From 89a9066624f0131115d7efb4acc4a0a1fa130f74 Mon Sep 17 00:00:00 2001
From: Thibault Charbonnier <thibaultcha@me.com>
Date: Fri, 7 Jul 2017 12:39:56 -0700
Subject: [PATCH] feat(apis) allow for specifying user-specified regexes

APIs can now receive user-specified regexes in their `uris` property.
Those regexes will be treated as such by the router.

A URI is considered a regex when its character set does not respect the
reserved list from RFC 3986.

Implements: #677 (partially)
---
 kong/dao/schemas/apis.lua                     |  9 +++-
 spec/01-unit/007-entities_schemas_spec.lua    | 44 ++++++++++++-------
 .../05-proxy/01-router_spec.lua               | 31 +++++++++++--
 3 files changed, 63 insertions(+), 21 deletions(-)

diff --git a/kong/dao/schemas/apis.lua b/kong/dao/schemas/apis.lua
index 91c2629838c1..37cf857a6133 100644
--- a/kong/dao/schemas/apis.lua
+++ b/kong/dao/schemas/apis.lua
@@ -83,8 +83,13 @@ local function check_uri(uri)
     return false, "invalid"
 
   elseif not match(uri, "^/[%w%.%-%_~%/%%]*$") then
-    -- Check if characters are in RFC 3986 unreserved list, and % for percent encoding
-    return false, "must only contain alphanumeric and '., -, _, ~, /, %' characters"
+    -- URI contains characters outside of the reserved list of
+    -- RFC 3986: the value will be interpreted as a regex;
+    -- but is it a valid one?
+    local _, _, err = ngx.re.find("", uri, "aj")
+    if err then
+      return false, "invalid regex '" .. uri .. "' PCRE returned: " .. err
+    end
   end
 
   local esc = uri:gsub("%%%x%x", "___") -- drop all proper %-encodings
diff --git a/spec/01-unit/007-entities_schemas_spec.lua b/spec/01-unit/007-entities_schemas_spec.lua
index 89f5116dac64..1a1abb9498f4 100644
--- a/spec/01-unit/007-entities_schemas_spec.lua
+++ b/spec/01-unit/007-entities_schemas_spec.lua
@@ -270,6 +270,18 @@ describe("Entities Schemas", function()
         assert.is_true(ok)
       end)
 
+      it("accepts reserved characters from RFC 3986 (considered as a regex)", function()
+        local t = {
+          name = "httpbin",
+          upstream_url = "http://httpbin.org",
+          uris = { "/users/[a-z]+/" },
+        }
+
+        local ok, errors = validate_entity(t, api_schema)
+        assert.is_nil(errors)
+        assert.is_true(ok)
+      end)
+
       it("accepts properly %-encoded characters", function()
         local valids = {"/abcd%aa%10%ff%AA%FF"}
 
@@ -359,22 +371,6 @@ describe("Entities Schemas", function()
           end
         end)
 
-        it("rejects reserved characters from RFC 3986", function()
-          local invalids = { "/[a-z]{3}" }
-
-          for _, v in ipairs(invalids) do
-            local t = {
-              name = "mockbin",
-              upstream_url = "http://mockbin.com",
-              uris = { v },
-            }
-
-            local ok, errors = validate_entity(t, api_schema)
-            assert.is_false(ok)
-            assert.matches("must only contain alphanumeric and '., -, _, ~, /, %' characters", errors.uris, nil, true)
-          end
-        end)
-
         it("rejects bad %-encoded characters", function()
           local invalids = {
             "/some%2words",
@@ -431,6 +427,22 @@ describe("Entities Schemas", function()
             assert.matches("invalid", errors.uris, nil, true)
           end
         end)
+
+        it("rejects regex URIs that are invalid regexes", function()
+          local invalids = { [[/users/(foo/profile]] }
+
+          for _, v in ipairs(invalids) do
+            local t = {
+              name = "mockbin",
+              upstream_url = "http://mockbin.com",
+              uris = { v },
+            }
+
+            local ok, errors = validate_entity(t, api_schema)
+            assert.is_false(ok)
+            assert.matches("invalid regex", errors.uris, nil, true)
+          end
+        end)
       end)
     end)
 
diff --git a/spec/02-integration/05-proxy/01-router_spec.lua b/spec/02-integration/05-proxy/01-router_spec.lua
index baa20d41f231..8c5837e6a5ab 100644
--- a/spec/02-integration/05-proxy/01-router_spec.lua
+++ b/spec/02-integration/05-proxy/01-router_spec.lua
@@ -77,9 +77,15 @@ describe("Router", function()
         {
           name = "api-4",
           upstream_url = "http://httpbin.org/basic-auth",
-          uris = { "/user" },
+          uris = { "/private" },
           strip_uri = false,
-        }
+        },
+        {
+          name = "api-5",
+          upstream_url = "http://httpbin.org/anything",
+          uris = { [[/users/\d+/profile]] },
+          strip_uri = true,
+        },
       }
 
       assert(helpers.start_kong())
@@ -140,13 +146,32 @@ describe("Router", function()
     it("with strip_uri = false", function()
         local res = assert(client:send {
           method = "GET",
-          path = "/user/passwd",
+          path = "/private/passwd",
           headers = { ["kong-debug"] = 1 },
         })
 
         assert.res_status(401, res)
         assert.equal("api-4", res.headers["kong-api-name"])
     end)
+
+    it("[uri] with a regex", function()
+      local res = assert(client:send {
+        method  = "GET",
+        path    = "/users/foo/profile",
+        headers = { ["kong-debug"] = 1 },
+      })
+
+      assert.res_status(404, res)
+
+      res = assert(client:send {
+        method  = "GET",
+        path    = "/users/123/profile",
+        headers = { ["kong-debug"] = 1 },
+      })
+
+      assert.res_status(200, res)
+      assert.equal("api-5", res.headers["kong-api-name"])
+    end)
   end)
 
   describe("URI arguments (querystring)", function()