Improve URI resolution to cover most of the abnormal cases from RFC3986.

Improve test coverage of uri.c.


git-svn-id: http://svn.drobilla.net/serd/trunk@266 490d8e77-9747-427b-9fa3-0b8f29cee8a0

4 files changed, 70 insertions, 10 deletions

diff --git a/ChangeLog b/ChangeLog
index a3fab995..5c73cd2f 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -18,6 +18,7 @@ serd (UNRELEASED) unstable; urgency=low
   * Add serd_file_sink for the common case of writing to a FILE* stream.
   * Escape ASCII control characters in output (e.g. fix problems with string
     literals that start with a backspace)
+  * Improve URI resolution to cover most of the abnormal cases from RFC3986
 
  -- David Robillard <d@drobilla.net>  (UNRELEASED)
 
diff --git a/src/uri.c b/src/uri.c
index 24dd91d1..cd89d818 100644
--- a/src/uri.c
+++ b/src/uri.c
@@ -287,7 +287,7 @@ serd_uri_serialise(const SerdURI* uri, SerdSink sink, void* stream)
 		WRITE("//", 2);
 		WRITE(uri->authority.buf, uri->authority.len);
 	}
-	if (uri->path_base.len) {
+	if (uri->path_base.len || uri->path.len) {
 		if (!uri->path.buf && (uri->fragment.buf || uri->query.buf)) {
 			WRITE_COMPONENT("", uri->path_base, "");
 		} else if (uri->path.buf) {
@@ -307,24 +307,47 @@ serd_uri_serialise(const SerdURI* uri, SerdSink sink, void* stream)
 						begin += 2;  // Chop leading "./"
 						break;
 					case '.':
-						++up;
 						switch (begin[2]) {
+						case '\0':
+							++up;
+							begin += 2;  // Chop input ".."
+							done = true;
+							break;
 						case '/':
-							begin += 3;  // Chop lading "../"
+							++up;
+							begin += 3;  // Chop leading "../"
 							break;
 						default:
-							begin += 2;  // Chop leading ".."
+							done = true;
+							break;
 						}
 						break;
+					case '\0':
+						++begin;  // Chop input "." (and fall-through)
 					default:
-						++begin;  // Chop leading "."
+						done = true;
+						break;
 					}
 					break;
 				case '/':
-					if (begin[1] == '/') {
-						++begin;  // Replace leading "//" with "/"
-						break;
-					}  // else fall through
+					switch (begin[1]) {
+					case '.':
+						switch (begin[2]) {
+						case '/':
+							begin += 2;  // Leading "/./" => "/"
+							break;
+						case '.':
+							switch (begin[3]) {
+							case '/':
+								++up;
+								begin += 3;  // Leading "/../" => "/"
+							}
+							break;
+						default:
+							done = true;
+							break;
+						}
+					} // else fall through
 				default:
 					done = true;  // Finished chopping dot components
 				}
diff --git a/tests/test-uri.out b/tests/test-uri.out
index 1517d2e4..8c984f79 100644
--- a/tests/test-uri.out
+++ b/tests/test-uri.out
@@ -21,6 +21,20 @@
 <http://a/> <http://www.w3.org/2002/07/owl#sameAs> <http://a/> .
 <http://a/> <http://www.w3.org/2002/07/owl#sameAs> <http://a/> .
 <http://a/g> <http://www.w3.org/2002/07/owl#sameAs> <http://a/g> .
+<http://a/g> <http://www.w3.org/2002/07/owl#sameAs> <http://a/g> .
+<http://a/g> <http://www.w3.org/2002/07/owl#sameAs> <http://a/g> .
+<http://a/g> <http://www.w3.org/2002/07/owl#sameAs> <http://a/g> .
+<http://a/g> <http://www.w3.org/2002/07/owl#sameAs> <http://a/g> .
+<http://a/b/c/g.> <http://www.w3.org/2002/07/owl#sameAs> <http://a/b/c/g.> .
+<http://a/b/c/.g> <http://www.w3.org/2002/07/owl#sameAs> <http://a/b/c/.g> .
+<http://a/b/c/g..> <http://www.w3.org/2002/07/owl#sameAs> <http://a/b/c/g..> .
+<http://a/b/c/..g> <http://www.w3.org/2002/07/owl#sameAs> <http://a/b/c/..g> .
+<http://a/b/g> <http://www.w3.org/2002/07/owl#sameAs> <http://a/b/g> .
+<http://a/b/c/g?y/./x> <http://www.w3.org/2002/07/owl#sameAs> <http://a/b/c/g?y/./x> .
+<http://a/b/c/g?y/../x> <http://www.w3.org/2002/07/owl#sameAs> <http://a/b/c/g?y/../x> .
+<http://a/b/c/g#s/./x> <http://www.w3.org/2002/07/owl#sameAs> <http://a/b/c/g#s/./x> .
+<http://a/b/c/g#s/../x> <http://www.w3.org/2002/07/owl#sameAs> <http://a/b/c/g#s/../x> .
+<http://a/.g> <http://www.w3.org/2002/07/owl#sameAs> <http://a/.g> .
 <http://A?aquery> <http://www.w3.org/2002/07/owl#sameAs> <http://A?aquery> .
 <http://A#afragment> <http://www.w3.org/2002/07/owl#sameAs> <http://A#afragment> .
 <http://a/b/c/d;p?aquery> <http://www.w3.org/2002/07/owl#sameAs> <http://a/b/c/d;p?aquery> .
diff --git a/tests/test-uri.ttl b/tests/test-uri.ttl
index 22799360..cbc699ab 100644
--- a/tests/test-uri.ttl
+++ b/tests/test-uri.ttl
@@ -29,7 +29,29 @@
 <../../>  owl:sameAs <http://a/> .
 <../../g> owl:sameAs <http://a/g> .
 
+# 5.4.2. Abnormal Examples
+<../../../g>    owl:sameAs <http://a/g> .
+<../../../../g> owl:sameAs <http://a/g> .
+</./g>          owl:sameAs <http://a/g> .
+</../g>         owl:sameAs <http://a/g> .
+<g.>            owl:sameAs <http://a/b/c/g.> .
+<.g>            owl:sameAs <http://a/b/c/.g> .
+<g..>           owl:sameAs <http://a/b/c/g..> .
+<..g>           owl:sameAs <http://a/b/c/..g> .
+<./../g>        owl:sameAs <http://a/b/g> .
+# Serd only resolves the leading components
+#<./g/.>         owl:sameAs <http://a/b/c/g/> .
+#<g/./h>         owl:sameAs <http://a/b/c/g/h> .
+#<g/../h>        owl:sameAs <http://a/b/c/h> .
+#<g;x=1/./y>     owl:sameAs <http://a/b/c/g;x=1/y> .
+#<g;x=1/../y>    owl:sameAs <http://a/b/c/y> .
+<g?y/./x>       owl:sameAs <http://a/b/c/g?y/./x> .
+<g?y/../x>      owl:sameAs <http://a/b/c/g?y/../x> .
+<g#s/./x>       owl:sameAs <http://a/b/c/g#s/./x> .
+<g#s/../x>      owl:sameAs <http://a/b/c/g#s/../x> .
+      
 # Additional tests for Serd
+</.g>                owl:sameAs <http://a/.g> . 
 <http://A?aquery>    owl:sameAs <http://A?aquery> .
 <http://A#afragment> owl:sameAs <http://A#afragment> .
 <?aquery>            owl:sameAs <http://a/b/c/d;p?aquery> .
@@ -41,4 +63,4 @@
 
 @base <http://C#cfragment> .
 
-</bar>       owl:sameAs <http://C/bar> .
+</bar> owl:sameAs <http://C/bar> .