1 files changed, 121 insertions, 0 deletions
diff --git a/Homework/cs5300/project-one/p1-scanner/scanner/Scanner.java b/Homework/cs5300/project-one/p1-scanner/scanner/Scanner.java
new file mode 100644
index 0000000..2581dcb
--- /dev/null
+++ b/Homework/cs5300/project-one/p1-scanner/scanner/Scanner.java
@@ -0,0 +1,121 @@
+package scanner;
+
+import java.util.HashMap;
+import java.util.Stack;
+
+public class Scanner {
+  HashMap<Character, String> charCategories;
+  HashMap<String, HashMap<String, String>> transitions;
+  HashMap<String, String> tokenTypes;
+  String s0;
+
+  /**
+   * Builds the tables needed for the scanner.
+   */
+  public Scanner(TableReader tableReader) {
+    charCategories = new HashMap<>();
+    transitions = new HashMap<>();
+    tokenTypes = new HashMap<>();
+
+    // Build catMap, mapping a character to a category.
+    for (TableReader.CharCat cat : tableReader.getClassifier())
+      charCategories.put(cat.getC(), cat.getCategory());
+
+    // Build the transition table. Given a state and a character category,
+    // give a new state.
+    for (TableReader.Transition t : tableReader.getTransitions()) {
+      if (s0 == null)
+        s0 = t.getFromStateName();
+      if (!transitions.containsKey(t.getFromStateName()))
+        transitions.put(t.getFromStateName(), new HashMap<String, String>());
+
+      transitions.get(t.getFromStateName())
+          .put(t.getCategory(), t.getToStateName());
+    }
+
+    // Build the token types table
+    for (TableReader.TokenType tt : tableReader.getTokens())
+      tokenTypes.put(tt.getState(), tt.getType());
+  }
+
+  /**
+   * Returns the category for c or "not in alphabet" if c has no category. Do
+   * not hardcode this. That is, this function should have nothing more than a
+   * table lookup or two. You should not have any character literals in here
+   * such as 'r' or '3'.
+   */
+  public String getCategory(Character c) {
+    if (charCategories.containsKey(c))
+      return charCategories.get(c);
+
+    return "not in alphabet";
+  }
+
+  /**
+   * Returns the new state given a current state and category. This models
+   * the transition table. Returns "error" if there is no transition.
+   * Do not hardcode any state names or categories. You should have only
+   * table lookups here.
+   */
+  public String getNewState(String state, String category) {
+    if (transitionExists(state, category))
+      return transitions.get(state).get(category);
+
+    return "error";
+  }
+
+  private boolean transitionExists(String state, String category) {
+    if (transitions.containsKey(state)) {
+      HashMap<String, String> transitionTable = transitions.get(state);
+      return transitionTable.containsKey(category);
+    }
+    return false;
+  }
+
+  /**
+   * Returns the type of token corresponding to a given state. If the state
+   * is not accepting then return "error".
+   * Do not hardcode any state names or token types.
+   */
+  public String getTokenType(String state) {
+    if (tokenTypes.containsKey(state))
+      return tokenTypes.get(state);
+    return "error";
+  }
+
+  /**
+   * Return the next token or null if there's a lexical error.
+   */
+  public Token nextToken(ScanStream ss) {
+    String state = s0;
+    String lexeme = "";
+
+    Stack<String> stack = new Stack<>();
+    stack.push("bad");
+
+    while (!state.equals("error") && !ss.eof()) {
+      char c = ss.next();
+      lexeme += c;
+
+      if (tokenTypes.containsKey(state))
+        stack.clear();
+
+      stack.push(state);
+
+      state = getNewState(state, getCategory(c));
+    }
+
+    while (!tokenTypes.containsKey(state) && !state.equals("bad")) {
+      state = stack.pop();
+      if (lexeme.length() > 0)
+        lexeme = lexeme.substring(0, lexeme.length() - 1);
+
+      ss.rollback();
+    }
+
+    if (tokenTypes.containsKey(state))
+      return new Token(tokenTypes.get(state), lexeme);
+
+    return null;
+  }
+}