caddy/caddyfile/parse_test.go

747 lines
18 KiB
Go
Raw Normal View History

// 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 caddyfile
import (
"io/ioutil"
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
"os"
"path/filepath"
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
"strings"
"testing"
)
func TestAllTokens(t *testing.T) {
tests := []struct {
name string
input string
expected []string
}{
{
name: "not-empty",
input: "a b c\nd e",
expected: []string{"a", "b", "c", "d", "e"},
}, {
name: "empty",
input: "",
}, {
name: "newline",
input: "\n",
}, {
name: "space",
input: " ",
}, {
name: "tab and newline",
input: "\t\n",
},
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 _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
tokens, err := allTokens(strings.NewReader(tt.input))
if err != nil {
t.Fatalf("Expected no error, got %v", err)
}
if len(tokens) != len(tt.expected) {
t.Fatalf("Expected %d tokens, got %d", len(tt.expected), len(tokens))
}
for i, val := range tt.expected {
if tokens[i].Text != val {
t.Errorf("Token %d should be '%s' but was '%s'", i, val, tokens[i].Text)
}
}
})
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
}
}
func TestParseOneAndImport(t *testing.T) {
testParseOne := func(input string) (ServerBlock, error) {
p := testParser(input)
p.Next() // parseOne doesn't call Next() to start, so we must
err := p.parseOne()
return p.block, err
}
for i, test := range []struct {
input string
shouldErr bool
keys []string
tokens map[string]int // map of directive name to number of tokens expected
}{
{`localhost`, false, []string{
"localhost",
}, map[string]int{}},
{`localhost
dir1`, false, []string{
"localhost",
}, map[string]int{
"dir1": 1,
}},
{`localhost:1234
dir1 foo bar`, false, []string{
"localhost:1234",
}, map[string]int{
"dir1": 3,
}},
{`localhost {
dir1
}`, false, []string{
"localhost",
}, map[string]int{
"dir1": 1,
}},
{`localhost:1234 {
dir1 foo bar
dir2
}`, false, []string{
"localhost:1234",
}, map[string]int{
"dir1": 3,
"dir2": 1,
}},
{`http://localhost https://localhost
dir1 foo bar`, false, []string{
"http://localhost",
"https://localhost",
}, map[string]int{
"dir1": 3,
}},
{`http://localhost https://localhost {
dir1 foo bar
}`, false, []string{
"http://localhost",
"https://localhost",
}, map[string]int{
"dir1": 3,
}},
{`http://localhost, https://localhost {
dir1 foo bar
}`, false, []string{
"http://localhost",
"https://localhost",
}, map[string]int{
"dir1": 3,
}},
{`http://localhost, {
}`, true, []string{
"http://localhost",
}, map[string]int{}},
{`host1:80, http://host2.com
dir1 foo bar
dir2 baz`, false, []string{
"host1:80",
"http://host2.com",
}, map[string]int{
"dir1": 3,
"dir2": 2,
}},
{`http://host1.com,
http://host2.com,
https://host3.com`, false, []string{
"http://host1.com",
"http://host2.com",
"https://host3.com",
}, map[string]int{}},
{`http://host1.com:1234, https://host2.com
dir1 foo {
bar baz
}
dir2`, false, []string{
"http://host1.com:1234",
"https://host2.com",
}, map[string]int{
"dir1": 6,
"dir2": 1,
}},
{`127.0.0.1
dir1 {
bar baz
}
dir2 {
foo bar
}`, false, []string{
"127.0.0.1",
}, map[string]int{
"dir1": 5,
"dir2": 5,
}},
{`localhost
dir1 {
foo`, true, []string{
"localhost",
}, map[string]int{
"dir1": 3,
}},
{`localhost
dir1 {
}`, false, []string{
"localhost",
}, map[string]int{
"dir1": 3,
}},
{`localhost
dir1 {
} }`, true, []string{
"localhost",
}, map[string]int{
"dir1": 3,
}},
{`localhost
dir1 {
nested {
foo
}
}
dir2 foo bar`, false, []string{
"localhost",
}, map[string]int{
"dir1": 7,
"dir2": 3,
}},
{``, false, []string{}, map[string]int{}},
{`localhost
dir1 arg1
import testdata/import_test1.txt`, false, []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
"localhost",
}, map[string]int{
"dir1": 2,
"dir2": 3,
"dir3": 1,
}},
{`import testdata/import_test2.txt`, false, []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
"host1",
}, map[string]int{
"dir1": 1,
"dir2": 2,
}},
{`import testdata/import_test1.txt testdata/import_test2.txt`, true, []string{}, map[string]int{}},
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
{`import testdata/not_found.txt`, true, []string{}, map[string]int{}},
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
{`""`, false, []string{}, map[string]int{}},
{``, false, []string{}, map[string]int{}},
// test cases found by fuzzing!
{`import }{$"`, true, []string{}, map[string]int{}},
{`import /*/*.txt`, true, []string{}, map[string]int{}},
{`import /???/?*?o`, true, []string{}, map[string]int{}},
{`import /??`, true, []string{}, map[string]int{}},
{`import /[a-z]`, true, []string{}, map[string]int{}},
{`import {$}`, true, []string{}, map[string]int{}},
{`import {%}`, true, []string{}, map[string]int{}},
{`import {$$}`, true, []string{}, map[string]int{}},
{`import {%%}`, true, []string{}, map[string]int{}},
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
} {
result, err := testParseOne(test.input)
if test.shouldErr && err == nil {
t.Errorf("Test %d: Expected an error, but didn't get one", i)
}
if !test.shouldErr && err != nil {
t.Errorf("Test %d: Expected no error, but got: %v", i, err)
}
if len(result.Keys) != len(test.keys) {
t.Errorf("Test %d: Expected %d keys, got %d",
i, len(test.keys), len(result.Keys))
continue
}
for j, addr := range result.Keys {
if addr != test.keys[j] {
t.Errorf("Test %d, key %d: Expected '%s', but was '%s'",
i, j, test.keys[j], addr)
}
}
if len(result.Tokens) != len(test.tokens) {
t.Errorf("Test %d: Expected %d directives, had %d",
i, len(test.tokens), len(result.Tokens))
continue
}
for directive, tokens := range result.Tokens {
if len(tokens) != test.tokens[directive] {
t.Errorf("Test %d, directive '%s': Expected %d tokens, counted %d",
i, directive, test.tokens[directive], len(tokens))
continue
}
}
}
}
func TestRecursiveImport(t *testing.T) {
testParseOne := func(input string) (ServerBlock, error) {
p := testParser(input)
p.Next() // parseOne doesn't call Next() to start, so we must
err := p.parseOne()
return p.block, err
}
isExpected := func(got ServerBlock) bool {
if len(got.Keys) != 1 || got.Keys[0] != "localhost" {
t.Errorf("got keys unexpected: expect localhost, got %v", got.Keys)
return false
}
if len(got.Tokens) != 2 {
t.Errorf("got wrong number of tokens: expect 2, got %d", len(got.Tokens))
return false
}
if len(got.Tokens["dir1"]) != 1 || len(got.Tokens["dir2"]) != 2 {
t.Errorf("got unexpect tokens: %v", got.Tokens)
return false
}
return true
}
recursiveFile1, err := filepath.Abs("testdata/recursive_import_test1")
if err != nil {
t.Fatal(err)
}
recursiveFile2, err := filepath.Abs("testdata/recursive_import_test2")
if err != nil {
t.Fatal(err)
}
// test relative recursive import
err = ioutil.WriteFile(recursiveFile1, []byte(
`localhost
dir1
import recursive_import_test2`), 0644)
if err != nil {
t.Fatal(err)
}
defer os.Remove(recursiveFile1)
err = ioutil.WriteFile(recursiveFile2, []byte("dir2 1"), 0644)
if err != nil {
t.Fatal(err)
}
defer os.Remove(recursiveFile2)
// import absolute path
result, err := testParseOne("import " + recursiveFile1)
if err != nil {
t.Fatal(err)
}
if !isExpected(result) {
t.Error("absolute+relative import failed")
}
// import relative path
result, err = testParseOne("import testdata/recursive_import_test1")
if err != nil {
t.Fatal(err)
}
if !isExpected(result) {
t.Error("relative+relative import failed")
}
// test absolute recursive import
err = ioutil.WriteFile(recursiveFile1, []byte(
`localhost
dir1
import `+recursiveFile2), 0644)
if err != nil {
t.Fatal(err)
}
// import absolute path
result, err = testParseOne("import " + recursiveFile1)
if err != nil {
t.Fatal(err)
}
if !isExpected(result) {
t.Error("absolute+absolute import failed")
}
// import relative path
result, err = testParseOne("import testdata/recursive_import_test1")
if err != nil {
t.Fatal(err)
}
if !isExpected(result) {
t.Error("relative+absolute import failed")
}
}
func TestDirectiveImport(t *testing.T) {
testParseOne := func(input string) (ServerBlock, error) {
p := testParser(input)
p.Next() // parseOne doesn't call Next() to start, so we must
err := p.parseOne()
return p.block, err
}
isExpected := func(got ServerBlock) bool {
if len(got.Keys) != 1 || got.Keys[0] != "localhost" {
t.Errorf("got keys unexpected: expect localhost, got %v", got.Keys)
return false
}
if len(got.Tokens) != 2 {
t.Errorf("got wrong number of tokens: expect 2, got %d", len(got.Tokens))
return false
}
if len(got.Tokens["dir1"]) != 1 || len(got.Tokens["proxy"]) != 8 {
t.Errorf("got unexpect tokens: %v", got.Tokens)
return false
}
return true
}
directiveFile, err := filepath.Abs("testdata/directive_import_test")
if err != nil {
t.Fatal(err)
}
err = ioutil.WriteFile(directiveFile, []byte(`prop1 1
prop2 2`), 0644)
if err != nil {
t.Fatal(err)
}
defer os.Remove(directiveFile)
// import from existing file
result, err := testParseOne(`localhost
dir1
proxy {
import testdata/directive_import_test
transparent
}`)
if err != nil {
t.Fatal(err)
}
if !isExpected(result) {
t.Error("directive import failed")
}
// import from nonexistent file
_, err = testParseOne(`localhost
dir1
proxy {
import testdata/nonexistent_file
transparent
}`)
if err == nil {
t.Fatal("expected error when importing a nonexistent file")
}
}
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
func TestParseAll(t *testing.T) {
for i, test := range []struct {
input string
shouldErr bool
keys [][]string // keys per server block, in order
}{
{`localhost`, false, [][]string{
{"localhost"},
}},
{`localhost:1234`, false, [][]string{
{"localhost:1234"},
}},
{`localhost:1234 {
}
localhost:2015 {
}`, false, [][]string{
{"localhost:1234"},
{"localhost:2015"},
}},
{`localhost:1234, http://host2`, false, [][]string{
{"localhost:1234", "http://host2"},
}},
{`localhost:1234, http://host2,`, true, [][]string{}},
{`http://host1.com, http://host2.com {
}
https://host3.com, https://host4.com {
}`, false, [][]string{
{"http://host1.com", "http://host2.com"},
{"https://host3.com", "https://host4.com"},
}},
{`import testdata/import_glob*.txt`, false, [][]string{
{"glob0.host0"},
{"glob0.host1"},
{"glob1.host0"},
{"glob2.host0"},
}},
{`import notfound/*`, false, [][]string{}}, // glob needn't error with no matches
{`import notfound/file.conf`, true, [][]string{}}, // but a specific file should
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
} {
p := testParser(test.input)
blocks, err := p.parseAll()
if test.shouldErr && err == nil {
t.Errorf("Test %d: Expected an error, but didn't get one", i)
}
if !test.shouldErr && err != nil {
t.Errorf("Test %d: Expected no error, but got: %v", i, err)
}
if len(blocks) != len(test.keys) {
t.Errorf("Test %d: Expected %d server blocks, got %d",
i, len(test.keys), len(blocks))
continue
}
for j, block := range blocks {
if len(block.Keys) != len(test.keys[j]) {
t.Errorf("Test %d: Expected %d keys in block %d, got %d",
i, len(test.keys[j]), j, len(block.Keys))
continue
}
for k, addr := range block.Keys {
if addr != test.keys[j][k] {
t.Errorf("Test %d, block %d, key %d: Expected '%s', but got '%s'",
i, j, k, test.keys[j][k], addr)
}
}
}
}
}
func TestEnvironmentReplacement(t *testing.T) {
os.Setenv("PORT", "8080")
os.Setenv("ADDRESS", "servername.com")
os.Setenv("FOOBAR", "foobar")
os.Setenv("PARTIAL_DIR", "r1")
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
// basic test; unix-style env vars
p := testParser(`{$ADDRESS}`)
blocks, _ := p.parseAll()
if actual, expected := blocks[0].Keys[0], "servername.com"; expected != actual {
t.Errorf("Expected key to be '%s' but was '%s'", expected, actual)
}
// basic test; unix-style env vars
p = testParser(`di{$PARTIAL_DIR}`)
blocks, _ = p.parseAll()
if actual, expected := blocks[0].Keys[0], "dir1"; expected != actual {
t.Errorf("Expected key to be '%s' but was '%s'", expected, actual)
}
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
// multiple vars per token
p = testParser(`{$ADDRESS}:{$PORT}`)
blocks, _ = p.parseAll()
if actual, expected := blocks[0].Keys[0], "servername.com:8080"; expected != actual {
t.Errorf("Expected key to be '%s' but was '%s'", expected, actual)
}
// windows-style var and unix style in same token
p = testParser(`{%ADDRESS%}:{$PORT}`)
blocks, _ = p.parseAll()
if actual, expected := blocks[0].Keys[0], "servername.com:8080"; expected != actual {
t.Errorf("Expected key to be '%s' but was '%s'", expected, actual)
}
// reverse order
p = testParser(`{$ADDRESS}:{%PORT%}`)
blocks, _ = p.parseAll()
if actual, expected := blocks[0].Keys[0], "servername.com:8080"; expected != actual {
t.Errorf("Expected key to be '%s' but was '%s'", expected, actual)
}
// env var in server block body as argument
p = testParser(":{%PORT%}\ndir1 {$FOOBAR}")
blocks, _ = p.parseAll()
if actual, expected := blocks[0].Keys[0], ":8080"; expected != actual {
t.Errorf("Expected key to be '%s' but was '%s'", expected, actual)
}
if actual, expected := blocks[0].Tokens["dir1"][1].Text, "foobar"; expected != actual {
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
t.Errorf("Expected argument to be '%s' but was '%s'", expected, actual)
}
// combined windows env vars in argument
p = testParser(":{%PORT%}\ndir1 {%ADDRESS%}/{%FOOBAR%}")
blocks, _ = p.parseAll()
if actual, expected := blocks[0].Tokens["dir1"][1].Text, "servername.com/foobar"; expected != actual {
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
t.Errorf("Expected argument to be '%s' but was '%s'", expected, actual)
}
// malformed env var (windows)
p = testParser(":1234\ndir1 {%ADDRESS}")
blocks, _ = p.parseAll()
if actual, expected := blocks[0].Tokens["dir1"][1].Text, "{%ADDRESS}"; expected != actual {
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
t.Errorf("Expected host to be '%s' but was '%s'", expected, actual)
}
// malformed (non-existent) env var (unix)
p = testParser(`:{$PORT$}`)
blocks, _ = p.parseAll()
if actual, expected := blocks[0].Keys[0], ":"; expected != actual {
t.Errorf("Expected key to be '%s' but was '%s'", expected, actual)
}
// in quoted field
p = testParser(":1234\ndir1 \"Test {$FOOBAR} test\"")
blocks, _ = p.parseAll()
if actual, expected := blocks[0].Tokens["dir1"][1].Text, "Test foobar test"; expected != actual {
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
t.Errorf("Expected argument to be '%s' but was '%s'", expected, actual)
}
// after end token
p = testParser(":1234\nanswer \"{{ .Name }} {$FOOBAR}\"")
blocks, _ = p.parseAll()
if actual, expected := blocks[0].Tokens["answer"][1].Text, "{{ .Name }} foobar"; expected != actual {
t.Errorf("Expected argument to be '%s' but was '%s'", expected, actual)
}
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
}
func testParser(input string) parser {
buf := strings.NewReader(input)
p := parser{Dispenser: NewDispenser("Caddyfile", buf)}
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
return p
}
2017-11-01 11:56:24 +08:00
func TestSnippets(t *testing.T) {
p := testParser(`
(common) {
gzip foo
errors stderr
}
http://example.com {
import common
}
`)
2017-10-16 07:10:56 +08:00
blocks, err := p.parseAll()
if err != nil {
t.Fatal(err)
}
for _, b := range blocks {
t.Log(b.Keys)
t.Log(b.Tokens)
}
if len(blocks) != 1 {
t.Fatalf("Expect exactly one server block. Got %d.", len(blocks))
}
if actual, expected := blocks[0].Keys[0], "http://example.com"; expected != actual {
t.Errorf("Expected server name to be '%s' but was '%s'", expected, actual)
}
if len(blocks[0].Tokens) != 2 {
t.Fatalf("Server block should have tokens from import")
}
if actual, expected := blocks[0].Tokens["gzip"][0].Text, "gzip"; expected != actual {
t.Errorf("Expected argument to be '%s' but was '%s'", expected, actual)
}
if actual, expected := blocks[0].Tokens["errors"][1].Text, "stderr"; expected != actual {
t.Errorf("Expected argument to be '%s' but was '%s'", expected, actual)
}
}
func writeStringToTempFileOrDie(t *testing.T, str string) (pathToFile string) {
file, err := ioutil.TempFile("", t.Name())
if err != nil {
panic(err) // get a stack trace so we know where this was called from.
}
if _, err := file.WriteString(str); err != nil {
panic(err)
}
if err := file.Close(); err != nil {
panic(err)
}
return file.Name()
}
func TestImportedFilesIgnoreNonDirectiveImportTokens(t *testing.T) {
fileName := writeStringToTempFileOrDie(t, `
http://example.com {
# This isn't an import directive, it's just an arg with value 'import'
basicauth / import password
}
`)
// Parse the root file that imports the other one.
p := testParser(`import ` + fileName)
blocks, err := p.parseAll()
if err != nil {
t.Fatal(err)
}
for _, b := range blocks {
t.Log(b.Keys)
t.Log(b.Tokens)
}
auth := blocks[0].Tokens["basicauth"]
line := auth[0].Text + " " + auth[1].Text + " " + auth[2].Text + " " + auth[3].Text
if line != "basicauth / import password" {
// Previously, it would be changed to:
// basicauth / import /path/to/test/dir/password
// referencing a file that (probably) doesn't exist and changing the
// password!
t.Errorf("Expected basicauth tokens to be 'basicauth / import password' but got %#q", line)
}
}
func TestSnippetAcrossMultipleFiles(t *testing.T) {
// Make the derived Caddyfile that expects (common) to be defined.
fileName := writeStringToTempFileOrDie(t, `
http://example.com {
import common
}
`)
// Parse the root file that defines (common) and then imports the other one.
p := testParser(`
(common) {
gzip foo
}
import ` + fileName + `
`)
blocks, err := p.parseAll()
if err != nil {
t.Fatal(err)
}
for _, b := range blocks {
t.Log(b.Keys)
t.Log(b.Tokens)
}
if len(blocks) != 1 {
t.Fatalf("Expect exactly one server block. Got %d.", len(blocks))
}
if actual, expected := blocks[0].Keys[0], "http://example.com"; expected != actual {
t.Errorf("Expected server name to be '%s' but was '%s'", expected, actual)
}
if len(blocks[0].Tokens) != 1 {
t.Fatalf("Server block should have tokens from import")
}
if actual, expected := blocks[0].Tokens["gzip"][0].Text, "gzip"; expected != actual {
t.Errorf("Expected argument to be '%s' but was '%s'", expected, actual)
}
}