2017-09-23 13:56:58 +08:00
|
|
|
// Copyright 2015 Light Code Labs, LLC
|
|
|
|
//
|
|
|
|
// Licensed under the Apache License, Version 2.0 (the "License");
|
|
|
|
// you may not use this file except in compliance with the License.
|
|
|
|
// You may obtain a copy of the License at
|
|
|
|
//
|
|
|
|
// http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
//
|
|
|
|
// Unless required by applicable law or agreed to in writing, software
|
|
|
|
// distributed under the License is distributed on an "AS IS" BASIS,
|
|
|
|
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
|
|
// See the License for the specific language governing permissions and
|
|
|
|
// limitations under the License.
|
|
|
|
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
package caddy
|
|
|
|
|
|
|
|
import (
|
|
|
|
"fmt"
|
2017-03-28 08:05:59 +08:00
|
|
|
"log"
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
"net"
|
|
|
|
"sort"
|
2018-02-03 14:52:53 +08:00
|
|
|
"sync"
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
|
2019-07-03 02:49:20 +08:00
|
|
|
"github.com/caddyserver/caddy/caddyfile"
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
)
|
|
|
|
|
|
|
|
// These are all the registered plugins.
|
|
|
|
var (
|
|
|
|
// serverTypes is a map of registered server types.
|
|
|
|
serverTypes = make(map[string]ServerType)
|
|
|
|
|
|
|
|
// plugins is a map of server type to map of plugin name to
|
|
|
|
// Plugin. These are the "general" plugins that may or may
|
|
|
|
// not be associated with a specific server type. If it's
|
|
|
|
// applicable to multiple server types or the server type is
|
|
|
|
// irrelevant, the key is empty string (""). But all plugins
|
|
|
|
// must have a name.
|
|
|
|
plugins = make(map[string]map[string]Plugin)
|
|
|
|
|
2017-03-28 08:05:59 +08:00
|
|
|
// eventHooks is a map of hook name to Hook. All hooks plugins
|
|
|
|
// must have a name.
|
2018-03-18 08:29:22 +08:00
|
|
|
eventHooks = &sync.Map{}
|
2017-03-28 08:05:59 +08:00
|
|
|
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
// parsingCallbacks maps server type to map of directive
|
|
|
|
// to list of callback functions. These aren't really
|
|
|
|
// plugins on their own, but are often registered from
|
|
|
|
// plugins.
|
2016-06-21 01:44:20 +08:00
|
|
|
parsingCallbacks = make(map[string]map[string][]ParsingCallback)
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
|
|
|
|
// caddyfileLoaders is the list of all Caddyfile loaders
|
|
|
|
// in registration order.
|
|
|
|
caddyfileLoaders []caddyfileLoader
|
|
|
|
)
|
|
|
|
|
|
|
|
// DescribePlugins returns a string describing the registered plugins.
|
|
|
|
func DescribePlugins() string {
|
2018-02-11 03:59:23 +08:00
|
|
|
pl := ListPlugins()
|
|
|
|
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
str := "Server types:\n"
|
2018-02-11 03:59:23 +08:00
|
|
|
for _, name := range pl["server_types"] {
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
str += " " + name + "\n"
|
|
|
|
}
|
|
|
|
|
|
|
|
str += "\nCaddyfile loaders:\n"
|
2018-02-11 03:59:23 +08:00
|
|
|
for _, name := range pl["caddyfile_loaders"] {
|
|
|
|
str += " " + name + "\n"
|
|
|
|
}
|
|
|
|
|
2018-02-17 14:20:08 +08:00
|
|
|
if len(pl["event_hooks"]) > 0 {
|
2018-02-11 03:59:23 +08:00
|
|
|
str += "\nEvent hook plugins:\n"
|
|
|
|
for _, name := range pl["event_hooks"] {
|
|
|
|
str += " hook." + name + "\n"
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-12-12 03:13:48 +08:00
|
|
|
if len(pl["clustering"]) > 0 {
|
|
|
|
str += "\nClustering plugins:\n"
|
|
|
|
for _, name := range pl["clustering"] {
|
|
|
|
str += " " + name + "\n"
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-02-11 03:59:23 +08:00
|
|
|
str += "\nOther plugins:\n"
|
|
|
|
for _, name := range pl["others"] {
|
|
|
|
str += " " + name + "\n"
|
|
|
|
}
|
|
|
|
|
|
|
|
return str
|
|
|
|
}
|
|
|
|
|
|
|
|
// ListPlugins makes a list of the registered plugins,
|
|
|
|
// keyed by plugin type.
|
|
|
|
func ListPlugins() map[string][]string {
|
|
|
|
p := make(map[string][]string)
|
|
|
|
|
|
|
|
// server type plugins
|
|
|
|
for name := range serverTypes {
|
|
|
|
p["server_types"] = append(p["server_types"], name)
|
|
|
|
}
|
|
|
|
|
|
|
|
// caddyfile loaders in registration order
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
for _, loader := range caddyfileLoaders {
|
2018-02-11 03:59:23 +08:00
|
|
|
p["caddyfile_loaders"] = append(p["caddyfile_loaders"], loader.name)
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
}
|
|
|
|
if defaultCaddyfileLoader.name != "" {
|
2018-02-11 03:59:23 +08:00
|
|
|
p["caddyfile_loaders"] = append(p["caddyfile_loaders"], defaultCaddyfileLoader.name)
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
}
|
|
|
|
|
2018-02-03 14:52:53 +08:00
|
|
|
// List the event hook plugins
|
|
|
|
eventHooks.Range(func(k, _ interface{}) bool {
|
2018-02-17 14:20:08 +08:00
|
|
|
p["event_hooks"] = append(p["event_hooks"], k.(string))
|
2018-02-03 14:52:53 +08:00
|
|
|
return true
|
|
|
|
})
|
2017-03-28 22:19:29 +08:00
|
|
|
|
2018-02-11 03:59:23 +08:00
|
|
|
// alphabetize the rest of the plugins
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
var others []string
|
|
|
|
for stype, stypePlugins := range plugins {
|
|
|
|
for name := range stypePlugins {
|
|
|
|
var s string
|
|
|
|
if stype != "" {
|
|
|
|
s = stype + "."
|
|
|
|
}
|
|
|
|
s += name
|
|
|
|
others = append(others, s)
|
|
|
|
}
|
|
|
|
}
|
2017-03-28 22:19:29 +08:00
|
|
|
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
sort.Strings(others)
|
|
|
|
for _, name := range others {
|
2018-02-11 03:59:23 +08:00
|
|
|
p["others"] = append(p["others"], name)
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
}
|
|
|
|
|
2018-02-11 03:59:23 +08:00
|
|
|
return p
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
// ValidDirectives returns the list of all directives that are
|
|
|
|
// recognized for the server type serverType. However, not all
|
|
|
|
// directives may be installed. This makes it possible to give
|
|
|
|
// more helpful error messages, like "did you mean ..." or
|
|
|
|
// "maybe you need to plug in ...".
|
|
|
|
func ValidDirectives(serverType string) []string {
|
|
|
|
stype, err := getServerType(serverType)
|
|
|
|
if err != nil {
|
|
|
|
return nil
|
|
|
|
}
|
2016-08-25 14:15:18 +08:00
|
|
|
return stype.Directives()
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
}
|
|
|
|
|
2016-08-19 08:19:45 +08:00
|
|
|
// ServerListener pairs a server to its listener and/or packetconn.
|
|
|
|
type ServerListener struct {
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
server Server
|
|
|
|
listener net.Listener
|
2016-07-19 04:24:09 +08:00
|
|
|
packet net.PacketConn
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
}
|
|
|
|
|
2016-08-19 08:19:45 +08:00
|
|
|
// LocalAddr returns the local network address of the packetconn. It returns
|
|
|
|
// nil when it is not set.
|
|
|
|
func (s ServerListener) LocalAddr() net.Addr {
|
|
|
|
if s.packet == nil {
|
|
|
|
return nil
|
|
|
|
}
|
|
|
|
return s.packet.LocalAddr()
|
|
|
|
}
|
|
|
|
|
|
|
|
// Addr returns the listener's network address. It returns nil when it is
|
|
|
|
// not set.
|
|
|
|
func (s ServerListener) Addr() net.Addr {
|
|
|
|
if s.listener == nil {
|
|
|
|
return nil
|
|
|
|
}
|
|
|
|
return s.listener.Addr()
|
|
|
|
}
|
|
|
|
|
2016-06-07 08:27:54 +08:00
|
|
|
// Context is a type which carries a server type through
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
// the load and setup phase; it maintains the state
|
|
|
|
// between loading the Caddyfile, then executing its
|
|
|
|
// directives, then making the servers for Caddy to
|
|
|
|
// manage. Typically, such state involves configuration
|
|
|
|
// structs, etc.
|
|
|
|
type Context interface {
|
2016-06-07 08:27:54 +08:00
|
|
|
// Called after the Caddyfile is parsed into server
|
|
|
|
// blocks but before the directives are executed,
|
|
|
|
// this method gives you an opportunity to inspect
|
|
|
|
// the server blocks and prepare for the execution
|
|
|
|
// of directives. Return the server blocks (which
|
|
|
|
// you may modify, if desired) and an error, if any.
|
|
|
|
// The first argument is the name or path to the
|
|
|
|
// configuration file (Caddyfile).
|
|
|
|
//
|
|
|
|
// This function can be a no-op and simply return its
|
|
|
|
// input if there is nothing to do here.
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
InspectServerBlocks(string, []caddyfile.ServerBlock) ([]caddyfile.ServerBlock, error)
|
2016-06-07 08:27:54 +08:00
|
|
|
|
|
|
|
// This is what Caddy calls to make server instances.
|
|
|
|
// By this time, all directives have been executed and,
|
|
|
|
// presumably, the context has enough state to produce
|
|
|
|
// server instances for Caddy to start.
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
MakeServers() ([]Server, error)
|
|
|
|
}
|
|
|
|
|
|
|
|
// RegisterServerType registers a server type srv by its
|
|
|
|
// name, typeName.
|
|
|
|
func RegisterServerType(typeName string, srv ServerType) {
|
|
|
|
if _, ok := serverTypes[typeName]; ok {
|
|
|
|
panic("server type already registered")
|
|
|
|
}
|
|
|
|
serverTypes[typeName] = srv
|
|
|
|
}
|
|
|
|
|
|
|
|
// ServerType contains information about a server type.
|
|
|
|
type ServerType struct {
|
2016-08-25 14:15:18 +08:00
|
|
|
// Function that returns the list of directives, in
|
|
|
|
// execution order, that are valid for this server
|
|
|
|
// type. Directives should be one word if possible
|
|
|
|
// and lower-cased.
|
|
|
|
Directives func() []string
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
|
|
|
|
// DefaultInput returns a default config input if none
|
2016-06-07 08:27:54 +08:00
|
|
|
// is otherwise loaded. This is optional, but highly
|
|
|
|
// recommended, otherwise a blank Caddyfile will be
|
|
|
|
// used.
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
DefaultInput func() Input
|
|
|
|
|
2016-06-07 08:27:54 +08:00
|
|
|
// The function that produces a new server type context.
|
|
|
|
// This will be called when a new Caddyfile is being
|
|
|
|
// loaded, parsed, and executed independently of any
|
|
|
|
// startup phases before this one. It's a way to keep
|
|
|
|
// each set of server instances separate and to reduce
|
|
|
|
// the amount of global state you need.
|
tls: Restructure and improve certificate management
- Expose the list of Caddy instances through caddy.Instances()
- Added arbitrary storage to caddy.Instance
- The cache of loaded certificates is no longer global; now scoped
per-instance, meaning upon reload (like SIGUSR1) the old cert cache
will be discarded entirely, whereas before, aggressively reloading
config that added and removed lots of sites would cause unnecessary
build-up in the cache over time.
- Key certificates in the cache by their SHA-256 hash instead of
by their names. This means certificates will not be duplicated in
memory (within each instance), making Caddy much more memory-efficient
for large-scale deployments with thousands of sites sharing certs.
- Perform name-to-certificate lookups scoped per caddytls.Config instead
of a single global lookup. This prevents certificates from stepping on
each other when they overlap in their names.
- Do not allow TLS configurations keyed by the same hostname to be
different; this now throws an error.
- Updated relevant tests, with a stark awareness that more tests are
needed.
- Change the NewContext function signature to include an *Instance.
- Strongly recommend (basically require) use of caddytls.NewConfig()
to create a new *caddytls.Config, to ensure pointers to the instance
certificate cache are initialized properly.
- Update the TLS-SNI challenge solver (even though TLS-SNI is disabled
currently on the CA side). Store temporary challenge cert in instance
cache, but do so directly by the ACME challenge name, not the hash.
Modified the getCertificate function to check the cache directly for
a name match if one isn't found otherwise. This will allow any
caddytls.Config to be able to help solve a TLS-SNI challenge, with one
extra side-effect that might actually be kind of interesting (and
useless): clients could send a certificate's hash as the SNI and
Caddy would be able to serve that certificate for the handshake.
- Do not attempt to match a "default" (random) certificate when SNI
is present but unrecognized; return no certificate so a TLS alert
happens instead.
- Store an Instance in the list of instances even while the instance
is still starting up (this allows access to the cert cache for
performing renewals at startup, etc). Will be removed from list again
if instance startup fails.
- Laid groundwork for ACMEv2 and Let's Encrypt wildcard support.
Server type plugins will need to be updated slightly to accommodate
minor adjustments to their API (like passing in an Instance). This
commit includes the changes for the HTTP server.
Certain Caddyfile configurations might error out with this change, if
they configured different TLS settings for the same hostname.
This change trades some complexity for other complexity, but ultimately
this new complexity is more correct and robust than earlier logic.
Fixes #1991
Fixes #1994
Fixes #1303
2018-02-04 15:58:27 +08:00
|
|
|
NewContext func(inst *Instance) Context
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
// Plugin is a type which holds information about a plugin.
|
|
|
|
type Plugin struct {
|
|
|
|
// ServerType is the type of server this plugin is for.
|
|
|
|
// Can be empty if not applicable, or if the plugin
|
|
|
|
// can associate with any server type.
|
|
|
|
ServerType string
|
|
|
|
|
|
|
|
// Action is the plugin's setup function, if associated
|
|
|
|
// with a directive in the Caddyfile.
|
|
|
|
Action SetupFunc
|
|
|
|
}
|
|
|
|
|
|
|
|
// RegisterPlugin plugs in plugin. All plugins should register
|
|
|
|
// themselves, even if they do not perform an action associated
|
|
|
|
// with a directive. It is important for the process to know
|
|
|
|
// which plugins are available.
|
2016-06-07 05:31:03 +08:00
|
|
|
//
|
|
|
|
// The plugin MUST have a name: lower case and one word.
|
|
|
|
// If this plugin has an action, it must be the name of
|
|
|
|
// the directive that invokes it. A name is always required
|
|
|
|
// and must be unique for the server type.
|
|
|
|
func RegisterPlugin(name string, plugin Plugin) {
|
|
|
|
if name == "" {
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
panic("plugin must have a name")
|
|
|
|
}
|
|
|
|
if _, ok := plugins[plugin.ServerType]; !ok {
|
|
|
|
plugins[plugin.ServerType] = make(map[string]Plugin)
|
|
|
|
}
|
2016-06-07 05:31:03 +08:00
|
|
|
if _, dup := plugins[plugin.ServerType][name]; dup {
|
|
|
|
panic("plugin named " + name + " already registered for server type " + plugin.ServerType)
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
}
|
2016-06-07 05:31:03 +08:00
|
|
|
plugins[plugin.ServerType][name] = plugin
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
}
|
|
|
|
|
2017-03-28 08:05:59 +08:00
|
|
|
// EventName represents the name of an event used with event hooks.
|
|
|
|
type EventName string
|
|
|
|
|
2017-10-02 01:25:30 +08:00
|
|
|
// Define names for the various events
|
2017-03-28 08:05:59 +08:00
|
|
|
const (
|
2017-10-02 10:36:23 +08:00
|
|
|
StartupEvent EventName = "startup"
|
2018-06-13 07:00:53 +08:00
|
|
|
ShutdownEvent = "shutdown"
|
|
|
|
CertRenewEvent = "certrenew"
|
|
|
|
InstanceStartupEvent = "instancestartup"
|
|
|
|
InstanceRestartEvent = "instancerestart"
|
2017-03-28 08:05:59 +08:00
|
|
|
)
|
|
|
|
|
|
|
|
// EventHook is a type which holds information about a startup hook plugin.
|
2017-06-03 21:28:16 +08:00
|
|
|
type EventHook func(eventType EventName, eventInfo interface{}) error
|
2017-03-28 08:05:59 +08:00
|
|
|
|
|
|
|
// RegisterEventHook plugs in hook. All the hooks should register themselves
|
|
|
|
// and they must have a name.
|
|
|
|
func RegisterEventHook(name string, hook EventHook) {
|
|
|
|
if name == "" {
|
|
|
|
panic("event hook must have a name")
|
|
|
|
}
|
2018-02-03 14:52:53 +08:00
|
|
|
_, dup := eventHooks.LoadOrStore(name, hook)
|
|
|
|
if dup {
|
2017-03-28 08:05:59 +08:00
|
|
|
panic("hook named " + name + " already registered")
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// EmitEvent executes the different hooks passing the EventType as an
|
|
|
|
// argument. This is a blocking function. Hook developers should
|
|
|
|
// use 'go' keyword if they don't want to block Caddy.
|
2017-06-03 21:28:16 +08:00
|
|
|
func EmitEvent(event EventName, info interface{}) {
|
2018-02-03 14:52:53 +08:00
|
|
|
eventHooks.Range(func(k, v interface{}) bool {
|
|
|
|
err := v.(EventHook)(event, info)
|
2017-03-28 08:05:59 +08:00
|
|
|
if err != nil {
|
2018-02-03 14:52:53 +08:00
|
|
|
log.Printf("error on '%s' hook: %v", k.(string), err)
|
2017-03-28 08:05:59 +08:00
|
|
|
}
|
2018-02-03 14:52:53 +08:00
|
|
|
return true
|
|
|
|
})
|
2017-03-28 08:05:59 +08:00
|
|
|
}
|
|
|
|
|
2018-03-18 08:29:22 +08:00
|
|
|
// cloneEventHooks return a clone of the event hooks *sync.Map
|
|
|
|
func cloneEventHooks() *sync.Map {
|
|
|
|
c := &sync.Map{}
|
|
|
|
eventHooks.Range(func(k, v interface{}) bool {
|
|
|
|
c.Store(k, v)
|
|
|
|
return true
|
|
|
|
})
|
|
|
|
return c
|
|
|
|
}
|
|
|
|
|
|
|
|
// purgeEventHooks purges all event hooks from the map
|
|
|
|
func purgeEventHooks() {
|
|
|
|
eventHooks.Range(func(k, _ interface{}) bool {
|
|
|
|
eventHooks.Delete(k)
|
|
|
|
return true
|
|
|
|
})
|
|
|
|
}
|
|
|
|
|
|
|
|
// restoreEventHooks restores eventHooks with a provided *sync.Map
|
|
|
|
func restoreEventHooks(m *sync.Map) {
|
|
|
|
// Purge old event hooks
|
|
|
|
purgeEventHooks()
|
|
|
|
|
|
|
|
// Restore event hooks
|
|
|
|
m.Range(func(k, v interface{}) bool {
|
|
|
|
eventHooks.Store(k, v)
|
|
|
|
return true
|
|
|
|
})
|
|
|
|
}
|
|
|
|
|
2016-06-21 01:44:20 +08:00
|
|
|
// ParsingCallback is a function that is called after
|
|
|
|
// a directive's setup functions have been executed
|
|
|
|
// for all the server blocks.
|
|
|
|
type ParsingCallback func(Context) error
|
|
|
|
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
// RegisterParsingCallback registers callback to be called after
|
|
|
|
// executing the directive afterDir for server type serverType.
|
2016-06-21 01:44:20 +08:00
|
|
|
func RegisterParsingCallback(serverType, afterDir string, callback ParsingCallback) {
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
if _, ok := parsingCallbacks[serverType]; !ok {
|
2016-06-21 01:44:20 +08:00
|
|
|
parsingCallbacks[serverType] = make(map[string][]ParsingCallback)
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
}
|
|
|
|
parsingCallbacks[serverType][afterDir] = append(parsingCallbacks[serverType][afterDir], callback)
|
|
|
|
}
|
|
|
|
|
|
|
|
// SetupFunc is used to set up a plugin, or in other words,
|
|
|
|
// execute a directive. It will be called once per key for
|
|
|
|
// each server block it appears in.
|
|
|
|
type SetupFunc func(c *Controller) error
|
|
|
|
|
|
|
|
// DirectiveAction gets the action for directive dir of
|
|
|
|
// server type serverType.
|
|
|
|
func DirectiveAction(serverType, dir string) (SetupFunc, error) {
|
|
|
|
if stypePlugins, ok := plugins[serverType]; ok {
|
|
|
|
if plugin, ok := stypePlugins[dir]; ok {
|
|
|
|
return plugin.Action, nil
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if genericPlugins, ok := plugins[""]; ok {
|
|
|
|
if plugin, ok := genericPlugins[dir]; ok {
|
|
|
|
return plugin.Action, nil
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return nil, fmt.Errorf("no action found for directive '%s' with server type '%s' (missing a plugin?)",
|
|
|
|
dir, serverType)
|
|
|
|
}
|
|
|
|
|
|
|
|
// Loader is a type that can load a Caddyfile.
|
|
|
|
// It is passed the name of the server type.
|
|
|
|
// It returns an error only if something went
|
|
|
|
// wrong, not simply if there is no Caddyfile
|
|
|
|
// for this loader to load.
|
|
|
|
//
|
|
|
|
// A Loader should only load the Caddyfile if
|
|
|
|
// a certain condition or requirement is met,
|
|
|
|
// as returning a non-nil Input value along with
|
|
|
|
// another Loader will result in an error.
|
|
|
|
// In other words, loading the Caddyfile must
|
|
|
|
// be deliberate & deterministic, not haphazard.
|
|
|
|
//
|
|
|
|
// The exception is the default Caddyfile loader,
|
|
|
|
// which will be called only if no other Caddyfile
|
|
|
|
// loaders return a non-nil Input. The default
|
|
|
|
// loader may always return an Input value.
|
|
|
|
type Loader interface {
|
2016-06-07 08:27:54 +08:00
|
|
|
Load(serverType string) (Input, error)
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
// LoaderFunc is a convenience type similar to http.HandlerFunc
|
|
|
|
// that allows you to use a plain function as a Load() method.
|
2016-06-07 08:27:54 +08:00
|
|
|
type LoaderFunc func(serverType string) (Input, error)
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
|
|
|
|
// Load loads a Caddyfile.
|
|
|
|
func (lf LoaderFunc) Load(serverType string) (Input, error) {
|
|
|
|
return lf(serverType)
|
|
|
|
}
|
|
|
|
|
|
|
|
// RegisterCaddyfileLoader registers loader named name.
|
|
|
|
func RegisterCaddyfileLoader(name string, loader Loader) {
|
|
|
|
caddyfileLoaders = append(caddyfileLoaders, caddyfileLoader{name: name, loader: loader})
|
|
|
|
}
|
|
|
|
|
|
|
|
// SetDefaultCaddyfileLoader registers loader by name
|
|
|
|
// as the default Caddyfile loader if no others produce
|
|
|
|
// a Caddyfile. If another Caddyfile loader has already
|
|
|
|
// been set as the default, this replaces it.
|
|
|
|
//
|
|
|
|
// Do not call RegisterCaddyfileLoader on the same
|
|
|
|
// loader; that would be redundant.
|
|
|
|
func SetDefaultCaddyfileLoader(name string, loader Loader) {
|
|
|
|
defaultCaddyfileLoader = caddyfileLoader{name: name, loader: loader}
|
|
|
|
}
|
|
|
|
|
|
|
|
// loadCaddyfileInput iterates the registered Caddyfile loaders
|
|
|
|
// and, if needed, calls the default loader, to load a Caddyfile.
|
|
|
|
// It is an error if any of the loaders return an error or if
|
|
|
|
// more than one loader returns a Caddyfile.
|
|
|
|
func loadCaddyfileInput(serverType string) (Input, error) {
|
|
|
|
var loadedBy string
|
|
|
|
var caddyfileToUse Input
|
|
|
|
for _, l := range caddyfileLoaders {
|
2016-08-18 07:02:01 +08:00
|
|
|
cdyfile, err := l.loader.Load(serverType)
|
|
|
|
if err != nil {
|
|
|
|
return nil, fmt.Errorf("loading Caddyfile via %s: %v", l.name, err)
|
|
|
|
}
|
|
|
|
if cdyfile != nil {
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
if caddyfileToUse != nil {
|
|
|
|
return nil, fmt.Errorf("Caddyfile loaded multiple times; first by %s, then by %s", loadedBy, l.name)
|
|
|
|
}
|
|
|
|
loaderUsed = l
|
|
|
|
caddyfileToUse = cdyfile
|
|
|
|
loadedBy = l.name
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if caddyfileToUse == nil && defaultCaddyfileLoader.loader != nil {
|
|
|
|
cdyfile, err := defaultCaddyfileLoader.loader.Load(serverType)
|
|
|
|
if err != nil {
|
|
|
|
return nil, err
|
|
|
|
}
|
|
|
|
if cdyfile != nil {
|
|
|
|
loaderUsed = defaultCaddyfileLoader
|
|
|
|
caddyfileToUse = cdyfile
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return caddyfileToUse, nil
|
|
|
|
}
|
|
|
|
|
2018-02-14 04:23:09 +08:00
|
|
|
// OnProcessExit is a list of functions to run when the process
|
|
|
|
// exits -- they are ONLY for cleanup and should not block,
|
|
|
|
// return errors, or do anything fancy. They will be run with
|
|
|
|
// every signal, even if "shutdown callbacks" are not executed.
|
|
|
|
// This variable must only be modified in the main goroutine
|
|
|
|
// from init() functions.
|
|
|
|
var OnProcessExit []func()
|
|
|
|
|
Rewrote Caddy from the ground up; initial commit of 0.9 branch
These changes span work from the last ~4 months in an effort to make
Caddy more extensible, reduce the coupling between its components, and
lay a more robust foundation of code going forward into 1.0. A bunch of
new features have been added, too, with even higher future potential.
The most significant design change is an overall inversion of
dependencies. Instead of the caddy package knowing about the server
and the notion of middleware and config, the caddy package exposes an
interface that other components plug into. This does introduce more
indirection when reading the code, but every piece is very modular and
pluggable. Even the HTTP server is pluggable.
The caddy package has been moved to the top level, and main has been
pushed into a subfolder called caddy. The actual logic of the main
file has been pushed even further into caddy/caddymain/run.go so that
custom builds of Caddy can be 'go get'able.
The HTTPS logic was surgically separated into two parts to divide the
TLS-specific code and the HTTPS-specific code. The caddytls package can
now be used by any type of server that needs TLS, not just HTTP. I also
added the ability to customize nearly every aspect of TLS at the site
level rather than all sites sharing the same TLS configuration. Not all
of this flexibility is exposed in the Caddyfile yet, but it may be in
the future. Caddy can also generate self-signed certificates in memory
for the convenience of a developer working on localhost who wants HTTPS.
And Caddy now supports the DNS challenge, assuming at least one DNS
provider is plugged in.
Dozens, if not hundreds, of other minor changes swept through the code
base as I literally started from an empty main function, copying over
functions or files as needed, then adjusting them to fit in the new
design. Most tests have been restored and adapted to the new API,
but more work is needed there.
A lot of what was "impossible" before is now possible, or can be made
possible with minimal disruption of the code. For example, it's fairly
easy to make plugins hook into another part of the code via callbacks.
Plugins can do more than just be directives; we now have plugins that
customize how the Caddyfile is loaded (useful when you need to get your
configuration from a remote store).
Site addresses no longer need be just a host and port. They can have a
path, allowing you to scope a configuration to a specific path. There is
no inheretance, however; each site configuration is distinct.
Thanks to amazing work by Lucas Clemente, this commit adds experimental
QUIC support. Turn it on using the -quic flag; your browser may have
to be configured to enable it.
Almost everything is here, but you will notice that most of the middle-
ware are missing. After those are transferred over, we'll be ready for
beta tests.
I'm very excited to get this out. Thanks for everyone's help and
patience these last few months. I hope you like it!!
2016-06-05 07:00:29 +08:00
|
|
|
// caddyfileLoader pairs the name of a loader to the loader.
|
|
|
|
type caddyfileLoader struct {
|
|
|
|
name string
|
|
|
|
loader Loader
|
|
|
|
}
|
|
|
|
|
|
|
|
var (
|
|
|
|
defaultCaddyfileLoader caddyfileLoader // the default loader if all else fail
|
|
|
|
loaderUsed caddyfileLoader // the loader that was used (relevant for reloads)
|
|
|
|
)
|