[ruby/uri] Fix RFC3986 regexps

ruby/uri@8e38592241
ruby · Jun 13, 2023 · 6d734a8 · 6d734a8
1 parent 3924dba
commit 6d734a8
Show file tree

Hide file tree

Showing 2 changed files with 13 additions and 4 deletions.
diff --git a/lib/uri/rfc3986_parser.rb b/lib/uri/rfc3986_parser.rb
@@ -34,6 +34,7 @@ class RFC3986_Parser # :nodoc:
 
     SCHEME = %r[[A-Za-z][+\-.0-9A-Za-z]*+].source
     SEG = %r[(?:%\h\h|[!$&-.0-9:;=@A-Z_a-z~/])].source
+    SEG_NC = %r[(?:%\h\h|[!$&-.0-9;=@A-Z_a-z~])].source
     FRAGMENT = %r[(?:%\h\h|[!$&-.0-9:;=@A-Z_a-z~/?])*+].source
 
     RFC3986_URI = %r[\A
@@ -43,8 +44,8 @@ class RFC3986_Parser # :nodoc:
       (?<hier-part>//
         (?<authority>#{AUTHORITY})
         (?<path-abempty>(?:/\g<seg>*+)?)
-      | (?<path-absolute>/\g<seg>*+)
-      | (?<path-rootless>(?!=/)\g<seg>++)
+      | (?<path-absolute>/((?!/)\g<seg>++)?)
+      | (?<path-rootless>(?!/)\g<seg>++)
       | (?<path-empty>)
       )
       (?:\?(?<query>[^\#]*+))?
@@ -58,7 +59,7 @@ class RFC3986_Parser # :nodoc:
         (?<authority>#{AUTHORITY})
         (?<path-abempty>(?:/\g<seg>*+)?)
       | (?<path-absolute>/\g<seg>*+)
-      | (?<path-noscheme>(?!=[:/])\g<seg>++)
+      | (?<path-noscheme>#{SEG_NC}++(?:/\g<seg>*+)?)
       | (?<path-empty>)
       )
       (?:\?(?<query>[^#]*+))?
@@ -156,7 +157,7 @@ def default_regexp # :nodoc:
         USERINFO: %r[\A#{USERINFO}\z]o,
         HOST: %r[\A#{HOST}\z]o,
         ABS_PATH: %r[\A/#{SEG}*+\z]o,
-        REL_PATH: %r[\A(?!=/)#{SEG}++\z]o,
+        REL_PATH: %r[\A(?!/)#{SEG}++\z]o,
         QUERY: %r[\A(?:%\h\h|[!$&-.0-9:;=@A-Z_a-z~/?])*+\z],
         FRAGMENT: %r[\A#{FRAGMENT}\z]o,
         OPAQUE: %r[\A(?:[^/].*)?\z],

diff --git a/test/uri/test_parser.rb b/test/uri/test_parser.rb
@@ -78,5 +78,13 @@ def test_split
     assert_equal(["http", nil, "[0::0]", nil, nil, "", nil, nil, nil], URI.split("http://[0::0]"))
     assert_equal([nil, nil, "example.com", nil, nil, "", nil, nil, nil], URI.split("//example.com"))
     assert_equal([nil, nil, "[0::0]", nil, nil, "", nil, nil, nil], URI.split("//[0::0]"))
+
+    assert_equal(["a", nil, nil, nil, nil, "", nil, nil, nil], URI.split("a:"))
+    assert_raise(URI::InvalidURIError) do
+      URI.parse("::")
+    end
+    assert_raise(URI::InvalidURIError) do
+      URI.parse("foo@example:foo")
+    end
   end
 end