = start .. end;
let mut builder = mp4::Mp4FileBuilder::new();
// There should be roughly ceil((end - start) / desired_recording_duration) recordings
// in the desired timespan if there are no gaps or overlap. Add a couple more to be safe:
// one for misalignment of the requested timespan with the rotate offset, another because
// rotation only happens at key frames.
let ceil_durations = ((end - start).0 + recording::DESIRED_RECORDING_DURATION - 1) /
recording::DESIRED_RECORDING_DURATION;
let est_records = (ceil_durations + 2) as usize;
let mut next_start = start;
builder.reserve(est_records);
{
let db = self.db.lock();
db.list_recordings(camera_id, &desired_range, |r| {
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
if builder.len() == 0 && r.start > next_start {
return Err(Error::new(format!("recording started late ({} vs requested {})",
r.start, start)));
} else if builder.len() != 0 && r.start != next_start {
return Err(Error::new(format!("gap/overlap in recording: {} to {} after row {}",
next_start, r.start, builder.len())));
}
next_start = r.start + recording::Duration(r.duration_90k as i64);
// TODO: check for inconsistent video sample entries.
let rel_start = if r.start < start {
(start - r.start).0 as i32
} else {
0
};
let rel_end = if r.start + recording::Duration(r.duration_90k as i64) > end {
(end - r.start).0 as i32
} else {
r.duration_90k
};
builder.append(&db, r, rel_start .. rel_end)?;
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
Ok(())
})?;
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
}
if next_start < end {
return Err(Error::new(format!(
"recording ends early: {}, not requested: {} after {} rows.",
next_start, end, builder.len())))
}
if builder.len() > est_records {
warn!("Estimated {} records for time [{}, {}); actually were {}",
est_records, start, end, builder.len());
} else {
debug!("Estimated {} records for time [{}, {}); actually were {}",
est_records, start, end, builder.len());
}
builder.include_timestamp_subtitle_track(include_ts);
let mp4 = builder.build(self.db.clone(), self.dir.clone())?;
http_entity::serve(&mp4, req, res)?;
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
Ok(())
}
/// Parses optional `start_time_90k` and `end_time_90k` query parameters, defaulting to the
/// full range of possible values.
fn get_optional_range(query: &str) -> Result<Range<recording::Time>> {
let mut start = i64::min_value();
let mut end = i64::max_value();
for (key, value) in form_urlencoded::parse(query.as_bytes()) {
let (key, value) = (key.borrow(), value.borrow());
match key {
"start_time_90k" => start = i64::from_str(value)?,
"end_time_90k" => end = i64::from_str(value)?,
_ => {},
}
};
Ok(recording::Time(start) .. recording::Time(end))
}
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
}
impl server::Handler for Handler {
fn handle(&self, req: server::Request, res: server::Response) {
let (path, query) = get_path_and_query(&req.uri);
let res = match decode_path(path) {
Path::CamerasList => self.list_cameras(&req, res),
Path::Camera(uuid) => self.camera(uuid, query, &req, res),
Path::CameraRecordings(uuid) => self.camera_recordings(uuid, query, &req, res),
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
Path::CameraViewMp4(uuid) => self.camera_view_mp4(uuid, query, &req, res),
Path::NotFound => self.not_found(res),
};
if let Err(ref e) = res {
warn!("Error handling request: {}", e);
}
}
}
#[cfg(test)]
mod tests {
use super::{HtmlEscaped, Humanized};
#[test]
fn test_humanize() {
assert_eq!("1.0 B", format!("{:b}B", Humanized(1)));
assert_eq!("1.0 EiB", format!("{:b}B", Humanized(1i64 << 60)));
assert_eq!("1.5 EiB", format!("{:b}B", Humanized((1i64 << 60) + (1i64 << 59))));
assert_eq!("8.0 EiB", format!("{:b}B", Humanized(i64::max_value())));
assert_eq!("1.0 Mbps", format!("{}bps", Humanized(1_000_000)));
}
#[test]
fn test_html_escaped() {
assert_eq!("", format!("{}", HtmlEscaped("")));
assert_eq!("no special chars", format!("{}", HtmlEscaped("no special chars")));
assert_eq!("a &lt;tag> &amp; text", format!("{}", HtmlEscaped("a <tag> & text")));
}
}