Split out HttpUtil::SpecForRequest() into a more generic function of net_util.h.

This was a TODO, since that function is useful outside of HTTP.

In the process, I uncovered some test cases in proxy_service that are passing in invalid URLs (by virtue of the extra DCHECK). This doesn't make much sense to me to support that, so I have changed them.

Review URL: http://codereview.chromium.org/160558

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@22359 0039d316-1c4b-4281-b951-d872f2087c98

3 files changed, 58 insertions, 0 deletions

diff --git a/net/base/net_util.cc b/net/base/net_util.cc
index 7afd4c5..492f7a8 100644
--- a/net/base/net_util.cc
+++ b/net/base/net_util.cc
@@ -1297,4 +1297,13 @@ std::wstring FormatUrl(const GURL& url,
   return url_string;
 }
 
+GURL SimplifyUrlForRequest(const GURL& url) {
+  DCHECK(url.is_valid());
+  GURL::Replacements replacements;
+  replacements.ClearUsername();
+  replacements.ClearPassword();
+  replacements.ClearRef();
+  return url.ReplaceComponents(replacements);
+}
+
 }  // namespace net
diff --git a/net/base/net_util.h b/net/base/net_util.h
index 4320e1c..0e247b8 100644
--- a/net/base/net_util.h
+++ b/net/base/net_util.h
@@ -225,6 +225,11 @@ inline std::wstring FormatUrl(const GURL& url, const std::wstring& languages) {
   return FormatUrl(url, languages, true, UnescapeRule::SPACES, NULL, NULL);
 }
 
+// Strip the portions of |url| that aren't core to the network request.
+//   - user name / password
+//   - reference section
+GURL SimplifyUrlForRequest(const GURL& url);
+
 }  // namespace net
 
 #endif  // NET_BASE_NET_UTIL_H__
diff --git a/net/base/net_util_unittest.cc b/net/base/net_util_unittest.cc
index feb4be7..726529e 100644
--- a/net/base/net_util_unittest.cc
+++ b/net/base/net_util_unittest.cc
@@ -1269,3 +1269,47 @@ TEST(NetUtilTest, FormatUrlParsed) {
   EXPECT_EQ(L"query", formatted.substr(parsed.query.begin, parsed.query.len));
   EXPECT_EQ(L"ref", formatted.substr(parsed.ref.begin, parsed.ref.len));
 }
+
+TEST(NetUtilTest, SimplifyUrlForRequest) {
+  struct {
+    const char* input_url;
+    const char* expected_simplified_url;
+  } tests[] = {
+    {
+      // Reference section should be stripped.
+      "http://www.google.com:78/foobar?query=1#hash",
+      "http://www.google.com:78/foobar?query=1",
+    },
+    {
+      // Reference section can itself contain #.
+      "http://192.168.0.1?query=1#hash#10#11#13#14",
+      "http://192.168.0.1?query=1",
+    },
+    { // Strip username/password.
+      "http://user:pass@google.com",
+      "http://google.com/",
+    },
+    { // Strip both the reference and the username/password.
+      "http://user:pass@google.com:80/sup?yo#X#X",
+      "http://google.com/sup?yo",
+    },
+    { // Try an HTTPS URL -- strip both the reference and the username/password.
+      "https://user:pass@google.com:80/sup?yo#X#X",
+      "https://google.com:80/sup?yo",
+    },
+    { // Try an FTP URL -- strip both the reference and the username/password.
+      "ftp://user:pass@google.com:80/sup?yo#X#X",
+      "ftp://google.com:80/sup?yo",
+    },
+    { // Try an standard URL with unknow scheme.
+      "foobar://user:pass@google.com:80/sup?yo#X#X",
+      "foobar://google.com:80/sup?yo",
+    },
+  };
+  for (size_t i = 0; i < ARRAYSIZE_UNSAFE(tests); ++i) {
+    SCOPED_TRACE(StringPrintf("Test[%d]: %s", i, tests[i].input_url));
+    GURL input_url(GURL(tests[i].input_url));
+    GURL expected_url(GURL(tests[i].expected_simplified_url));
+    EXPECT_EQ(expected_url, net::SimplifyUrlForRequest(input_url));
+  }
+}