From 4e3c61ef4d63dd6c3c735e901ca798728e4e1fbd Mon Sep 17 00:00:00 2001
From: Tatsuhiro Tsujikawa <tatsuhiro.t@gmail.com>
Date: Sat, 19 Sep 2020 23:58:37 +0900
Subject: [PATCH] nghttpx: Add accesslog variables to record request path
 without query

This commit the following variables to construct request line without
including query component:

* $method
* $path
* $path_without_query
* $protocol_version
---
 gennghttpxfun.py    |  4 ++++
 src/shrpx.cc        |  8 ++++++++
 src/shrpx_config.cc | 24 ++++++++++++++++++++++++
 src/shrpx_log.cc    | 23 +++++++++++++++++++++++
 src/shrpx_log.h     |  4 ++++
 5 files changed, 63 insertions(+)

diff --git a/gennghttpxfun.py b/gennghttpxfun.py
index 0db75875..1be4d56c 100755
--- a/gennghttpxfun.py
+++ b/gennghttpxfun.py
@@ -205,6 +205,10 @@ LOGVARS = [
     "tls_client_serial",
     "backend_host",
     "backend_port",
+    "method",
+    "path",
+    "path_without_query",
+    "protocol_version",
 ]
 
 if __name__ == '__main__':
diff --git a/src/shrpx.cc b/src/shrpx.cc
index 4ff5e478..363d564c 100644
--- a/src/shrpx.cc
+++ b/src/shrpx.cc
@@ -2607,6 +2607,14 @@ Logging:
                 request.  "-" if backend host is not available.
               * $backend_port:  backend  port   used  to  fulfill  the
                 request.  "-" if backend host is not available.
+              * $method: HTTP method
+              * $path:  Request  path  including query.   For  CONNECT
+                request, authority is recorded.
+              * $path_without_query:  $path   up  to  the   first  '?'
+                character.    For   CONNECT  request,   authority   is
+                recorded.
+              * $protocol_version:   HTTP  version   (e.g.,  HTTP/1.1,
+                HTTP/2)
 
               The  variable  can  be  enclosed  by  "{"  and  "}"  for
               disambiguation (e.g., ${remote_addr}).
diff --git a/src/shrpx_config.cc b/src/shrpx_config.cc
index 769ad9b8..8f5cd507 100644
--- a/src/shrpx_config.cc
+++ b/src/shrpx_config.cc
@@ -387,6 +387,11 @@ LogFragmentType log_var_lookup_token(const char *name, size_t namelen) {
     break;
   case 4:
     switch (name[3]) {
+    case 'h':
+      if (util::strieq_l("pat", name, 3)) {
+        return LogFragmentType::PATH;
+      }
+      break;
     case 'n':
       if (util::strieq_l("alp", name, 3)) {
         return LogFragmentType::ALPN;
@@ -396,6 +401,11 @@ LogFragmentType log_var_lookup_token(const char *name, size_t namelen) {
     break;
   case 6:
     switch (name[5]) {
+    case 'd':
+      if (util::strieq_l("metho", name, 5)) {
+        return LogFragmentType::METHOD;
+      }
+      break;
     case 's':
       if (util::strieq_l("statu", name, 5)) {
         return LogFragmentType::STATUS;
@@ -502,6 +512,15 @@ LogFragmentType log_var_lookup_token(const char *name, size_t namelen) {
       break;
     }
     break;
+  case 16:
+    switch (name[15]) {
+    case 'n':
+      if (util::strieq_l("protocol_versio", name, 15)) {
+        return LogFragmentType::PROTOCOL_VERSION;
+      }
+      break;
+    }
+    break;
   case 17:
     switch (name[16]) {
     case 'l':
@@ -521,6 +540,11 @@ LogFragmentType log_var_lookup_token(const char *name, size_t namelen) {
         return LogFragmentType::TLS_SESSION_REUSED;
       }
       break;
+    case 'y':
+      if (util::strieq_l("path_without_quer", name, 17)) {
+        return LogFragmentType::PATH_WITHOUT_QUERY;
+      }
+      break;
     }
     break;
   case 22:
diff --git a/src/shrpx_log.cc b/src/shrpx_log.cc
index cc5d018b..3789c779 100644
--- a/src/shrpx_log.cc
+++ b/src/shrpx_log.cc
@@ -603,6 +603,11 @@ void upstream_accesslog(const std::vector<LogFragment> &lfv,
                                                  ? StringRef::from_lit("*")
                                                  : StringRef::from_lit("-")
                                            : req.path;
+  auto path_without_query =
+      req.method == HTTP_CONNECT
+          ? path
+          : StringRef{std::begin(path),
+                      std::find(std::begin(path), std::end(path), '?')};
 
   auto p = std::begin(buf);
   auto last = std::end(buf) - 2;
@@ -632,6 +637,24 @@ void upstream_accesslog(const std::vector<LogFragment> &lfv,
         std::tie(p, last) = copy(req.http_minor, p, last);
       }
       break;
+    case LogFragmentType::METHOD:
+      std::tie(p, last) = copy(method, p, last);
+      std::tie(p, last) = copy(' ', p, last);
+      break;
+    case LogFragmentType::PATH:
+      std::tie(p, last) = copy_escape(path, p, last);
+      break;
+    case LogFragmentType::PATH_WITHOUT_QUERY:
+      std::tie(p, last) = copy_escape(path_without_query, p, last);
+      break;
+    case LogFragmentType::PROTOCOL_VERSION:
+      std::tie(p, last) = copy_l("HTTP/", p, last);
+      std::tie(p, last) = copy(req.http_major, p, last);
+      if (req.http_major < 2) {
+        std::tie(p, last) = copy('.', p, last);
+        std::tie(p, last) = copy(req.http_minor, p, last);
+      }
+      break;
     case LogFragmentType::STATUS:
       std::tie(p, last) = copy(resp.http_status, p, last);
       break;
diff --git a/src/shrpx_log.h b/src/shrpx_log.h
index 7b0b914e..81035b2e 100644
--- a/src/shrpx_log.h
+++ b/src/shrpx_log.h
@@ -249,6 +249,10 @@ enum class LogFragmentType {
   TLS_CLIENT_SUBJECT_NAME,
   BACKEND_HOST,
   BACKEND_PORT,
+  METHOD,
+  PATH,
+  PATH_WITHOUT_QUERY,
+  PROTOCOL_VERSION,
 };
 
 struct LogFragment {