Fix incorrect parsing of strange quote escape patterns

6 files changed, 136 insertions, 5 deletions

diff --git a/.reuse/dep5 b/.reuse/dep5
index 53511c48..d7d4151f 100644
--- a/.reuse/dep5
+++ b/.reuse/dep5
@@ -14,7 +14,7 @@ Comment: Extra tests for serd (potential contributions to the W3C suites)
 License: BSD-3-Clause OR ISC
 
 Files: AUTHORS NEWS serd.ttl
-Copyright: 2011-2022 David Robillard <d@drobilla.net>
+Copyright: 2011-2023 David Robillard <d@drobilla.net>
 Comment: Contributed to the Commons as a representation of simple facts
 License: 0BSD OR ISC
 
diff --git a/NEWS b/NEWS
index b9a34beb..4d8fd5fa 100644
--- a/NEWS
+++ b/NEWS
@@ -9,6 +9,7 @@ serd (0.31.0) unstable; urgency=medium
   * Clean up code
   * Fix crash when trying to read chunks without starting
   * Fix hang when skipping an error at EOF when lax parsing
+  * Fix incorrect parsing of strange quote escape patterns
   * Gracefully handle bad characters in Turtle blank node syntax
   * Gracefully handle bad characters in Turtle datatype syntax
   * Improve serdi man page
@@ -16,7 +17,7 @@ serd (0.31.0) unstable; urgency=medium
   * Replace duplicated dox_to_sphinx script with sphinxygen dependency
   * Test header for warnings more strictly
 
- -- David Robillard <d@drobilla.net>  Tue, 07 Feb 2023 23:34:12 +0000
+ -- David Robillard <d@drobilla.net>  Wed, 01 Mar 2023 00:36:43 +0000
 
 serd (0.30.16) stable; urgency=medium
 
diff --git a/src/n3.c b/src/n3.c
index 71a2fc56..e14cb8ce 100644
--- a/src/n3.c
+++ b/src/n3.c
@@ -361,9 +361,15 @@ read_STRING_LITERAL_LONG(SerdReader* const    reader,
         skip_byte(reader, q3);
         break;
       }
-      *flags |= SERD_HAS_QUOTE;
-      push_byte(reader, ref, c);
-      st = read_character(reader, ref, flags, (uint8_t)q2);
+
+      if (q2 == '\\') {
+        push_byte(reader, ref, c);
+        st = read_string_escape(reader, ref, flags);
+      } else {
+        *flags |= SERD_HAS_QUOTE;
+        push_byte(reader, ref, c);
+        st = read_character(reader, ref, flags, (uint8_t)q2);
+      }
     } else if (c == EOF) {
       return r_err(reader, SERD_ERR_BAD_SYNTAX, "end of file in long string\n");
     } else {
diff --git a/test/good/manifest.ttl b/test/good/manifest.ttl
index 00009c4f..25359fdc 100644
--- a/test/good/manifest.ttl
+++ b/test/good/manifest.ttl
@@ -45,6 +45,7 @@
 		<#test-out-of-range-unicode>
 		<#test-prefix>
 		<#test-pretty>
+		<#test-quote-escapes>
 		<#test-rel>
 		<#test-semi-dot>
 		<#test-uri-escape>
@@ -281,6 +282,12 @@
 	mf:action <test-pretty.ttl> ;
 	mf:result <test-pretty.nt> .
 
+<#test-quote-escapes>
+	rdf:type rdft:TestTurtleEval ;
+	mf:name "test-quote-escapes" ;
+	mf:action <test-quote-escapes.ttl> ;
+	mf:result <test-quote-escapes.nt> .
+
 <#test-rel>
 	rdf:type rdft:TestTurtleEval ;
 	mf:name "test-rel" ;
diff --git a/test/good/test-quote-escapes.nt b/test/good/test-quote-escapes.nt
new file mode 100644
index 00000000..26c06a8d
--- /dev/null
+++ b/test/good/test-quote-escapes.nt
@@ -0,0 +1,52 @@
+<http://example.org/s> <http://example.org/p> "\"" .
+<http://example.org/s> <http://example.org/p> "\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"" .
+<http://example.org/s> <http://example.org/p> "\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"\"" .
+<http://example.org/s> <http://example.org/p> "single \" within" .
+<http://example.org/s> <http://example.org/p> "single \" within" .
+<http://example.org/s> <http://example.org/p> "double \"\" within" .
+<http://example.org/s> <http://example.org/p> "double \"\" within" .
+<http://example.org/s> <http://example.org/p> "double \"\" within" .
+<http://example.org/s> <http://example.org/p> "double \"\" within" .
+<http://example.org/s> <http://example.org/p> "triple \"\"\" within" .
+<http://example.org/s> <http://example.org/p> "triple \"\"\" within" .
+<http://example.org/s> <http://example.org/p> "triple \"\"\" within" .
+<http://example.org/s> <http://example.org/p> "triple \"\"\" within" .
+<http://example.org/s> <http://example.org/p> "triple \"\"\" within" .
+<http://example.org/s> <http://example.org/p> "triple \"\"\" within" .
+<http://example.org/s> <http://example.org/p> "triple \"\"\" within" .
+<http://example.org/s> <http://example.org/p> "\" start single" .
+<http://example.org/s> <http://example.org/p> "\" start single" .
+<http://example.org/s> <http://example.org/p> "\"\" start double" .
+<http://example.org/s> <http://example.org/p> "\"\" start double" .
+<http://example.org/s> <http://example.org/p> "\"\" start double" .
+<http://example.org/s> <http://example.org/p> "\"\" start double" .
+<http://example.org/s> <http://example.org/p> "\"\"\" start triple" .
+<http://example.org/s> <http://example.org/p> "\"\"\" start triple" .
+<http://example.org/s> <http://example.org/p> "\"\"\" start triple" .
+<http://example.org/s> <http://example.org/p> "\"\"\" start triple" .
+<http://example.org/s> <http://example.org/p> "\"\"\" start triple" .
+<http://example.org/s> <http://example.org/p> "\"\"\" start triple" .
+<http://example.org/s> <http://example.org/p> "\"\"\" start triple" .
+<http://example.org/s> <http://example.org/p> "end single \"" .
+<http://example.org/s> <http://example.org/p> "end double \"\"" .
+<http://example.org/s> <http://example.org/p> "end double \"\"" .
+<http://example.org/s> <http://example.org/p> "end triple \"\"\"" .
+<http://example.org/s> <http://example.org/p> "end triple \"\"\"" .
+<http://example.org/s> <http://example.org/p> "end triple \"\"\"" .
+<http://example.org/s> <http://example.org/p> "end triple \"\"\"" .
diff --git a/test/good/test-quote-escapes.ttl b/test/good/test-quote-escapes.ttl
new file mode 100644
index 00000000..ffe14807
--- /dev/null
+++ b/test/good/test-quote-escapes.ttl
@@ -0,0 +1,65 @@
+<http://example.org/s> <http://example.org/p> "\"" .
+<http://example.org/s> <http://example.org/p> "\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"" .
+<http://example.org/s> <http://example.org/p> "\"\"\"\"" .
+
+<http://example.org/s> <http://example.org/p> """\"""" .
+
+<http://example.org/s> <http://example.org/p> """"\"""" .
+<http://example.org/s> <http://example.org/p> """\"\"""" .
+
+<http://example.org/s> <http://example.org/p> """""\"""" .
+<http://example.org/s> <http://example.org/p> """\""\"""" .
+<http://example.org/s> <http://example.org/p> """\""\"""" .
+<http://example.org/s> <http://example.org/p> """\"\"\"""" .
+<http://example.org/s> <http://example.org/p> """"\"\"""" .
+
+<http://example.org/s> <http://example.org/p> """""\"\"""" .
+<http://example.org/s> <http://example.org/p> """"\""\"""" .
+<http://example.org/s> <http://example.org/p> """"\"\"\"""" .
+<http://example.org/s> <http://example.org/p> """\"""\"""" .
+<http://example.org/s> <http://example.org/p> """\""\"\"""" .
+<http://example.org/s> <http://example.org/p> """\"\""\"""" .
+<http://example.org/s> <http://example.org/p> """\"\"\"\"""" .
+
+<http://example.org/s> <http://example.org/p> """single " within""" .
+<http://example.org/s> <http://example.org/p> """single \" within""" .
+
+<http://example.org/s> <http://example.org/p> """double "" within""" .
+<http://example.org/s> <http://example.org/p> """double "\" within""" .
+<http://example.org/s> <http://example.org/p> """double \"" within""" .
+<http://example.org/s> <http://example.org/p> """double \"\" within""" .
+
+<http://example.org/s> <http://example.org/p> """triple ""\" within""" .
+<http://example.org/s> <http://example.org/p> """triple "\"" within""" .
+<http://example.org/s> <http://example.org/p> """triple "\"\" within""" .
+<http://example.org/s> <http://example.org/p> """triple \""" within""" .
+<http://example.org/s> <http://example.org/p> """triple \""\" within""" .
+<http://example.org/s> <http://example.org/p> """triple \"\"" within""" .
+<http://example.org/s> <http://example.org/p> """triple \"\"\" within""" .
+
+<http://example.org/s> <http://example.org/p> """" start single""" .
+<http://example.org/s> <http://example.org/p> """\" start single""" .
+
+<http://example.org/s> <http://example.org/p> """"" start double""" .
+<http://example.org/s> <http://example.org/p> """"\" start double""" .
+<http://example.org/s> <http://example.org/p> """\"" start double""" .
+<http://example.org/s> <http://example.org/p> """\"\" start double""" .
+
+<http://example.org/s> <http://example.org/p> """""\" start triple""" .
+<http://example.org/s> <http://example.org/p> """"\"" start triple""" .
+<http://example.org/s> <http://example.org/p> """"\"\" start triple""" .
+<http://example.org/s> <http://example.org/p> """\""" start triple""" .
+<http://example.org/s> <http://example.org/p> """\""\" start triple""" .
+<http://example.org/s> <http://example.org/p> """\"\"" start triple""" .
+<http://example.org/s> <http://example.org/p> """\"\"\" start triple""" .
+
+<http://example.org/s> <http://example.org/p> """end single \"""" .
+
+<http://example.org/s> <http://example.org/p> """end double "\"""" .
+<http://example.org/s> <http://example.org/p> """end double \"\"""" .
+
+<http://example.org/s> <http://example.org/p> """end triple ""\"""" .
+<http://example.org/s> <http://example.org/p> """end triple "\"\"""" .
+<http://example.org/s> <http://example.org/p> """end triple \""\"""" .
+<http://example.org/s> <http://example.org/p> """end triple \"\"\"""" .