ENHANCED: Improved support for RDF import/output formats.

The following improvements are made:

  - HDT files can now be created based on TriG input.

  - HDT files can now be exported in the TriG format.

  - Command-line tools now correctly advertize all and only supported
    formats.

Implementation details:

  - `RDFNotation::TRIG' was added.

  - `RDFNotation::NQUADS' was added.  The old `RDFNotation::NQUAD' is
    left in place, and is marked deprecated.

The support for RDF serialization formats contained many bugs:

  - It was not possible to generate/export HDTs from/to TriG files,
    even though Serd parses/generates TriG.

  - Serialization formats `n3' and `rdfxml' were still advertized in
    command-line tools, even though using these results in an error:

        $ ./rdf2hdt -f n3 example.ttl example.hdt
        Catch exception load: Item not found
        ERROR: Item not found

  - The Windows-based command-line utility (`rdf2hdt/rdf2hdt.cpp') did
    not support the abbreviation formats (`nq', `nt', and `ttl').

For consistency, the Git ignore file is updated to exclude N-Quads,
TriG, and Turtle files in addition to N-Triples files.

Backwards compatibility is ensured by leaving all old `RDFNotation'
items in place (with the same iteger value).  This includes formats
that are not supported, i.e., `N3', `JSON', and `XML'.

Author: wouterbeek

.gitignore

@@ -54,8 +54,11 @@ stamp-h1
 
 *.hdt
 *.hdt.index
+*.nq
 *.nt
 *.rdf
+*.trig
+*.ttl
 *.a
 **/examples/*
 **/tests/*

libhdt/include/HDTEnums.hpp

@@ -108,37 +108,14 @@ TripleComponentOrder parseOrder(const char *str);
 const char *getOrderStr(TripleComponentOrder order);
 
 enum RDFNotation {
-	/**
-	 * XML notation
-	 *
-	 * @see <a href="http://en.wikipedia.org/wiki/RDF/XML">Wikipedia</a>
-	 */
-	XML,
-	/**
-	 * N-TRIPLES notation
-	 *
-	 * @see <a href="http://en.wikipedia.org/wiki/N-Triples">Wikipedia</a>
-	 */
+	XML, // No longer supported.
 	NTRIPLES,
-	/**
-	 * TURTLE notation
-	 *
-	 * @see <a href="http://en.wikipedia.org/wiki/Turtle_(syntax)">Wikipedia</a>
-	 */
 	TURTLE,
-	/**
-	 * Notation 3 notation
-	 *
-	 * @see <a href="http://en.wikipedia.org/wiki/Notation_3">Wikipedia</a>
-	 */
-	N3,
-	/**
-	 * NQUAD notation
-	 *
-	 * @see <a href="http://sw.deri.org/2008/07/n-quads/">Wikipedia</a>
-	 */
-	NQUAD,
-	JSON,
+	N3, // Not supported.
+	NQUAD, // Deprecated: use `NQUADS' instead.
+	JSON, // Not supported.
+	NQUADS,
+	TRIG,
 };
 
 enum ResultEstimationType {

libhdt/src/rdf/RDFParser.cpp

@@ -14,7 +14,11 @@ namespace hdt {
 
 RDFParserCallback *RDFParserCallback::getParserCallback(RDFNotation notation) {
 #ifdef HAVE_SERD
-    if(notation==TURTLE || notation==NTRIPLES  || notation==NQUAD) {
+    if(notation==NQUAD || // Deprecated: use `NQUADS' instead.
+       notation==NQUADS ||
+       notation==NTRIPLES ||
+       notation==TRIG ||
+       notation==TURTLE) {
         return new RDFParserSerd();
     }
 #else

libhdt/src/rdf/RDFParserSerd.cpp

@@ -152,14 +152,18 @@ RDFParserSerd::~RDFParserSerd() {
 
 SerdSyntax RDFParserSerd::getParserType(RDFNotation notation) {
 	switch(notation){
+	case NQUAD: // Deprecated: use `NQUADS' instead.
+		return SERD_NQUADS;
+	case NQUADS:
+		return SERD_NQUADS;
 	case NTRIPLES:
 		return SERD_NTRIPLES;
-	case NQUAD:
-		return SERD_NQUADS;
+	case TRIG:
+		return SERD_TRIG;
 	case TURTLE:
 		return SERD_TURTLE;
 	default:
-		throw ParseException("Serd parser only supports ntriples, nquads, and turtle.");
+		throw ParseException("Serd parser only supports N-Triples, N-Quads, TriG, and Turtle.");
 	}
 }
 

libhdt/src/rdf/RDFSerializerSerd.cpp

@@ -11,14 +11,18 @@ namespace hdt {
 
 static SerdSyntax getType(RDFNotation notation) {
 	switch (notation) {
+	case NQUAD: // Deprecated: use `NQUADS' instead.
+		return SERD_NQUADS;
+	case NQUADS:
+		return SERD_NQUADS;
 	case NTRIPLES:
 		return SERD_NTRIPLES;
-	case NQUAD:
-		return SERD_NQUADS;
+	case TRIG:
+		return SERD_TRIG;
 	case TURTLE:
 		return SERD_TURTLE;
 	default:
-		throw std::runtime_error("Serd seriaizer only supports ntriples, nquads, and turtle.");
+		throw std::runtime_error("Serd seriaizer only supports N-Quads, N-Triples, TriG, and Turtle.");
 	}
 }
 

libhdt/tools/rdf2hdt.cpp

@@ -51,7 +51,7 @@ void help() {
     cout << "\t-i\t\t\tAlso generate index to solve all triple patterns." << endl;
     cout << "\t-c\t<configfile>\tHDT Config options file" << endl;
     cout << "\t-o\t<options>\tHDT Additional options (option1=value1;option2=value2;...)" << endl;
-    cout << "\t-f\t<format>\tFormat of the RDF input (n3, ntriples or nt, nquads or nq, turtle or ttl)" << endl;
+    cout << "\t-f\t<format>\tFormat of the RDF input (nquads,nq,ntriples,nt,trig,turtle,ttl)" << endl;
     cout << "\t-B\t\"<base URI>\"\tBase URI of the dataset." << endl;
     cout << "\t-V\tPrints the HDT version number." << endl;
     cout << "\t-p\tPrints a progress indicator." << endl;
@@ -173,7 +173,7 @@ int main(int argc, char **argv) {
         if (rdfFormat == "" || rdfFormat == "gz")
         {
             rdfFormat = "nt";
-            vout << "No input format detected. Using default: NTRIPLES." << endl;
+            vout << "No input format detected: using N-Triples by default." << endl;
         }
     }
 
@@ -183,22 +183,22 @@ int main(int argc, char **argv) {
     transform (rdfFormat.begin (), rdfFormat.end (), rdfFormat.begin (), ::tolower);
 
     // Detect input format
-    if (rdfFormat == "n3")
-        notation = N3;
-    else if (rdfFormat == "ntriples" || rdfFormat == "nt")
+    if (rdfFormat=="nquads" || rdfFormat=="nq") {
+        notation = NQUADS;
+    } else if (rdfFormat== "ntriples" || rdfFormat=="nt") {
         notation = NTRIPLES;
-    else if (rdfFormat == "nquads" || rdfFormat == "nq")
-        notation = NQUAD;
-    else if (rdfFormat == "turtle" || rdfFormat == "ttl")
+    } else if (rdfFormat=="trig") {
+        notation = TRIG;
+    } else if (rdfFormat=="turtle" || rdfFormat=="ttl") {
         notation = TURTLE;
     // -f or file extension detected, but didn't match any valid format.
-    else {
-        cerr << "ERROR: Detected \"" << rdfFormat << "\" input format. Must be one of:" << endl
-             << "\t- n3" << endl
-             << "\t- ntriples or nt" << endl
-             << "\t- nquads or nq" << endl
-             << "\t- turtle or ttl" << endl;
-
+    } else {
+        cerr << "ERROR: Input format `" << rdfFormat << "' is not supported.\n"
+             << "Use either of the following:\n"
+             << "\t- `ntriples' or `nt' for N-Triples\n"
+             << "\t- `nquads' or `nq' for N-Quads\n"
+             << "\t- `turtle' or `ttl' for Turtle\n"
+             << "\t- `trig' for TriG" << endl;
         return 1;
     }
 

rdf2hdt/rdf2hdt.cpp

@@ -51,7 +51,7 @@ void help() {
     cout << "\t-i\t\tAlso generate index to solve all triple patterns." << endl;
     cout << "\t-c\t<configfile>\tHDT Config options file" << endl;
     cout << "\t-o\t<options>\tHDT Additional options (option1=value1;option2=value2;...)" << endl;
-    cout << "\t-f\t<format>\tFormat of the RDF input (ntriples, nquad, n3, turtle, rdfxml)" << endl;
+    cout << "\t-f\t<format>\tFormat of the RDF input (nquads,nq,ntriples,nt,trig,turtle,ttl)" << endl;
     cout << "\t-B\t\"<base URI>\"\tBase URI of the dataset." << endl;
     cout << "\t-V\tPrints the HDT version number." << endl;
     //cout << "\t-v\tVerbose output" << endl;
@@ -129,18 +129,21 @@ int main(int argc, char **argv) {
     }
 
     if(rdfFormat!="") {
-        if(rdfFormat=="ntriples") {
+        if(rdfFormat=="nquads" || rdfFormat=="nq") {
+            notation = NQUADS;
+        } else if(rdfFormat=="ntriples" || rdfFormat=="nt") {
             notation = NTRIPLES;
-        } else if(rdfFormat=="nquad") {
-            notation = NQUAD;
-        } else if(rdfFormat=="n3") {
-            notation = N3;
-        } else if(rdfFormat=="turtle") {
+        } else if(rdfFormat=="trig") {
+            notation = TRIG;
+        } else if(rdfFormat=="turtle" || rdfFormat=="ttl") {
             notation = TURTLE;
-        } else if(rdfFormat=="rdfxml") {
-            notation = XML;
         } else {
-            cout << "ERROR: The RDF input format must be one of: (ntriples, nquad, n3, turtle, rdfxml)" << endl;
+            cerr << "ERROR: Input format `" << rdfFormat << "' is not supported.\n"
+                 << "Use either of the following:\n"
+                 << "\t- `ntriples' or `nt' for N-Triples\n"
+                 << "\t- `nquads' or `nq' for N-Quads\n"
+                 << "\t- `turtle' or `ttl' for Turtle\n"
+                 << "\t- `trig' for TriG" << endl;
             help();
             return 1;
         }