url: file URL path normalization

Refs: https://github.com/whatwg/url/pull/544
Refs: https://github.com/web-platform-tests/wpt/pull/25716

PR-URL: https://github.com/nodejs/node/pull/35477
Fixes: https://github.com/nodejs/node/issues/35429
Reviewed-By: Guy Bedford <guybedford@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: David Carlier <devnexen@gmail.com>
Reviewed-By: Bradley Farias <bradley.meck@gmail.com>
Reviewed-By: Michaël Zasso <targos@protonmail.com>

1 files changed, 19 insertions, 33 deletions

diff --git a/src/node_url.cc b/src/node_url.cc
index b7aeb8b0062..466f3c689a5 100644
--- a/src/node_url.cc
+++ b/src/node_url.cc
@@ -1904,16 +1904,19 @@ void URL::Parse(const char* input,
               state = kFragment;
               break;
             default:
+              url->query.clear();
+              if (base->flags & URL_FLAGS_HAS_HOST) {
+                url->flags |= URL_FLAGS_HAS_HOST;
+                url->host = base->host;
+              }
+              if (base->flags & URL_FLAGS_HAS_PATH) {
+                url->flags |= URL_FLAGS_HAS_PATH;
+                url->path = base->path;
+              }
               if (!StartsWithWindowsDriveLetter(p, end)) {
-                if (base->flags & URL_FLAGS_HAS_HOST) {
-                  url->flags |= URL_FLAGS_HAS_HOST;
-                  url->host = base->host;
-                }
-                if (base->flags & URL_FLAGS_HAS_PATH) {
-                  url->flags |= URL_FLAGS_HAS_PATH;
-                  url->path = base->path;
-                }
                 ShortenUrlPath(url);
+              } else {
+                url->path.clear();
               }
               state = kPath;
               continue;
@@ -1927,20 +1930,13 @@ void URL::Parse(const char* input,
         if (ch == '/' || ch == '\\') {
           state = kFileHost;
         } else {
-          if (has_base &&
-              base->scheme == "file:" &&
-              !StartsWithWindowsDriveLetter(p, end)) {
-            if (IsNormalizedWindowsDriveLetter(base->path[0])) {
+          if (has_base && base->scheme == "file:") {
+            url->flags |= URL_FLAGS_HAS_HOST;
+            url->host = base->host;
+            if (!StartsWithWindowsDriveLetter(p, end) &&
+                IsNormalizedWindowsDriveLetter(base->path[0])) {
               url->flags |= URL_FLAGS_HAS_PATH;
               url->path.push_back(base->path[0]);
-            } else {
-              if (base->flags & URL_FLAGS_HAS_HOST) {
-                url->flags |= URL_FLAGS_HAS_HOST;
-                url->host = base->host;
-              } else {
-                url->flags &= ~URL_FLAGS_HAS_HOST;
-                url->host.clear();
-              }
             }
           }
           state = kPath;
@@ -2024,29 +2020,19 @@ void URL::Parse(const char* input,
                 url->path.empty() &&
                 buffer.size() == 2 &&
                 IsWindowsDriveLetter(buffer)) {
-              if ((url->flags & URL_FLAGS_HAS_HOST) &&
-                  !url->host.empty()) {
-                url->host.clear();
-                url->flags |= URL_FLAGS_HAS_HOST;
-              }
               buffer[1] = ':';
             }
             url->flags |= URL_FLAGS_HAS_PATH;
             url->path.emplace_back(std::move(buffer));
           }
           buffer.clear();
-          if (url->scheme == "file:" &&
-              (ch == kEOL ||
-               ch == '?' ||
-               ch == '#')) {
-            while (url->path.size() > 1 && url->path[0].empty()) {
-              url->path.erase(url->path.begin());
-            }
-          }
           if (ch == '?') {
             url->flags |= URL_FLAGS_HAS_QUERY;
+            url->query.clear();
             state = kQuery;
           } else if (ch == '#') {
+            url->flags |= URL_FLAGS_HAS_FRAGMENT;
+            url->fragment.clear();
             state = kFragment;
           }
         } else {