Improve overall parsing

8 files changed, 370 insertions, 91 deletions

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 1106056..b86b74c 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -9,6 +9,7 @@ set(PGNP_INCLUDE_DIR ${CMAKE_CURRENT_BINARY_DIR}/includes) # For conveniance
 set(PGNP_INCLUDE_DIR ${PGNP_INCLUDE_DIR} PARENT_SCOPE) # To be used by other projects with add_subdirectory()
 file(MAKE_DIRECTORY ${PGNP_INCLUDE_DIR})
 configure_file(src/pgnp.hpp ${PGNP_INCLUDE_DIR} COPYONLY)
+include_directories(${PGNP_INCLUDE_DIR})
 
 # Unit tests
 enable_testing()
diff --git a/src/pgnp.cpp b/src/pgnp.cpp
index 391e6ee..bb9fd29 100644
--- a/src/pgnp.cpp
+++ b/src/pgnp.cpp
@@ -2,80 +2,274 @@
 #include "pgnp.hpp"
 #include <iostream>
 
-#define IS_BLANK(c) (c==' ' || c=='\n' || c=='\t')
-#define IS_EOF(loc) (loc>=pgn_content.size())
-#define EOF_CHECK(loc) {if(IS_EOF(loc)) throw UnexpectedEOF();}
+#define IS_BLANK(c) (c == ' ' || c == '\n' || c == '\t')
+#define IS_DIGIT(c)                                                            \
+  (c == '0' || c == '1' || c == '2' || c == '3' || c == '4' || c == '5' ||     \
+   c == '6' || c == '7' || c == '8' || c == '9')
+#define IS_EOF(loc) (loc >= pgn_content.size())
+#define EOF_CHECK(loc)                                                         \
+  {                                                                            \
+    if (IS_EOF(loc))                                                           \
+      throw UnexpectedEOF();                                                   \
+  }
 
 namespace pgnp {
 
-  void PGN::FromFile(std::string filepath){
-    std::ifstream file(filepath);
-    
-    std::string content((std::istreambuf_iterator<char>(file)),
-                    std::istreambuf_iterator<char>());
-    FromString(content);
-  }
-
-
-  void PGN::FromString(std::string pgn_content){
-    this->pgn_content=pgn_content;
-    int loc=0;
-    while(!IS_EOF(loc)) {
-      char c=pgn_content[loc];
-      if(!IS_BLANK(c)){
-        switch (c) {
-          case '[':
-            loc=ParseNextTag(loc);
-            break;
-        }
+HalfMove::HalfMove() : isBlack(false), MainLine(NULL) {}
+
+HalfMove::~HalfMove() {
+  for (auto *move : variations) {
+    delete move;
+  }
+}
+
+void HalfMove::NestedDump(HalfMove *m, int indent) {
+  for (int i = 0; i < indent; i++) {
+    std::cout << "   ";
+  }
+  std::cout << "  "
+            << " Move=" << m->move << " Count=" << m->count << " Comment=\""
+            << m->comment << "\""
+            << " IsBlack=" << m->isBlack
+            << " Variations=" << m->variations.size() << std::endl;
+
+  for (auto *var : m->variations) {
+    NestedDump(var, indent + 1);
+  }
+
+  if (m->MainLine != NULL) {
+    NestedDump(m->MainLine, indent);
+  }
+}
+
+void HalfMove::Dump() { NestedDump(this, 0); }
+
+int HalfMove::GetLength() {
+  int length = 0;
+  HalfMove *m = this;
+  while (m != NULL) {
+    length++;
+    m = m->MainLine;
+  }
+  return length;
+}
+
+PGN::~PGN() {
+  if (moves != NULL)
+    delete moves;
+}
+
+void PGN::FromFile(std::string filepath) {
+  std::ifstream file(filepath);
+
+  std::string content((std::istreambuf_iterator<char>(file)),
+                      std::istreambuf_iterator<char>());
+  FromString(content);
+}
+
+void PGN::FromString(std::string pgn_content) {
+  this->pgn_content = pgn_content;
+  moves = NULL;
+  int loc = 0;
+  while (!IS_EOF(loc)) {
+    char c = pgn_content[loc];
+    if (!IS_BLANK(c)) {
+      if (c == '[') {
+        loc = ParseNextTag(loc);
+      } else if (IS_DIGIT(c)) {
+        moves = new HalfMove();
+        loc = ParseLine(loc, moves);
+        break;
       }
-      loc++;
     }
+    loc++;
+  }
+}
 
-    /*for (auto const& [key, val] : tags){
-      std::cout << key <<"="<<val<<std::endl;
-    }*/
+void PGN::STRCheck() {
+  int i = 0;
+  // Locate Event tag
+  while (i < tagkeys.size()) {
+    if (tagkeys[i] == "Event") {
+      break;
+    }
+    i++;
   }
 
-  int PGN::ParseNextTag(int start_loc){
-    // Parse key
-    std::string key;
-    int keyloc=start_loc+1;
-    EOF_CHECK(keyloc);
-    char c=pgn_content[keyloc];
-    while(!IS_BLANK(c)){
-      key+=c;
-      keyloc++;
-      EOF_CHECK(keyloc);
-      c=pgn_content[keyloc];
+  // Check tags
+  if (i + 6 < tagkeys.size()) {
+    bool valid = (tagkeys[i] == "Event") && (tagkeys[i + 1] == "Site") &&
+                 (tagkeys[i + 2] == "Date") && (tagkeys[i + 3] == "Round") &&
+                 (tagkeys[i + 4] == "White") && (tagkeys[i + 5] == "Black") &&
+                 (tagkeys[i + 6] == "Result");
+    if (!valid) {
+      throw STRCheckFailed();
     }
+  } else {
+    throw STRCheckFailed();
+  }
+}
 
-    // Parse value
-    std::string value;
-    int valueloc=NextNonBlank(keyloc)+1;
-    EOF_CHECK(keyloc);
-    c=pgn_content[valueloc];
-    while(c!='"' or IS_EOF(valueloc)){
-      value+=c;
-      valueloc++;
-      EOF_CHECK(keyloc);
-      c=pgn_content[valueloc];
+bool PGN::HasTag(std::string key) {
+  auto tags = GetTagList();
+  return (std::find(tags.begin(), tags.end(), key) != tags.end());
+}
+
+int PGN::ParseLine(int loc, HalfMove *hm) {
+  // Goto next char
+  loc = NextNonBlank(loc);
+  EOF_CHECK(loc);
+  char c = pgn_content[loc];
+
+  // Check if we reach score entry (* or 1-0 or 0-1 or 1/2-1/2)
+  if (!IS_EOF(loc + 1)) {
+    char nc = pgn_content[loc + 1]; // Next c
+    if ((IS_DIGIT(c) && nc == '-') or (IS_DIGIT(c) && nc == '/')) {
+      return (loc);
     }
+  }
 
-    // Add tag
-    tags[key]=value;
+  // Parse (move number
+  if (IS_DIGIT(c)) {
+    std::string move_nb;
+    while (IS_DIGIT(c)) {
+      move_nb += c;
+      loc++;
+      c = pgn_content[loc];
+      EOF_CHECK(loc);
+    }
+    hm->count = std::stoi(move_nb);
+    loc++;
+    EOF_CHECK(loc);
+    if (pgn_content[loc] == '.') {
+      hm->isBlack = true;
+      loc += 2; // Skip two dots
+      EOF_CHECK(loc);
+    }
+  } else {
+    hm->isBlack = true;
+  }
+
+  // Parse the HalfMove
+  loc = NextNonBlank(loc);
+  EOF_CHECK(loc);
+  c = pgn_content[loc];
+  std::string move;
+  while (!IS_BLANK(c) && c != ')') {
+    move += c;
+    loc++;
+    c = pgn_content[loc];
+    EOF_CHECK(loc);
+  }
+  hm->move = move;
 
- 
-    return(valueloc+1); // +1 For the last char of the tag which is ']'
+  // Skip end of variation
+  if (c == ')') {
+    loc++;
+    return (loc);
   }
 
-  int PGN::NextNonBlank(int loc){
-    char c=pgn_content[loc];
-    while(IS_BLANK(c)){
+  // Check for comment
+  loc = NextNonBlank(loc);
+  if (!IS_EOF(loc) && pgn_content[loc] == '{') {
+    loc++; // Skip '{'
+    c = pgn_content[loc];
+    while (c != '}') {
+      hm->comment += c;
       loc++;
-      c=pgn_content[loc];
+      EOF_CHECK(loc);
+      c = pgn_content[loc];
     }
-    return(loc);
+    loc++; // Skip '}'
+  }
+
+  // Check for variations
+  loc = NextNonBlank(loc);
+  while (!IS_EOF(loc) && pgn_content[loc] == '(') {
+    loc++; // Skip '('
+    HalfMove *var = new HalfMove;
+    loc = ParseLine(loc, var);
+    hm->variations.push_back(var);
+    loc++; // Skip ')'
+  }
+
+  // Parse next HalfMove
+  loc = NextNonBlank(loc);
+  if (!IS_EOF(loc)) {
+    HalfMove *next_hm = new HalfMove;
+    next_hm->count = hm->count;
+    loc = ParseLine(loc, next_hm);
+    // Check if move parsed successfuly
+    if (next_hm->move.size() > 0) {
+      hm->MainLine = next_hm;
+    } else {
+      delete next_hm;
+    }
+  }
+
+  return (loc);
+}
+
+int PGN::ParseNextTag(int start_loc) {
+  // Parse key
+  std::string key;
+  int keyloc = start_loc + 1;
+  EOF_CHECK(keyloc);
+  char c = pgn_content[keyloc];
+  while (!IS_BLANK(c)) {
+    key += c;
+    keyloc++;
+    EOF_CHECK(keyloc);
+    c = pgn_content[keyloc];
+  }
+
+  // Parse value
+  std::string value;
+  int valueloc = NextNonBlank(keyloc) + 1;
+  EOF_CHECK(keyloc);
+  c = pgn_content[valueloc];
+  while (c != '"' or IS_EOF(valueloc)) {
+    value += c;
+    valueloc++;
+    EOF_CHECK(keyloc);
+    c = pgn_content[valueloc];
+  }
+
+  // Add tag
+  tags[key] = value;
+  tagkeys.push_back(key);
+
+  // TODO: Check that caracters if a ]
+  return (valueloc + 1); // +1 For the last char of the tag which is ']'
+}
+
+HalfMove *PGN::GetMoves() { return (moves); }
+
+std::vector<std::string> PGN::GetTagList() { return tagkeys; }
+
+std::string PGN::GetTagValue(std::string key) { return tags[key]; }
+
+void PGN::Dump() {
+  std::cout << "---------- PGN DUMP ----------" << std::endl;
+  std::cout << "Tags:" << std::endl;
+  for (auto &tag : GetTagList()) {
+    std::cout << "  " << tag << "=" << GetTagValue(tag) << std::endl;
+  }
+  std::cout << "Moves:" << std::endl;
+
+  if (moves != NULL)
+    moves->Dump();
+}
+
+int PGN::NextNonBlank(int loc) {
+  char c = pgn_content[loc];
+  while (IS_BLANK(c)) {
+    loc++;
+    if (IS_EOF(loc))
+      return (loc);
+    c = pgn_content[loc];
   }
+  return (loc);
+}
 
-}
\ No newline at end of file
+} // namespace pgnp
\ No newline at end of file
diff --git a/src/pgnp.hpp b/src/pgnp.hpp
index 25e74c8..a89de73 100644
--- a/src/pgnp.hpp
+++ b/src/pgnp.hpp
@@ -1,46 +1,75 @@
-#include <unordered_map>
-#include <string>
+#include <algorithm>
+#include <exception>
 #include <fstream>
-#include <streambuf>
 #include <iostream>
-#include <exception>
+#include <streambuf>
+#include <string>
+#include <unordered_map>
+#include <vector>
 
 namespace pgnp {
 
+class HalfMove {
+private:
+  /// @brief Recursive dump
+  void NestedDump(HalfMove *, int);
 
-    typedef struct HalfMove {
-
-    } HalfMove;
+public:
+  int count;
+  bool isBlack;
+  std::string move;
+  std::string comment;
+  HalfMove *MainLine;
+  std::vector<HalfMove *> variations;
 
+  HalfMove();
+  ~HalfMove();
+  int GetLength();
+  /// @brief Dump move and all its variations
+  void Dump();
+};
 
-    class PGN {
-        private:
-        std::unordered_map<std::string,std::string> tags;
-        HalfMove moves;
-        std::string pgn_content;
+class PGN {
+private:
+  std::unordered_map<std::string, std::string> tags;
+  std::vector<std::string> tagkeys;
 
-        public:
-        void FromFile(std::string);
-        void FromString(std::string);
+  HalfMove *moves;
+  std::string pgn_content;
 
+public:
+  ~PGN();
+  void FromFile(std::string);
+  void FromString(std::string);
+  bool HasTag(std::string);
+  /// @brief Perform a Seven Tag Roster compliance check
+  void STRCheck();
+  /// @brief Dump parsed PGN
+  void Dump();
+  std::vector<std::string> GetTagList();
+  std::string GetTagValue(std::string);
+  HalfMove *GetMoves();
 
-        private:
+private:
+  /// @brief Populate @a tags with by parsing the one starting at location in
+  /// argument
+  int ParseNextTag(int);
 
-        /// @brief Populate @a tags with by parsing the one starting at location in argument
-        int ParseNextTag(int);
-        
-        /// @brief Get the next non-blank char location starting from location in argument
-        int NextNonBlank(int);
-    };
+  /// @brief Get the next non-blank char location starting from location in
+  /// argument
+  int NextNonBlank(int);
 
+  int ParseLine(int, HalfMove *);
+};
 
+struct UnexpectedEOF : public std::exception {
+  const char *what() const throw() { return "Unexpected end of pgn file"; }
+};
 
-    struct UnexpectedEOF : public std::exception
-    {
-        const char * what () const throw ()
-        {
-            return "Unexpected end of pgn file";
-        }
-    };
+struct STRCheckFailed : public std::exception {
+  const char *what() const throw() {
+    return "Seven Tag Roster compliance check failed";
+  }
+};
 
-}
+} // namespace pgnp
diff --git a/tests/CMakeLists.txt b/tests/CMakeLists.txt
index cb7b080..fe07b37 100644
--- a/tests/CMakeLists.txt
+++ b/tests/CMakeLists.txt
@@ -5,4 +5,5 @@ file(COPY pgn_files DESTINATION ${CMAKE_CURRENT_BINARY_DIR}/)
 
 # Run tests
 add_executable(pgnp_tests tests.cpp ./catch3/catch_amalgamated.cpp)
+target_link_libraries(pgnp_tests pgnp)
 add_test(PGNP_Tests pgnp_tests)
diff --git a/tests/pgn_files/str/pgn1.pgn b/tests/pgn_files/str/pgn1.pgn
new file mode 100644
index 0000000..34bd807
--- /dev/null
+++ b/tests/pgn_files/str/pgn1.pgn
@@ -0,0 +1,15 @@
+[Event "F/S Return Match"]
+[Site "Belgrade, Serbia JUG"]
+[Date "1992.11.04"]
+[Round "29"]
+[White "Fischer, Robert J."]
+[Black "Spassky, Boris V."]
+[Result "1/2-1/2"]
+
+1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 d6 8. c3
+O-O 9. h3 Nb8 10. d4 Nbd7 11. c4 c6 12. cxb5 axb5 13. Nc3 Bb7 14. Bg5 b4 15.
+Nb1 h6 16. Bh4 c5 17. dxe5 Nxe4 18. Bxe7 Qxe7 19. exd6 Qf6 20. Nbd2 Nxd6 21.
+Nc4 Nxc4 22. Bxc4 Nb6 23. Ne5 Rae8 24. Bxf7+ Rxf7 25. Nxf7 Rxe1+ 26. Qxe1 Kxf7
+27. Qe3 Qg5 28. Qxg5 hxg5 29. b3 Ke6 30. a3 Kd6 31. axb4 cxb4 32. Ra5 Nd5 33.
+f3 Bc8 34. Kf2 Bf5 35. Ra7 g6 36. Ra6+ Kc5 37. Ke1 Nf4 38. g3 Nxh3 39. Kd2 Kb5
+40. Rd6 Kc5 41. Ra6 Nf2 42. g4 Bd3 43. Re6 1/2-1/2
diff --git a/tests/pgn_files/valid/pgn1.pgn b/tests/pgn_files/valid/pgn1.pgn
index bca769e..f593f02 100644
--- a/tests/pgn_files/valid/pgn1.pgn
+++ b/tests/pgn_files/valid/pgn1.pgn
@@ -17,4 +17,4 @@
 [Termination "Normal"]
 [Annotator "lichess.org"]
 
-1. g3 { A00 Hungarian Opening } d5 2. Bg2 Nf6 3. c4 c6 4. Nf3 h6?! { (0.22 → 0.74) Inaccuracy. dxc4 was best. } (4... dxc4 5. O-O Nbd7 6. Qc2 Nb6 7. a4 a5 8. Na3 Be6 9. Ne5) 5. e3?! { (0.74 → -0.15) Inaccuracy. O-O was best. } (5. O-O Bf5 6. d3 e6 7. cxd5 cxd5 8. Qb3 Qb6 9. Nd4 Bg6) 5... Bf5 6. Nc3?! { (-0.15 → -0.75) Inaccuracy. d4 was best. } (6. d4 Nbd7 7. b3 e6 8. O-O Be7 9. Nc3 O-O 10. Bb2 a5 11. Qe2 Ne4 12. Nxe4 Bxe4) 6... e6 7. d4 Be7 8. Qe2 O-O 9. a3 a5 10. Bd2?! { (-0.33 → -0.84) Inaccuracy. b3 was best. } (10. b3 Re8) 10... Bc2 11. O-O Bb3 12. c5?? { (-0.22 → -3.09) Blunder. cxd5 was best. } (12. cxd5) 12... Bc4 13. Qd1 Bxf1 14. Bxf1 b6 15. cxb6 Qxb6 16. Na4 Qa7 17. Rc1 Rc8 18. Ne5 c5 19. dxc5 Bxc5 20. Bb5? { (-2.12 → -4.00) Mistake. Rc2 was best. } (20. Rc2 Nbd7) 20... Ne4?? { (-4.00 → -1.02) Blunder. Qb7 was best. } (20... Qb7 21. Be8) 21. Kg2?! { (-1.02 → -2.00) Inaccuracy. b4 was best. } (21. b4) 21... Qb7 22. f3?? { (-2.13 → -7.56) Blunder. Bd3 was best. } (22. Bd3 Ba7) 22... Nxd2 23. Bc6 Nxc6 24. Nxc5 Qxb2 25. Ned3 Qxa3 26. Qxd2 Nb4 27. Nb2 Rc7 28. Kh3 Rac8 29. Nba4 Na6 30. Rc3 Qa1 31. Rc1 Qf6 32. Qxa5 Nxc5 33. Nb6 Nb3 { White resigns. } 0-1
+1. g3 d5 2. Bg2 Nf6 3. c4 c6
diff --git a/tests/pgn_files/valid/pgn2.pgn b/tests/pgn_files/valid/pgn2.pgn
new file mode 100644
index 0000000..bca769e
--- /dev/null
+++ b/tests/pgn_files/valid/pgn2.pgn
@@ -0,0 +1,20 @@
+[Event "Rated Rapid game"]
+[Site "https://lichess.org/cMjillK3"]
+[Date "2021.11.08"]
+[White "UpSideGravity"]
+[Black "manzerbredes"]
+[Result "0-1"]
+[UTCDate "2021.11.08"]
+[UTCTime "11:06:47"]
+[WhiteElo "1830"]
+[BlackElo "1818"]
+[WhiteRatingDiff "-45"]
+[BlackRatingDiff "+13"]
+[Variant "Standard"]
+[TimeControl "600+5"]
+[ECO "A00"]
+[Opening "Hungarian Opening"]
+[Termination "Normal"]
+[Annotator "lichess.org"]
+
+1. g3 { A00 Hungarian Opening } d5 2. Bg2 Nf6 3. c4 c6 4. Nf3 h6?! { (0.22 → 0.74) Inaccuracy. dxc4 was best. } (4... dxc4 5. O-O Nbd7 6. Qc2 Nb6 7. a4 a5 8. Na3 Be6 9. Ne5) 5. e3?! { (0.74 → -0.15) Inaccuracy. O-O was best. } (5. O-O Bf5 6. d3 e6 7. cxd5 cxd5 8. Qb3 Qb6 9. Nd4 Bg6) 5... Bf5 6. Nc3?! { (-0.15 → -0.75) Inaccuracy. d4 was best. } (6. d4 Nbd7 7. b3 e6 8. O-O Be7 9. Nc3 O-O 10. Bb2 a5 11. Qe2 Ne4 12. Nxe4 Bxe4) 6... e6 7. d4 Be7 8. Qe2 O-O 9. a3 a5 10. Bd2?! { (-0.33 → -0.84) Inaccuracy. b3 was best. } (10. b3 Re8) 10... Bc2 11. O-O Bb3 12. c5?? { (-0.22 → -3.09) Blunder. cxd5 was best. } (12. cxd5) 12... Bc4 13. Qd1 Bxf1 14. Bxf1 b6 15. cxb6 Qxb6 16. Na4 Qa7 17. Rc1 Rc8 18. Ne5 c5 19. dxc5 Bxc5 20. Bb5? { (-2.12 → -4.00) Mistake. Rc2 was best. } (20. Rc2 Nbd7) 20... Ne4?? { (-4.00 → -1.02) Blunder. Qb7 was best. } (20... Qb7 21. Be8) 21. Kg2?! { (-1.02 → -2.00) Inaccuracy. b4 was best. } (21. b4) 21... Qb7 22. f3?? { (-2.13 → -7.56) Blunder. Bd3 was best. } (22. Bd3 Ba7) 22... Nxd2 23. Bc6 Nxc6 24. Nxc5 Qxb2 25. Ned3 Qxa3 26. Qxd2 Nb4 27. Nb2 Rc7 28. Kh3 Rac8 29. Nba4 Na6 30. Rc3 Qa1 31. Rc1 Qf6 32. Qxa5 Nxc5 33. Nb6 Nb3 { White resigns. } 0-1
diff --git a/tests/tests.cpp b/tests/tests.cpp
index 2877213..9c65975 100644
--- a/tests/tests.cpp
+++ b/tests/tests.cpp
@@ -1,6 +1,25 @@
+#include "pgnp.hpp"
 #include <catch_amalgamated.hpp>
 
+using namespace pgnp;
 
-TEST_CASE( "DUMMY TEST", "[test]" ) {
-    REQUIRE( 1 == 1 );
+TEST_CASE("Valid PGN", "[pgn1]") {
+  PGN pgn;
+  REQUIRE_NOTHROW(pgn.FromFile("pgn_files/valid/pgn1.pgn"));
+  REQUIRE_THROWS(pgn.STRCheck());
+  REQUIRE(pgn.GetMoves()->GetLength() == 6);
+}
+
+TEST_CASE("Valid PGN", "[pgn2]") {
+  PGN pgn;
+  REQUIRE_NOTHROW(pgn.FromFile("pgn_files/valid/pgn2.pgn"));
+  REQUIRE_THROWS(pgn.STRCheck());
+  REQUIRE(pgn.GetMoves()->GetLength() == 66);
+}
+
+TEST_CASE("Seven Tag Roster", "[pgn1]") {
+  PGN pgn;
+  REQUIRE_NOTHROW(pgn.FromFile("pgn_files/str/pgn1.pgn"));
+  REQUIRE_NOTHROW(pgn.STRCheck());
+  REQUIRE(pgn.GetMoves()->GetLength() == 85);
 }