0
Fork 0
mirror of https://github.com/caddyserver/caddy.git synced 2024-12-23 22:27:38 -05:00
caddy/caddyhttp/httpserver/replacer.go

472 lines
12 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!!
2016-06-04 18:00:29 -05:00
package httpserver
2015-01-13 14:43:45 -05:00
import (
"bytes"
"io"
"io/ioutil"
"net"
2015-01-13 14:43:45 -05:00
"net/http"
"net/http/httputil"
"net/url"
"os"
"path"
2015-01-13 14:43:45 -05:00
"strconv"
"strings"
"time"
"github.com/mholt/caddy"
"github.com/mholt/caddy/caddytls"
2015-01-13 14:43:45 -05:00
)
// requestReplacer is a strings.Replacer which is used to
// encode literal \r and \n characters and keep everything
// on one line
var requestReplacer = strings.NewReplacer(
"\r", "\\r",
"\n", "\\n",
)
var now = time.Now
2015-05-03 14:43:50 -05:00
// Replacer is a type which can replace placeholder
2015-01-13 14:43:45 -05:00
// substrings in a string with actual values from a
// http.Request and ResponseRecorder. Always use
// NewReplacer to get one of these. Any placeholders
// made with Set() should overwrite existing values if
// the key is already used.
2015-05-03 14:43:50 -05:00
type Replacer interface {
Replace(string) string
2015-12-30 14:42:03 -05:00
Set(key, value string)
2015-05-03 14:43:50 -05:00
}
// replacer implements Replacer. customReplacements
// is used to store custom replacements created with
// Set() until the time of replacement, at which point
// they will be used to overwrite other replacements
// if there is a name conflict.
type replacer struct {
customReplacements map[string]string
emptyValue string
responseRecorder *ResponseRecorder
request *http.Request
requestBody *limitWriter
}
type limitWriter struct {
w bytes.Buffer
remain int
}
func newLimitWriter(max int) *limitWriter {
return &limitWriter{
w: bytes.Buffer{},
remain: max,
}
}
func (lw *limitWriter) Write(p []byte) (int, error) {
// skip if we are full
if lw.remain <= 0 {
return len(p), nil
}
if n := len(p); n > lw.remain {
p = p[:lw.remain]
}
n, err := lw.w.Write(p)
lw.remain -= n
return n, err
}
func (lw *limitWriter) String() string {
return lw.w.String()
}
2015-01-13 14:43:45 -05:00
// NewReplacer makes a new replacer based on r and rr which
// are used for request and response placeholders, respectively.
// Request placeholders are created immediately, whereas
// response placeholders are not created until Replace()
// is invoked. rr may be nil if it is not available.
// emptyValue should be the string that is used in place
// of empty string (can still be empty string).
func NewReplacer(r *http.Request, rr *ResponseRecorder, emptyValue string) Replacer {
repl := &replacer{
request: r,
responseRecorder: rr,
emptyValue: emptyValue,
}
// extract customReplacements from a request replacer when present.
if existing, ok := r.Context().Value(ReplacerCtxKey).(*replacer); ok {
repl.requestBody = existing.requestBody
repl.customReplacements = existing.customReplacements
} else {
// if there is no existing replacer, build one from scratch.
rb := newLimitWriter(MaxLogBodySize)
if r.Body != nil {
r.Body = struct {
io.Reader
io.Closer
}{io.TeeReader(r.Body, rb), io.Closer(r.Body)}
}
repl.requestBody = rb
repl.customReplacements = make(map[string]string)
2015-05-03 14:38:06 -05:00
}
return repl
2015-01-13 14:43:45 -05:00
}
2016-08-11 18:08:49 -05:00
func canLogRequest(r *http.Request) bool {
if r.Method == "POST" || r.Method == "PUT" {
for _, cType := range r.Header[headerContentType] {
// the cType could have charset and other info
if strings.Contains(cType, contentTypeJSON) || strings.Contains(cType, contentTypeXML) {
2016-08-11 18:08:49 -05:00
return true
}
}
}
2016-08-11 18:08:49 -05:00
return false
}
// unescapeBraces finds escaped braces in s and returns
// a string with those braces unescaped.
func unescapeBraces(s string) string {
s = strings.Replace(s, "\\{", "{", -1)
s = strings.Replace(s, "\\}", "}", -1)
return s
}
// Replace performs a replacement of values on s and returns
2015-01-13 14:43:45 -05:00
// the string with the replaced values.
func (r *replacer) Replace(s string) string {
// Do not attempt replacements if no placeholder is found.
if !strings.ContainsAny(s, "{}") {
return s
}
result := ""
Placeholders: // process each placeholder in sequence
for {
var idxStart, idxEnd int
idxOffset := 0
for { // find first unescaped opening brace
searchSpace := s[idxOffset:]
idxStart = strings.Index(searchSpace, "{")
if idxStart == -1 {
// no more placeholders
break Placeholders
}
if idxStart == 0 || searchSpace[idxStart-1] != '\\' {
// preceding character is not an escape
idxStart += idxOffset
break
}
// the brace we found was escaped
// search the rest of the string next
idxOffset += idxStart + 1
}
idxOffset = 0
for { // find first unescaped closing brace
searchSpace := s[idxStart+idxOffset:]
idxEnd = strings.Index(searchSpace, "}")
if idxEnd == -1 {
// unpaired placeholder
break Placeholders
}
if idxEnd == 0 || searchSpace[idxEnd-1] != '\\' {
// preceding character is not an escape
idxEnd += idxOffset + idxStart
break
}
// the brace we found was escaped
// search the rest of the string next
idxOffset += idxEnd + 1
}
2015-01-13 14:43:45 -05:00
// get a replacement for the unescaped placeholder
placeholder := unescapeBraces(s[idxStart : idxEnd+1])
replacement := r.getSubstitution(placeholder)
// append unescaped prefix + replacement
result += strings.TrimPrefix(unescapeBraces(s[:idxStart]), "\\") + replacement
// strip out scanned parts
s = s[idxEnd+1:]
2015-01-13 14:43:45 -05:00
}
// append unscanned parts
return result + unescapeBraces(s)
2015-01-13 14:43:45 -05:00
}
func roundDuration(d time.Duration) time.Duration {
if d >= time.Millisecond {
return round(d, time.Millisecond)
} else if d >= time.Microsecond {
return round(d, time.Microsecond)
}
return d
}
// round rounds d to the nearest r
func round(d, r time.Duration) time.Duration {
if r <= 0 {
return d
}
neg := d < 0
if neg {
d = -d
}
if m := d % r; m+m < r {
d = d - m
} else {
d = d + r - m
}
if neg {
return -d
}
return d
}
// getSubstitution retrieves value from corresponding key
func (r *replacer) getSubstitution(key string) string {
// search custom replacements first
if value, ok := r.customReplacements[key]; ok {
return value
}
// search request headers then
if key[1] == '>' {
want := key[2 : len(key)-1]
for key, values := range r.request.Header {
// Header placeholders (case-insensitive)
if strings.EqualFold(key, want) {
return strings.Join(values, ",")
}
}
}
// search response headers then
if r.responseRecorder != nil && key[1] == '<' {
want := key[2 : len(key)-1]
for key, values := range r.responseRecorder.Header() {
// Header placeholders (case-insensitive)
if strings.EqualFold(key, want) {
return strings.Join(values, ",")
}
}
}
// next check for cookies
if key[1] == '~' {
name := key[2 : len(key)-1]
if cookie, err := r.request.Cookie(name); err == nil {
return cookie.Value
}
}
// next check for query argument
if key[1] == '?' {
query := r.request.URL.Query()
name := key[2 : len(key)-1]
return query.Get(name)
}
// search default replacements in the end
switch key {
case "{method}":
return r.request.Method
case "{scheme}":
if r.request.TLS != nil {
return "https"
}
return "http"
case "{hostname}":
name, err := os.Hostname()
if err != nil {
return r.emptyValue
}
return name
case "{host}":
return r.request.Host
case "{hostonly}":
host, _, err := net.SplitHostPort(r.request.Host)
if err != nil {
return r.request.Host
}
return host
case "{path}":
u, _ := r.request.Context().Value(OriginalURLCtxKey).(url.URL)
return u.Path
case "{path_escaped}":
u, _ := r.request.Context().Value(OriginalURLCtxKey).(url.URL)
return url.QueryEscape(u.Path)
case "{request_id}":
reqid, _ := r.request.Context().Value(RequestIDCtxKey).(string)
return reqid
case "{rewrite_path}":
return r.request.URL.Path
case "{rewrite_path_escaped}":
return url.QueryEscape(r.request.URL.Path)
case "{query}":
u, _ := r.request.Context().Value(OriginalURLCtxKey).(url.URL)
return u.RawQuery
case "{query_escaped}":
u, _ := r.request.Context().Value(OriginalURLCtxKey).(url.URL)
return url.QueryEscape(u.RawQuery)
case "{fragment}":
u, _ := r.request.Context().Value(OriginalURLCtxKey).(url.URL)
return u.Fragment
case "{proto}":
return r.request.Proto
case "{remote}":
host, _, err := net.SplitHostPort(r.request.RemoteAddr)
if err != nil {
return r.request.RemoteAddr
}
return host
case "{port}":
_, port, err := net.SplitHostPort(r.request.RemoteAddr)
if err != nil {
return r.emptyValue
}
return port
case "{uri}":
u, _ := r.request.Context().Value(OriginalURLCtxKey).(url.URL)
return u.RequestURI()
case "{uri_escaped}":
u, _ := r.request.Context().Value(OriginalURLCtxKey).(url.URL)
return url.QueryEscape(u.RequestURI())
case "{rewrite_uri}":
return r.request.URL.RequestURI()
case "{rewrite_uri_escaped}":
return url.QueryEscape(r.request.URL.RequestURI())
case "{when}":
return now().Format(timeFormat)
case "{when_iso}":
return now().UTC().Format(timeFormatISOUTC)
2017-07-05 14:43:54 -05:00
case "{when_unix}":
return strconv.FormatInt(now().Unix(), 10)
case "{file}":
_, file := path.Split(r.request.URL.Path)
return file
case "{dir}":
dir, _ := path.Split(r.request.URL.Path)
return dir
case "{request}":
dump, err := httputil.DumpRequest(r.request, false)
if err != nil {
return r.emptyValue
}
return requestReplacer.Replace(string(dump))
case "{request_body}":
if !canLogRequest(r.request) {
return r.emptyValue
}
_, err := ioutil.ReadAll(r.request.Body)
if err != nil {
if err == ErrMaxBytesExceeded {
2016-11-03 19:25:49 -05:00
return r.emptyValue
}
}
return requestReplacer.Replace(r.requestBody.String())
case "{mitm}":
if val, ok := r.request.Context().Value(caddy.CtxKey("mitm")).(bool); ok {
if val {
return "likely"
}
2017-04-29 19:28:18 -05:00
return "unlikely"
}
return "unknown"
case "{status}":
if r.responseRecorder == nil {
return r.emptyValue
}
return strconv.Itoa(r.responseRecorder.status)
case "{size}":
if r.responseRecorder == nil {
return r.emptyValue
}
return strconv.Itoa(r.responseRecorder.size)
case "{latency}":
if r.responseRecorder == nil {
return r.emptyValue
}
return roundDuration(time.Since(r.responseRecorder.start)).String()
case "{latency_ms}":
if r.responseRecorder == nil {
return r.emptyValue
}
elapsedDuration := time.Since(r.responseRecorder.start)
return strconv.FormatInt(convertToMilliseconds(elapsedDuration), 10)
case "{tls_protocol}":
if r.request.TLS != nil {
for k, v := range caddytls.SupportedProtocols {
if v == r.request.TLS.Version {
return k
}
}
return "tls" // this should never happen, but guard in case
}
return r.emptyValue // because not using a secure channel
case "{tls_cipher}":
if r.request.TLS != nil {
for k, v := range caddytls.SupportedCiphersMap {
if v == r.request.TLS.CipherSuite {
return k
}
}
return "UNKNOWN" // this should never happen, but guard in case
}
return r.emptyValue
default:
// {labelN}
if strings.HasPrefix(key, "{label") {
nStr := key[6 : len(key)-1] // get the integer N in "{labelN}"
n, err := strconv.Atoi(nStr)
if err != nil || n < 1 {
return r.emptyValue
}
labels := strings.Split(r.request.Host, ".")
if n > len(labels) {
return r.emptyValue
}
return labels[n-1]
}
}
return r.emptyValue
}
// convertToMilliseconds returns the number of milliseconds in the given duration
func convertToMilliseconds(d time.Duration) int64 {
return d.Nanoseconds() / 1e6
}
// Set sets key to value in the r.customReplacements map.
func (r *replacer) Set(key, value string) {
r.customReplacements["{"+key+"}"] = value
2015-12-30 14:42:03 -05:00
}
2015-01-13 14:43:45 -05:00
const (
timeFormat = "02/Jan/2006:15:04:05 -0700"
timeFormatISOUTC = "2006-01-02T15:04:05Z" // ISO 8601 with timezone to be assumed as UTC
headerContentType = "Content-Type"
contentTypeJSON = "application/json"
contentTypeXML = "application/xml"
// MaxLogBodySize limits the size of logged request's body
MaxLogBodySize = 100 * 1024
2015-01-13 14:43:45 -05:00
)