Final changes in 1-1 with Johannes

src/parser/RdfParser.cpp

@@ -823,10 +823,17 @@ bool RdfStreamParser<T>::resetStateAndRead(
 template <class T>
 void RdfStreamParser<T>::initialize(const string& filename) {
   this->clear();
-  // IMPORTANT: The current buffer must not end with a `.` (unless we are at the
-  // end of the file). The reason is that with a `.` at the end, we cannot
-  // decide whether we are in the middle of a `PN_LOCAL` (that continues in the
-  // next buffer) or at the end of a statement.
+  // Make sure that a block of data ends with a newline. This is important for
+  // two reasons:
+  //
+  // 1. A block of data must not end in the middle of a comment. Otherwise the
+  // remaining part of the comment, which is prepended to the next block, is
+  // not recognized as a comment.
+  //
+  // 2. A block of data must not end with a `.` (without subsequent newline).
+  // The reason is that with a `.` at the end, we cannot decide whether we are
+  // in the middle of a `PN_LOCAL` (that continues in the next buffer) or at the
+  // end of a statement.
   fileBuffer_ =
       std::make_unique<ParallelBufferWithEndRegex>(bufferSize_, "([\\r\\n]+)");
   fileBuffer_->open(filename);

src/parser/RdfParser.h

@@ -193,28 +193,22 @@ class TurtleParser : public RdfParserBase {
   // `TripleComponent` since it can hold any parsing result, not only objects.
   TripleComponent lastParseResult_;
 
-  // Map that maps prefix names to their IRI, initially empty.
-  ad_utility::HashMap<std::string, TripleComponent::Iri> prefixMap_{
-      {baseForRelativeIriKey_, TripleComponent::Iri{}},
-      {baseForAbsoluteIriKey_, TripleComponent::Iri{}}};
+  // Map that maps prefix names to their IRI. For our tests, it is important
+  // that without any BASE declaration, the two base prefixes are mapped to the
+  // empty IRI.
+  static const inline ad_utility::HashMap<std::string, TripleComponent::Iri>
+      prefixMapDefault_{{baseForRelativeIriKey_, TripleComponent::Iri{}},
+                        {baseForAbsoluteIriKey_, TripleComponent::Iri{}}};
+  ad_utility::HashMap<std::string, TripleComponent::Iri> prefixMap_ =
+      prefixMapDefault_;
 
   // Getters for the two base prefixes. Without BASE declaration, these will
   // both return the empty IRI.
-  //
-  // TODO<hannah> I would prefer to just call `prefixMap_.at(...)`, but then
-  // some of the tests fails because the keys are not in the map (despite the
-  // initialization above).
   const TripleComponent::Iri& baseForRelativeIri() {
-    // return prefixMap_.at(baseForRelativeIriKey_);
-    return prefixMap_
-        .try_emplace(baseForRelativeIriKey_, TripleComponent::Iri{})
-        .first->second;
+    return prefixMap_.at(baseForRelativeIriKey_);
   }
   const TripleComponent::Iri& baseForAbsoluteIri() {
-    // return prefixMap_.at(baseForAbsoluteIriKey_);
-    return prefixMap_
-        .try_emplace(baseForAbsoluteIriKey_, TripleComponent::Iri{})
-        .first->second;
+    return prefixMap_.at(baseForAbsoluteIriKey_);
   }
 
   // There are turtle constructs that reuse prefixes, subjects and predicates
@@ -248,7 +242,7 @@ class TurtleParser : public RdfParserBase {
     activePredicate_ = TripleComponent::Iri::fromIriref("<>");
     activePrefix_.clear();
 
-    prefixMap_.clear();
+    prefixMap_ = prefixMapDefault_;
 
     tok_.reset(nullptr, 0);
     triples_.clear();
@@ -544,8 +538,6 @@ class RdfStringParser : public Parser {
     this->tok_.reset(tmpToParse_.data(), tmpToParse_.size());
   }
 
-  void setPrefixMap(decltype(prefixMap_) m) { prefixMap_ = std::move(m); }
-
   const auto& getPrefixMap() const { return prefixMap_; }
 
   // __________________________________________________________

test/parser/LiteralOrIriTest.cpp

@@ -44,14 +44,8 @@ TEST(IriTest, getBaseIri) {
             "<http://purl.uniprot.org/>");
   EXPECT_EQ(getBaseIri("<http://purl.uniprot.org>", false),
             "<http://purl.uniprot.org/>");
-  // This leads to the following assertion failure and I don't understand why.
-  //
-  // C++ exception with description "Assertion `input.starts_with("<") &&
-  // input.ends_with(">")` failed. Please report this to the developers. In file
-  // "/local/data-ssd/qlever/qlever-code/src/parser/RdfEscaping.cpp " at line
-  // 232" thrown in the test body.
-  // EXPECT_EQ(getBaseIri("<http://purl.uniprot.org/", true),
-  //           "<http://purl.uniprot.org/>");
+  EXPECT_EQ(getBaseIri("<http://purl.uniprot.org/>", true),
+            "<http://purl.uniprot.org/>");
   EXPECT_EQ(getBaseIri("<http://purl.uniprot.org>", true),
             "<http://purl.uniprot.org/>");
   // IRI without scheme.