christoph/lecture-compilers
1
Fork 0
Code Actions Activity

rework Grammar

Browse Source
This commit is contained in:
ChUrl
2021-01-31 21:55:47 +01:00
parent 7e233d02a0
commit a2b35ad4da
10 changed files with 298 additions and 261 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/main/java/parser/ParsingTable.java
View File

@ -28,10 +28,6 @@ public class ParsingTable {
return this.parsetable.get(new SimpleEntry<>(nonterminal, terminal)); return this.parsetable.get(new SimpleEntry<>(nonterminal, terminal));
} }
public String getStartSymbol() {
return this.grammar.getStartSymbol();
}
public Set<String> getNonterminals() { public Set<String> getNonterminals() {
return this.grammar.getNonterminals(); return this.grammar.getNonterminals();
} }
@ -40,10 +36,6 @@ public class ParsingTable {
return this.grammar.getTerminals(); return this.grammar.getTerminals();
} }
public String getEpsilon() {
return this.grammar.getEpsilonSymbol();
}
@Override @Override
public String toString() { public String toString() {
final StringBuilder output = new StringBuilder(); final StringBuilder output = new StringBuilder();

4
src/main/java/parser/StupsParser.java
View File

@ -39,7 +39,7 @@ public class StupsParser {
public SyntaxTree parse(List<? extends Token> token, Vocabulary voc) { public SyntaxTree parse(List<? extends Token> token, Vocabulary voc) {
System.out.println(" - Parsing program..."); System.out.println(" - Parsing program...");
final SyntaxTreeNode root = new SyntaxTreeNode(this.parsetable.getStartSymbol(), 0); final SyntaxTreeNode root = new SyntaxTreeNode(Grammar.START_SYMBOL, 0);
final SyntaxTree tree = new SyntaxTree(root); final SyntaxTree tree = new SyntaxTree(root);
final Deque<SyntaxTreeNode> stack = new ArrayDeque<>(); final Deque<SyntaxTreeNode> stack = new ArrayDeque<>();
stack.push(root); stack.push(root);
@ -69,7 +69,7 @@ public class StupsParser {
final String prod = this.parsetable.get(top, currentTokenSym); final String prod = this.parsetable.get(top, currentTokenSym);
if (top.equals(this.parsetable.getEpsilon())) { if (top.equals(Grammar.EPSILON_SYMBOL)) {
// Wenn auf dem Stack das Epsilonsymbol liegt // Wenn auf dem Stack das Epsilonsymbol liegt
stack.pop(); stack.pop();

404
src/main/java/parser/grammar/Grammar.java
View File

@ -6,6 +6,7 @@ import java.io.IOException;
import java.nio.file.Files; import java.nio.file.Files;
import java.nio.file.Path; import java.nio.file.Path;
import java.util.Arrays; import java.util.Arrays;
import java.util.Collection;
import java.util.Collections; import java.util.Collections;
import java.util.EnumMap; import java.util.EnumMap;
import java.util.HashMap; import java.util.HashMap;
@ -13,7 +14,6 @@ import java.util.HashSet;
import java.util.List; import java.util.List;
import java.util.Map; import java.util.Map;
import java.util.Set; import java.util.Set;
import java.util.regex.Pattern;
import java.util.stream.Collectors; import java.util.stream.Collectors;
import static parser.grammar.GrammarAction.DELCHILD; import static parser.grammar.GrammarAction.DELCHILD;
@ -22,30 +22,56 @@ import static parser.grammar.GrammarAction.NAMETOVAL;
import static parser.grammar.GrammarAction.PROMOTE; import static parser.grammar.GrammarAction.PROMOTE;
import static parser.grammar.GrammarAction.RENAMETO; import static parser.grammar.GrammarAction.RENAMETO;
import static parser.grammar.GrammarAction.VALTOVAL; import static parser.grammar.GrammarAction.VALTOVAL;
import static parser.grammar.GrammarAction.values;
import static util.Logger.log; import static util.Logger.log;
/**
* Repräsentiert die Parse-Grammatik und die Kontextaktionen.
*/
public class Grammar { public class Grammar {
private static final Pattern EPS = Pattern.compile("EPS"); // Grammar
private static final Pattern ARROW = Pattern.compile("->"); public static final String START_SYMBOL = "S";
public static final String EPSILON_SYMBOL = "eps";
private final Set<String> terminals; private final Set<String> terminals;
private final Set<String> nonterminals; private final Set<String> nonterminals;
private final String startSymbol;
private final String epsilonSymbol;
// Actions // Actions
private final Map<GrammarAction, Set<String>> actions;
/**
* Jeder Kontextaktion werden alle leftsides zugewiesen, welche diese Aktion ausführen.
*/
private final Map<GrammarAction, Set<String>> actionMap;
/**
* Jeder leftside mit [renameto=name] wird der entsprechende neue Name zugewiesen.
*/
private final Map<String, String> renameMappings; private final Map<String, String> renameMappings;
/**
* Jeder leftside mit [nametoval=children] werden die entpsrechenden Children zugewiesen,
* deren Namen in die Parentvalue gschoben werden.
*/
private final Map<String, List<String>> nameToValMappings; private final Map<String, List<String>> nameToValMappings;
/**
* Jeder leftside mit [valtoval=children] werden die entpsrechenden Children zugewiesen,
* deren Values in die Parentvalue gschoben werden.
*/
private final Map<String, List<String>> valToValMappings; private final Map<String, List<String>> valToValMappings;
/**
* Jeder Leftside mit [delchild=children] werden die entpsrechenden Children zugewiesen, welche entfernt werden.
*/
private final Map<String, List<String>> delChildMappings; private final Map<String, List<String>> delChildMappings;
/**
* Die eigentlichen Produktionsregeln der Form leftside -> rightside.
*/
private final Set<GrammarRule> rules; private final Set<GrammarRule> rules;
public Grammar(Set<String> terminals, Set<String> nonterminals, public Grammar(Set<String> terminals, Set<String> nonterminals,
String startSymbol, String epsilonSymbol, Map<GrammarAction, Set<String>> actionMap,
Map<GrammarAction, Set<String>> actions,
Map<String, String> renameMappings, Map<String, String> renameMappings,
Map<String, List<String>> nameToValMappings, Map<String, List<String>> nameToValMappings,
Map<String, List<String>> valToValMappings, Map<String, List<String>> valToValMappings,
@ -55,10 +81,8 @@ public class Grammar {
this.terminals = Collections.unmodifiableSet(terminals); this.terminals = Collections.unmodifiableSet(terminals);
this.nonterminals = Collections.unmodifiableSet(nonterminals); this.nonterminals = Collections.unmodifiableSet(nonterminals);
this.rules = Collections.unmodifiableSet(rules); this.rules = Collections.unmodifiableSet(rules);
this.startSymbol = startSymbol;
this.epsilonSymbol = epsilonSymbol;
this.actions = Collections.unmodifiableMap(actions); this.actionMap = Collections.unmodifiableMap(actionMap);
this.renameMappings = Collections.unmodifiableMap(renameMappings); this.renameMappings = Collections.unmodifiableMap(renameMappings);
this.nameToValMappings = Collections.unmodifiableMap(nameToValMappings); this.nameToValMappings = Collections.unmodifiableMap(nameToValMappings);
this.valToValMappings = Collections.unmodifiableMap(valToValMappings); this.valToValMappings = Collections.unmodifiableMap(valToValMappings);
@ -69,141 +93,197 @@ public class Grammar {
System.out.println(" - Reading parser-grammar..."); System.out.println(" - Reading parser-grammar...");
List<String> lines = Files.readAllLines(path); List<String> lines = Files.readAllLines(path);
// Remove Whitespace + Comments
lines = lines.stream() lines = lines.stream()
.map(String::trim) .map(String::trim)
.filter(line -> !(line.isBlank() || line.startsWith("//"))) .filter(line -> !(line.isBlank() || line.startsWith("//")))
.collect(Collectors.toUnmodifiableList()); .collect(Collectors.toUnmodifiableList());
try { // Grammar
// Grammar final Set<String> terminals = new HashSet<>();
String startSymbol = ""; final Set<String> nonterminals = new HashSet<>();
String epsilonSymbol = ""; final Set<GrammarRule> rules = new HashSet<>();
final Set<String> terminals = new HashSet<>();
final Set<String> nonterminals = new HashSet<>();
final Set<GrammarRule> rules = new HashSet<>();
// Actions // Actions
final Map<GrammarAction, Set<String>> actions = new EnumMap<>(GrammarAction.class); final Map<GrammarAction, Set<String>> actionMap = new EnumMap<>(GrammarAction.class);
final Map<String, String> renameMappings = new HashMap<>(); final Map<String, String> renameMappings = new HashMap<>();
final Map<String, List<String>> nameToValMappings = new HashMap<>(); final Map<String, List<String>> nameToValMappings = new HashMap<>();
final Map<String, List<String>> valToValMappings = new HashMap<>(); final Map<String, List<String>> valToValMappings = new HashMap<>();
final Map<String, List<String>> delChildMappings = new HashMap<>(); final Map<String, List<String>> delChildMappings = new HashMap<>();
for (GrammarAction action : values()) { // Init actionMap
actions.put(action, new HashSet<>()); for (GrammarAction action : GrammarAction.values()) {
} actionMap.put(action, new HashSet<>());
// Init for validity check
final Set<String> actionSet = Arrays.stream(values())
.map(Enum::toString)
.collect(Collectors.toUnmodifiableSet());
log("Parsing Grammar from File:");
for (String line : lines) {
log("Parsed: " + line);
// Parse Keywords
if (line.startsWith("START:")) {
startSymbol = line.split(" ")[1];
} else if (line.startsWith("EPS:")) {
epsilonSymbol = line.split(" ")[1];
} else if (line.startsWith("TERM:")) {
terminals.addAll(Arrays.stream(line.split(" ")).skip(1).collect(Collectors.toSet()));
} else if (line.startsWith("NTERM:")) {
nonterminals.addAll(Arrays.stream(line.split(" ")).skip(1).collect(Collectors.toSet()));
} else {
// Parse Grammar Rules + Actions
// "S[...] -> E T2 | EPS" wird zu leftside = "S[...]" und rightside = "E T2 | eps"
final String[] split = ARROW.split(EPS.matcher(line).replaceAll(epsilonSymbol));
String leftside = split[0].trim();
final String rightside = split[1].trim();
if (leftside.indexOf('[') >= 0 && leftside.indexOf(']') >= 0) {
// Handle actions if they are given
final int open = leftside.indexOf('[');
final int close = leftside.indexOf(']');
// Aus "S[C R=...]" wird flags = {"C", "R=..."}
final String[] flags = leftside.substring(open + 1, close).split(" ");
final Set<String> flagSet = Arrays.stream(flags)
.map(String::trim)
.filter(flag -> !flag.isEmpty())
.collect(Collectors.toUnmodifiableSet());
// Check for action validity
for (String flag : flagSet) {
if (!actionSet.contains(flag.split("=")[0].toUpperCase())) {
throw new GrammarParseException("Invalid Action: " + flag);
}
}
// "S[C R=...]" wird zu "S"
leftside = leftside.substring(0, open).trim();
// Register actions, flagSet = {"C", "R=..."}
for (String flag : flagSet) {
final String[] flagSplit = flag.split("=");
final GrammarAction action = GrammarAction.valueOf(flagSplit[0].toUpperCase());
actions.get(action).add(leftside.trim());
log("Registered " + flag + ": " + leftside.trim());
if (flagSplit.length > 1) {
// Handle Action with arguments
// "R=A,B,C" -> argSplit = {"A", "B", "C"}
final int argStart = flag.indexOf('=');
final String[] argSplit = flag.substring(argStart + 1).split(",");
switch (action) {
case DELCHILD -> delChildMappings.put(leftside, Arrays.asList(argSplit));
case VALTOVAL -> valToValMappings.put(leftside, Arrays.asList(argSplit));
case NAMETOVAL -> nameToValMappings.put(leftside, Arrays.asList(argSplit));
case RENAMETO -> renameMappings.put(leftside, argSplit[0]);
}
}
}
}
// "E T2 | epsilon" wird zu prods[0] = "E T2" und prods[1] = "epsilon"
final String[] prods = rightside.split("\\|");
for (String prod : prods) {
final GrammarRule rule = new GrammarRule(leftside, prod.split(" "));
rules.add(rule);
}
}
}
log("\n" + actions);
log("-".repeat(100));
System.out.println("Grammar parsed successfully.");
return new Grammar(terminals, nonterminals,
startSymbol, epsilonSymbol,
actions,
renameMappings,
nameToValMappings,
valToValMappings,
delChildMappings,
rules);
} catch (Exception e) {
log("Die Grammatik kann nicht gelesen werden!");
log(path.toString());
e.printStackTrace();
} }
return null; log("Parsing Grammar from File:");
for (String currentLine : lines) {
log("Parsed: " + currentLine);
// Parse Keywords
if (currentLine.startsWith("TERM:")) {
terminals.addAll(Arrays.stream(currentLine.split(" ")).skip(1).collect(Collectors.toSet()));
} else if (currentLine.startsWith("NTERM:")) {
nonterminals.addAll(Arrays.stream(currentLine.split(" ")).skip(1).collect(Collectors.toSet()));
} else {
// Parse regular lines
parseRegularLine(currentLine, actionMap,
delChildMappings, valToValMappings, nameToValMappings, renameMappings,
rules);
}
}
log("\n" + actionMap);
log("-".repeat(100));
System.out.println("Grammar parsed successfully.");
return new Grammar(terminals, nonterminals,
actionMap, renameMappings, nameToValMappings,
valToValMappings, delChildMappings, rules);
} }
/**
* Es wird eine normale Zeile der Form leftside[actions] -> rightside geparst.
* Die Produktionsregeln sowie die Kontextaktionen werden registriert.
*/
private static void parseRegularLine(String currentLine,
Map<GrammarAction, Set<String>> actions,
Map<String, List<String>> delChildMappings,
Map<String, List<String>> valToValMappings,
Map<String, List<String>> nameToValMappings,
Map<String, String> renameMappings,
Collection<GrammarRule> rules) {
// "S[...] -> E T2 | eps" wird zu leftside = "S[...]" und rightside = "E T2 | eps"
final String[] split = currentLine.split("->");
String leftside = split[0].trim();
final String rightside = split[1].trim();
final int open = leftside.indexOf('[');
final int close = leftside.indexOf(']');
if (open >= 0 && close >= 0) {
// Handle actions if they are given
final Set<String> actionSet = parseActionSet(leftside, open, close);
// Validate Actions
throwOnInvalidActionSet(actionSet);
// "S[C R=...]" wird zu "S"
leftside = leftside.substring(0, open).trim();
// Register actions, flagSet = {"C", "R=..."}
for (String flag : actionSet) {
registerAction(flag, leftside, actions,
delChildMappings, valToValMappings, nameToValMappings, renameMappings);
}
}
registerProductionRules(leftside, rightside, rules);
}
/**
* Es wird die Menge an Kontextaktionen [action1,action2,...] ermittelt.
*/
private static Set<String> parseActionSet(String leftside, int open, int close) {
// Aus "S[C R=...]" wird flags = {"C", "R=..."}
final String[] flags = leftside.substring(open + 1, close).split(" ");
return Arrays.stream(flags)
.map(String::trim)
.filter(flag -> !flag.isEmpty())
.collect(Collectors.toUnmodifiableSet());
}
/**
* Es wird eine beliebige Kontextaktion geparst und der entsprechenden Map hinzugefügt.
*/
private static void registerAction(String flag, String leftside,
Map<GrammarAction, Set<String>> actions,
Map<String, List<String>> delChildMappings,
Map<String, List<String>> valToValMappings,
Map<String, List<String>> nameToValMappings,
Map<String, String> renameMappings) {
final String[] flagSplit = flag.split("=");
final GrammarAction action = GrammarAction.valueOf(flagSplit[0].toUpperCase());
registerRegularAction(action, leftside, flag, actions);
if (flagSplit.length > 1) {
registerActionArguments(flag, action, leftside,
delChildMappings, valToValMappings, nameToValMappings, renameMappings);
}
}
/**
* Es wird ein Eintrag in der action-Map mit der entsprechenden leftside hinzugefügt.
*/
private static void registerRegularAction(GrammarAction action, String leftside, String flag,
Map<GrammarAction, Set<String>> actions) {
actions.get(action).add(leftside.trim());
log("Registered " + flag + ": " + leftside.trim());
}
/**
* Es wird eine Kontextaktion der Form [action=arguments] geparst und der entsprechenden Map hinzugefügt.
*/
private static void registerActionArguments(String flag, GrammarAction action, String leftside,
Map<String, List<String>> delChildMappings,
Map<String, List<String>> valToValMappings,
Map<String, List<String>> nameToValMappings,
Map<String, String> renameMappings) {
// "R=A,B,C" -> argSplit = {"A", "B", "C"}
final int argStart = flag.indexOf('=');
final String[] argSplit = flag.substring(argStart + 1).split(",");
switch (action) {
case DELCHILD -> delChildMappings.put(leftside, Arrays.asList(argSplit));
case VALTOVAL -> valToValMappings.put(leftside, Arrays.asList(argSplit));
case NAMETOVAL -> nameToValMappings.put(leftside, Arrays.asList(argSplit));
case RENAMETO -> renameMappings.put(leftside, argSplit[0]);
default -> throw new GrammarParseException("Unexpected value for arguments: " + action);
}
}
/**
* Der Regelmenge wird eine neue Regel der Form leftside -> rightside hinzugefügt.
* Ist rightside dabei verodert, also leftside -> right1 | right2 | right3, dann
* wird rightside gesplittet.
*/
private static void registerProductionRules(String leftside, String rightside, Collection<GrammarRule> rules) {
// "E T2 | epsilon" wird zu prods[0] = "E T2" und prods[1] = "epsilon"
final String[] prods = rightside.split("\\|");
for (String prod : prods) {
final GrammarRule rule = new GrammarRule(leftside, prod.split(" "));
rules.add(rule);
}
}
private static void throwOnInvalidActionSet(Iterable<String> flagSet) {
final Set<String> actionSet = Arrays.stream(GrammarAction.values())
.map(Enum::toString)
.collect(Collectors.toUnmodifiableSet());
for (String flag : flagSet) {
if (!actionSet.contains(flag.split("=")[0].toUpperCase())) {
throw new GrammarParseException("Invalid Action: " + flag);
}
}
}
// Getters
public Set<String> getTerminals() { public Set<String> getTerminals() {
return this.terminals; return this.terminals;
} }
@ -212,18 +292,13 @@ public class Grammar {
return this.nonterminals; return this.nonterminals;
} }
public String getStartSymbol() {
return this.startSymbol;
}
public String getEpsilonSymbol() {
return this.epsilonSymbol;
}
public Set<GrammarRule> getRules() { public Set<GrammarRule> getRules() {
return this.rules; return this.rules;
} }
/**
* Ermittelt alle möglichen Produktionen, welche zu einer leftside gehören können.
*/
public Set<String> getRightsides(String leftside) { public Set<String> getRightsides(String leftside) {
return this.rules.stream() return this.rules.stream()
.filter(rule -> rule.getLeftside().equals(leftside)) .filter(rule -> rule.getLeftside().equals(leftside))
@ -249,11 +324,10 @@ public class Grammar {
&& root.getValue().isEmpty(); && root.getValue().isEmpty();
} }
private boolean canPromoteChild(String sym) { private boolean canPromoteChild(String rootName) {
return this.actions.get(PROMOTE).contains(sym); return this.actionMap.get(PROMOTE).contains(rootName);
} }
/** /**
* Checkt auch auf Anzahl der Kinder und vorhandene Value. * Checkt auch auf Anzahl der Kinder und vorhandene Value.
*/ */
@ -263,11 +337,10 @@ public class Grammar {
&& root.isEmpty(); && root.isEmpty();
} }
public boolean canDeleteIfEmpty(String sym) { public boolean canDeleteIfEmpty(String rootName) {
return this.actions.get(DELIFEMPTY).contains(sym); return this.actionMap.get(DELIFEMPTY).contains(rootName);
} }
/** /**
* Checkt auch auf Anzahl der Kinder. * Checkt auch auf Anzahl der Kinder.
* Epsilon-Knoten werden immer gelöscht. * Epsilon-Knoten werden immer gelöscht.
@ -277,40 +350,37 @@ public class Grammar {
&& child.isEmpty(); && child.isEmpty();
} }
public boolean canDeleteChild(String parent, String child) { public boolean canDeleteChild(String parentName, String childName) {
return (this.actions.get(DELCHILD).contains(parent) return (this.actionMap.get(DELCHILD).contains(parentName)
&& this.delChildMappings.get(parent).contains(child)) && this.delChildMappings.get(parentName).contains(childName))
|| (child.equals(this.epsilonSymbol)); || (Grammar.EPSILON_SYMBOL.equals(childName));
} }
public boolean canBeRenamed(SyntaxTreeNode root) { public boolean canBeRenamed(SyntaxTreeNode root) {
return this.canBeRenamed(root.getName()); return this.canBeRenamed(root.getName());
} }
public boolean canBeRenamed(String sym) { public boolean canBeRenamed(String rootName) {
return this.actions.get(RENAMETO).contains(sym); return this.actionMap.get(RENAMETO).contains(rootName);
} }
public String getNewName(SyntaxTreeNode root) { public String getNewName(SyntaxTreeNode root) {
return this.getNewName(root.getName()); return this.getNewName(root.getName());
} }
public String getNewName(String sym) { public String getNewName(String rootName) {
return this.renameMappings.get(sym); return this.renameMappings.get(rootName);
} }
public boolean hasValToVal(SyntaxTreeNode parent, SyntaxTreeNode child) { public boolean hasValToVal(SyntaxTreeNode parent, SyntaxTreeNode child) {
return this.hasValToVal(parent.getName(), child.getName()); return this.hasValToVal(parent.getName(), child.getName());
} }
public boolean hasValToVal(String parent, String child) { public boolean hasValToVal(String parentName, String childName) {
return this.actions.get(VALTOVAL).contains(parent) return this.actionMap.get(VALTOVAL).contains(parentName)
&& this.valToValMappings.get(parent).contains(child); && this.valToValMappings.get(parentName).contains(childName);
} }
/** /**
* Checkt auch auf bereits existierende Values. * Checkt auch auf bereits existierende Values.
*/ */
@ -319,8 +389,8 @@ public class Grammar {
&& parent.getValue().isEmpty(); && parent.getValue().isEmpty();
} }
public boolean canMoveNameToVal(String parent, String child) { public boolean canMoveNameToVal(String parentName, String childName) {
return this.actions.get(NAMETOVAL).contains(parent) return this.actionMap.get(NAMETOVAL).contains(parentName)
&& this.nameToValMappings.get(parent).contains(child); && this.nameToValMappings.get(parentName).contains(childName);
} }
} }

34
src/main/java/parser/grammar/GrammarAnalyzer.java
View File

@ -46,9 +46,9 @@ public class GrammarAnalyzer {
// Die Methode funktioniert erst, nachdem first initialisiert ist. // Die Methode funktioniert erst, nachdem first initialisiert ist.
// Deshalb hier doppelt. // Deshalb hier doppelt.
final Predicate<String> nullable = sym -> sym.equals(this.grammar.getEpsilonSymbol()) final Predicate<String> nullable = sym -> sym.equals(Grammar.EPSILON_SYMBOL)
|| sym.isBlank() || sym.isBlank()
|| firstOut.get(sym).contains(this.grammar.getEpsilonSymbol()); || firstOut.get(sym).contains(Grammar.EPSILON_SYMBOL);
final Predicate<String[]> allNullable = split -> split.length == 0 final Predicate<String[]> allNullable = split -> split.length == 0
|| Arrays.stream(split).allMatch(nullable); || Arrays.stream(split).allMatch(nullable);
@ -76,7 +76,7 @@ public class GrammarAnalyzer {
for (String rightside : this.grammar.getRightsides(leftside)) { for (String rightside : this.grammar.getRightsides(leftside)) {
// ...and X -> Y1 Y2 ... Yk is a production... // ...and X -> Y1 Y2 ... Yk is a production...
if (!rightside.equals(this.grammar.getEpsilonSymbol())) { if (!rightside.equals(Grammar.EPSILON_SYMBOL)) {
// ...for some k >= 1... // ...for some k >= 1...
final String[] split = rightside.split(" "); final String[] split = rightside.split(" ");
@ -94,7 +94,7 @@ public class GrammarAnalyzer {
// Because a != epsilon // Because a != epsilon
final Set<String> firstYiNoEps = firstOut.get(split[i]).stream() final Set<String> firstYiNoEps = firstOut.get(split[i]).stream()
.filter(sym -> !sym.equals(this.grammar.getEpsilonSymbol())) .filter(sym -> !sym.equals(Grammar.EPSILON_SYMBOL))
.collect(Collectors.toSet()); .collect(Collectors.toSet());
final boolean changeNow = firstOut.get(leftside).addAll(firstYiNoEps); final boolean changeNow = firstOut.get(leftside).addAll(firstYiNoEps);
@ -106,21 +106,21 @@ public class GrammarAnalyzer {
if (i == split.length - 1 && allNullable.test(split)) { if (i == split.length - 1 && allNullable.test(split)) {
// 2. (b) If epsilon is in first(Y1) ... first(Yk), then add epsilon to first(X). // 2. (b) If epsilon is in first(Y1) ... first(Yk), then add epsilon to first(X).
final boolean changeNow = firstOut.get(leftside).add(this.grammar.getEpsilonSymbol()); final boolean changeNow = firstOut.get(leftside).add(Grammar.EPSILON_SYMBOL);
change = change || changeNow; change = change || changeNow;
logIfTrue(changeNow, "First: Added " + this.grammar.getEpsilonSymbol() + " to " + leftside + " (All are nullable)"); logIfTrue(changeNow, "First: Added " + Grammar.EPSILON_SYMBOL + " to " + leftside + " (All are nullable)");
} }
} }
} }
if (rightside.equals(this.grammar.getEpsilonSymbol())) { if (rightside.equals(Grammar.EPSILON_SYMBOL)) {
// 3. If X -> epsilon is a production, then add epsilon to first(X). // 3. If X -> epsilon is a production, then add epsilon to first(X).
final boolean changeNow = firstOut.get(leftside).add(this.grammar.getEpsilonSymbol()); final boolean changeNow = firstOut.get(leftside).add(Grammar.EPSILON_SYMBOL);
change = change || changeNow; change = change || changeNow;
logIfTrue(changeNow, "First: Added " + this.grammar.getEpsilonSymbol() + " to " + leftside + " (X -> EPS exists)"); logIfTrue(changeNow, "First: Added " + Grammar.EPSILON_SYMBOL + " to " + leftside + " (X -> EPS exists)");
} }
} }
} }
@ -143,7 +143,7 @@ public class GrammarAnalyzer {
} }
// 1. Place $ in follow(S), where S is the start symbol, and $ is the input right endmarker // 1. Place $ in follow(S), where S is the start symbol, and $ is the input right endmarker
followOut.get(this.grammar.getStartSymbol()).add("$"); followOut.get(Grammar.START_SYMBOL).add("$");
boolean change; boolean change;
@ -176,7 +176,7 @@ public class GrammarAnalyzer {
if (this.allNullable(sub)) { if (this.allNullable(sub)) {
final Set<String> firstXkNoEps = this.first(split[k]).stream() final Set<String> firstXkNoEps = this.first(split[k]).stream()
.filter(sym -> !sym.equals(this.grammar.getEpsilonSymbol())) .filter(sym -> !sym.equals(Grammar.EPSILON_SYMBOL))
.collect(Collectors.toSet()); .collect(Collectors.toSet());
final boolean changeNow = followOut.get(split[i - 1]).addAll(firstXkNoEps); final boolean changeNow = followOut.get(split[i - 1]).addAll(firstXkNoEps);
@ -241,7 +241,7 @@ public class GrammarAnalyzer {
final Set<String> followLeftside = this.follow(leftside); final Set<String> followLeftside = this.follow(leftside);
if (firstRightside.contains(this.grammar.getEpsilonSymbol())) { if (firstRightside.contains(Grammar.EPSILON_SYMBOL)) {
// 2. If epsilon in first(a), then... // 2. If epsilon in first(a), then...
for (String sym : followLeftside) { for (String sym : followLeftside) {
@ -276,8 +276,8 @@ public class GrammarAnalyzer {
public boolean nullable(String sym) { public boolean nullable(String sym) {
return sym.isBlank() return sym.isBlank()
|| sym.equals(this.grammar.getEpsilonSymbol()) || sym.equals(Grammar.EPSILON_SYMBOL)
|| this.first.get(sym).contains(this.grammar.getEpsilonSymbol()); || this.first.get(sym).contains(Grammar.EPSILON_SYMBOL);
} }
public boolean allNullable(String[] split) { public boolean allNullable(String[] split) {
@ -304,7 +304,7 @@ public class GrammarAnalyzer {
// X1 ... Xi-1 are nullable, so first(X1 ... Xn) contains first(Xi) // X1 ... Xi-1 are nullable, so first(X1 ... Xn) contains first(Xi)
final Set<String> firstXiNoEps; final Set<String> firstXiNoEps;
if (split.length == 1 && split[0].equals(this.grammar.getEpsilonSymbol())) { if (split.length == 1 && split[0].equals(Grammar.EPSILON_SYMBOL)) {
// Stream collect has to be evaluated, doesn't work on empty stream // Stream collect has to be evaluated, doesn't work on empty stream
firstXiNoEps = Collections.emptySet(); firstXiNoEps = Collections.emptySet();
@ -312,7 +312,7 @@ public class GrammarAnalyzer {
// Only non-epsilon symbols // Only non-epsilon symbols
firstXiNoEps = this.first(split[i]).stream() firstXiNoEps = this.first(split[i]).stream()
.filter(sym -> !sym.equals(this.grammar.getEpsilonSymbol())) .filter(sym -> !sym.equals(Grammar.EPSILON_SYMBOL))
.collect(Collectors.toSet()); .collect(Collectors.toSet());
} }
@ -321,7 +321,7 @@ public class GrammarAnalyzer {
if (i == split.length - 1 && this.allNullable(split)) { if (i == split.length - 1 && this.allNullable(split)) {
// Finally, add epsilon to first(X1 X2 ... Xn) if, for all i, epsilon is in first(Xi). // Finally, add epsilon to first(X1 X2 ... Xn) if, for all i, epsilon is in first(Xi).
firstOut.add(this.grammar.getEpsilonSymbol()); firstOut.add(Grammar.EPSILON_SYMBOL);
} }
} }
} }

65
src/test/java/parser/grammar/GrammarAnalyzerTest.java
View File

@ -33,15 +33,12 @@ class GrammarAnalyzerTest {
final String[] tarray = {"a", "b", "e", "i", "t"}; final String[] tarray = {"a", "b", "e", "i", "t"};
terminals = new HashSet<>(Arrays.asList(tarray)); terminals = new HashSet<>(Arrays.asList(tarray));
final String startSymbol = "S";
final String epsilonSymbol = "epsilon";
final Set<GrammarRule> rules = new HashSet<>(); final Set<GrammarRule> rules = new HashSet<>();
rules.add(new GrammarRule("S", "a")); rules.add(new GrammarRule("S", "a"));
rules.add(new GrammarRule("S", "i", "E", "t", "S")); rules.add(new GrammarRule("S", "i", "E", "t", "S"));
rules.add(new GrammarRule("E", "b")); rules.add(new GrammarRule("E", "b"));
grammar0 = new Grammar(terminals, nonterminals, startSymbol, epsilonSymbol, grammar0 = new Grammar(terminals, nonterminals,
Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(),
Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(), rules); Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(), rules);
} }
@ -53,27 +50,24 @@ class GrammarAnalyzerTest {
*/ */
final Set<String> nonterminals; final Set<String> nonterminals;
final String[] narray = {"E", "T", "E2", "T2", "F"}; final String[] narray = {"S", "T", "E2", "T2", "F"};
nonterminals = new HashSet<>(Arrays.asList(narray)); nonterminals = new HashSet<>(Arrays.asList(narray));
final Set<String> terminals; final Set<String> terminals;
final String[] tarray = {"id", "+", "*", "(", ")"}; final String[] tarray = {"id", "+", "*", "(", ")"};
terminals = new HashSet<>(Arrays.asList(tarray)); terminals = new HashSet<>(Arrays.asList(tarray));
final String startSymbol = "E";
final String epsilonSymbol = "epsilon";
final Set<GrammarRule> rules = new HashSet<>(); final Set<GrammarRule> rules = new HashSet<>();
rules.add(new GrammarRule("E", "T", "E2")); rules.add(new GrammarRule("S", "T", "E2"));
rules.add(new GrammarRule("E2", "+", "T", "E2")); rules.add(new GrammarRule("E2", "+", "T", "E2"));
rules.add(new GrammarRule("E2", epsilonSymbol)); rules.add(new GrammarRule("E2", Grammar.EPSILON_SYMBOL));
rules.add(new GrammarRule("T", "F", "T2")); rules.add(new GrammarRule("T", "F", "T2"));
rules.add(new GrammarRule("T2", "*", "F", "T2")); rules.add(new GrammarRule("T2", "*", "F", "T2"));
rules.add(new GrammarRule("T2", epsilonSymbol)); rules.add(new GrammarRule("T2", Grammar.EPSILON_SYMBOL));
rules.add(new GrammarRule("F", "(", "E", ")")); rules.add(new GrammarRule("F", "(", "S", ")"));
rules.add(new GrammarRule("F", "id")); rules.add(new GrammarRule("F", "id"));
grammar1 = new Grammar(terminals, nonterminals, startSymbol, epsilonSymbol, grammar1 = new Grammar(terminals, nonterminals,
Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(),
Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(), rules); Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(), rules);
} }
@ -90,25 +84,22 @@ class GrammarAnalyzerTest {
*/ */
final Set<String> nonterminals; final Set<String> nonterminals;
final String[] narray = {"X", "Y", "Z"}; final String[] narray = {"X", "Y", "S"};
nonterminals = new HashSet<>(Arrays.asList(narray)); nonterminals = new HashSet<>(Arrays.asList(narray));
final Set<String> terminals; final Set<String> terminals;
final String[] tarray = {"a", "c", "d"}; final String[] tarray = {"a", "c", "d"};
terminals = new HashSet<>(Arrays.asList(tarray)); terminals = new HashSet<>(Arrays.asList(tarray));
final String startSymbol = "Z";
final String epsilonSymbol = "epsilon";
final Set<GrammarRule> rules = new HashSet<>(); final Set<GrammarRule> rules = new HashSet<>();
rules.add(new GrammarRule("Z", "d")); rules.add(new GrammarRule("S", "d"));
rules.add(new GrammarRule("Z", "X", "Y", "Z")); rules.add(new GrammarRule("S", "X", "Y", "S"));
rules.add(new GrammarRule("Y", epsilonSymbol)); rules.add(new GrammarRule("Y", Grammar.EPSILON_SYMBOL));
rules.add(new GrammarRule("Y", "c")); rules.add(new GrammarRule("Y", "c"));
rules.add(new GrammarRule("X", "Y")); rules.add(new GrammarRule("X", "Y"));
rules.add(new GrammarRule("X", "a")); rules.add(new GrammarRule("X", "a"));
grammar2 = new Grammar(terminals, nonterminals, startSymbol, epsilonSymbol, grammar2 = new Grammar(terminals, nonterminals,
Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(),
Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(), rules); Collections.emptyMap(), Collections.emptyMap(), Collections.emptyMap(), rules);
} }
@ -125,10 +116,10 @@ class GrammarAnalyzerTest {
void testFirstGrammar1() { void testFirstGrammar1() {
final GrammarAnalyzer analyzer = new GrammarAnalyzer(grammar1); final GrammarAnalyzer analyzer = new GrammarAnalyzer(grammar1);
assertThat(analyzer.getFirst().get("E")).containsOnly("id", "("); assertThat(analyzer.getFirst().get("S")).containsOnly("id", "(");
assertThat(analyzer.getFirst().get("E2")).containsOnly("+", grammar1.getEpsilonSymbol()); assertThat(analyzer.getFirst().get("E2")).containsOnly("+", Grammar.EPSILON_SYMBOL);
assertThat(analyzer.getFirst().get("T")).containsOnly("id", "("); assertThat(analyzer.getFirst().get("T")).containsOnly("id", "(");
assertThat(analyzer.getFirst().get("T2")).containsOnly("*", grammar1.getEpsilonSymbol()); assertThat(analyzer.getFirst().get("T2")).containsOnly("*", Grammar.EPSILON_SYMBOL);
assertThat(analyzer.getFirst().get("F")).containsOnly("id", "("); assertThat(analyzer.getFirst().get("F")).containsOnly("id", "(");
} }
@ -136,9 +127,9 @@ class GrammarAnalyzerTest {
void testFirstGrammar2() { void testFirstGrammar2() {
final GrammarAnalyzer analyzer = new GrammarAnalyzer(grammar2); final GrammarAnalyzer analyzer = new GrammarAnalyzer(grammar2);
assertThat(analyzer.getFirst().get("X")).containsOnly("c", "a", grammar2.getEpsilonSymbol()); assertThat(analyzer.getFirst().get("X")).containsOnly("c", "a", Grammar.EPSILON_SYMBOL);
assertThat(analyzer.getFirst().get("Y")).containsOnly("c", grammar2.getEpsilonSymbol()); assertThat(analyzer.getFirst().get("Y")).containsOnly("c", Grammar.EPSILON_SYMBOL);
assertThat(analyzer.getFirst().get("Z")).containsOnly("c", "a", "d"); assertThat(analyzer.getFirst().get("S")).containsOnly("c", "a", "d");
} }
@Test @Test
@ -153,7 +144,7 @@ class GrammarAnalyzerTest {
void testFollowGrammar1() { void testFollowGrammar1() {
final GrammarAnalyzer analyzer = new GrammarAnalyzer(grammar1); final GrammarAnalyzer analyzer = new GrammarAnalyzer(grammar1);
assertThat(analyzer.getFollow().get("E")).containsOnly(")", "$"); assertThat(analyzer.getFollow().get("S")).containsOnly(")", "$");
assertThat(analyzer.getFollow().get("E2")).containsOnly(")", "$"); assertThat(analyzer.getFollow().get("E2")).containsOnly(")", "$");
assertThat(analyzer.getFollow().get("T")).containsOnly("+", ")", "$"); assertThat(analyzer.getFollow().get("T")).containsOnly("+", ")", "$");
assertThat(analyzer.getFollow().get("T2")).containsOnly("+", ")", "$"); assertThat(analyzer.getFollow().get("T2")).containsOnly("+", ")", "$");
@ -166,7 +157,7 @@ class GrammarAnalyzerTest {
assertThat(analyzer.getFollow().get("X")).containsOnly("a", "c", "d"); assertThat(analyzer.getFollow().get("X")).containsOnly("a", "c", "d");
assertThat(analyzer.getFollow().get("Y")).containsOnly("a", "c", "d"); assertThat(analyzer.getFollow().get("Y")).containsOnly("a", "c", "d");
assertThat(analyzer.getFollow().get("Z")).containsOnly("$"); assertThat(analyzer.getFollow().get("S")).containsOnly("$");
} }
@Test @Test
@ -174,18 +165,18 @@ class GrammarAnalyzerTest {
final GrammarAnalyzer analyzer = new GrammarAnalyzer(grammar1); final GrammarAnalyzer analyzer = new GrammarAnalyzer(grammar1);
final ParsingTable table = analyzer.getTable(); final ParsingTable table = analyzer.getTable();
assertThat(table.get("E", "id")).isEqualTo("T E2"); assertThat(table.get("S", "id")).isEqualTo("T E2");
assertThat(table.get("E", "(")).isEqualTo("T E2"); assertThat(table.get("S", "(")).isEqualTo("T E2");
assertThat(table.get("E2", "+")).isEqualTo("+ T E2"); assertThat(table.get("E2", "+")).isEqualTo("+ T E2");
assertThat(table.get("E2", ")")).isEqualTo(grammar1.getEpsilonSymbol()); assertThat(table.get("E2", ")")).isEqualTo(Grammar.EPSILON_SYMBOL);
assertThat(table.get("E2", "$")).isEqualTo(grammar1.getEpsilonSymbol()); assertThat(table.get("E2", "$")).isEqualTo(Grammar.EPSILON_SYMBOL);
assertThat(table.get("T", "id")).isEqualTo("F T2"); assertThat(table.get("T", "id")).isEqualTo("F T2");
assertThat(table.get("T", "(")).isEqualTo("F T2"); assertThat(table.get("T", "(")).isEqualTo("F T2");
assertThat(table.get("T2", "+")).isEqualTo(grammar1.getEpsilonSymbol()); assertThat(table.get("T2", "+")).isEqualTo(Grammar.EPSILON_SYMBOL);
assertThat(table.get("T2", "*")).isEqualTo("* F T2"); assertThat(table.get("T2", "*")).isEqualTo("* F T2");
assertThat(table.get("T2", ")")).isEqualTo(grammar1.getEpsilonSymbol()); assertThat(table.get("T2", ")")).isEqualTo(Grammar.EPSILON_SYMBOL);
assertThat(table.get("T2", "$")).isEqualTo(grammar1.getEpsilonSymbol()); assertThat(table.get("T2", "$")).isEqualTo(Grammar.EPSILON_SYMBOL);
assertThat(table.get("F", "id")).isEqualTo("id"); assertThat(table.get("F", "id")).isEqualTo("id");
assertThat(table.get("F", "(")).isEqualTo("( E )"); assertThat(table.get("F", "(")).isEqualTo("( S )");
} }
} }

12
src/test/java/parser/grammar/GrammarTest.java
View File

@ -26,10 +26,7 @@ class GrammarTest {
final Path path = getPath("SimpleGrammar0.grammar"); final Path path = getPath("SimpleGrammar0.grammar");
final Grammar grammar = Grammar.fromFile(path); final Grammar grammar = Grammar.fromFile(path);
assert grammar != null;
assertThat(grammar.getEpsilonSymbol()).isEqualTo("epsilon");
assertThat(grammar.getStartSymbol()).isEqualTo("S");
assertThat(grammar.getTerminals()).containsOnly("a", "i", "t", "b"); assertThat(grammar.getTerminals()).containsOnly("a", "i", "t", "b");
assertThat(grammar.getNonterminals()).containsOnly("S", "E"); assertThat(grammar.getNonterminals()).containsOnly("S", "E");
assertThat(grammar.getRules()).containsOnly(new GrammarRule("S", "a"), assertThat(grammar.getRules()).containsOnly(new GrammarRule("S", "a"),
@ -42,14 +39,11 @@ class GrammarTest {
final Path path = getPath("SimpleGrammar1.grammar"); final Path path = getPath("SimpleGrammar1.grammar");
final Grammar grammar = Grammar.fromFile(path); final Grammar grammar = Grammar.fromFile(path);
assert grammar != null;
assertThat(grammar.getEpsilonSymbol()).isEqualTo("epsilon");
assertThat(grammar.getStartSymbol()).isEqualTo("E");
assertThat(grammar.getTerminals()).containsOnly("id", "+", "*", "(", ")"); assertThat(grammar.getTerminals()).containsOnly("id", "+", "*", "(", ")");
assertThat(grammar.getNonterminals()).containsOnly("E", "E2", "T", "T2", "F"); assertThat(grammar.getNonterminals()).containsOnly("S", "E2", "T", "T2", "F");
assertThat(grammar.getRules()).contains(new GrammarRule("E", "T", "E2"), assertThat(grammar.getRules()).contains(new GrammarRule("S", "T", "E2"),
new GrammarRule("E2", "+", "T", "E2"), new GrammarRule("E2", "+", "T", "E2"),
new GrammarRule("E2", "epsilon")); new GrammarRule("E2", Grammar.EPSILON_SYMBOL));
} }
} }

9
src/test/resources/exampleGrammars/Grammar.grammar
View File

@ -1,7 +1,4 @@
START: s // NTERM, TERM are reserved
EPS: eps
// START, EPS, NTERM, TERM are reserved
// Some Grammar-Symbols have to be named this way: // Some Grammar-Symbols have to be named this way:
// assignment, declaration (for TypeTable creation) // assignment, declaration (for TypeTable creation)
@ -11,7 +8,7 @@ EPS: eps
// Nonterminals: // Nonterminals:
NTERM: val type NTERM: val type
NTERM: op unary arith_op logic_op compare_op NTERM: op unary arith_op logic_op compare_op
NTERM: s class_cnt block_cnt NTERM: S class_cnt block_cnt
NTERM: statement stmt print NTERM: statement stmt print
NTERM: declaration assignment NTERM: declaration assignment
NTERM: par_expr expr expr_2 expr_f NTERM: par_expr expr expr_2 expr_f
@ -64,7 +61,7 @@ compare_op[promote] -> LESS | LESS_EQUAL | GREATER | GREATER_EQUAL | EQUAL | NOT
// ------------------------------------------------------------------------------------------------- // -------------------------------------------------------------------------------------------------
// START -> class IDENTIFIER { class_cnt } // START -> class IDENTIFIER { class_cnt }
s[promote] -> CLASS IDENTIFIER L_BRACE class_cnt R_BRACE | eps S[promote] -> CLASS IDENTIFIER L_BRACE class_cnt R_BRACE | eps
// class_cnt -> public static void main(String[] args) { block_cnt } // class_cnt -> public static void main(String[] args) { block_cnt }
class_cnt[promote delifempty] -> PUBLIC STATIC VOID_TYPE IDENTIFIER_MAIN L_PAREN STRING_TYPE L_BRACKET R_BRACKET IDENTIFIER R_PAREN L_BRACE block_cnt R_BRACE | eps class_cnt[promote delifempty] -> PUBLIC STATIC VOID_TYPE IDENTIFIER_MAIN L_PAREN STRING_TYPE L_BRACKET R_BRACKET IDENTIFIER R_PAREN L_BRACE block_cnt R_BRACE | eps

2
src/test/resources/exampleGrammars/SimpleGrammar0.grammar
View File

@ -1,5 +1,3 @@
START: S
EPS: epsilon
TERM: a i t b TERM: a i t b
NTERM: S E NTERM: S E

12
src/test/resources/exampleGrammars/SimpleGrammar1.grammar
View File

@ -1,13 +1,11 @@
START: E
EPS: epsilon
TERM: id + * ( ) TERM: id + * ( )
NTERM: E E2 T T2 F NTERM: S E2 T T2 F
// Leerzeilen sind egal // Leerzeilen sind egal
// Man kann Line-Comments schreiben und Produktionen verodern // Man kann Line-Comments schreiben und Produktionen verodern
E -> T E2 S -> T E2
E2 -> + T E2 | EPS E2 -> + T E2 | eps
T -> F T2 T -> F T2
T2 -> * F T2 | EPS T2 -> * F T2 | eps
F -> ( E ) | id F -> ( S ) | id

9
stups.grammar
View File

@ -1,7 +1,4 @@
START: s // NTERM, TERM are reserved
EPS: eps
// START, EPS, NTERM, TERM are reserved
// Some Grammar-Symbols have to be named this way: // Some Grammar-Symbols have to be named this way:
// assignment, declaration (for TypeTable creation) // assignment, declaration (for TypeTable creation)
@ -11,7 +8,7 @@ EPS: eps
// Nonterminals: // Nonterminals:
NTERM: val type NTERM: val type
NTERM: op unary arith_op logic_op compare_op NTERM: op unary arith_op logic_op compare_op
NTERM: s class_cnt block_cnt NTERM: S class_cnt block_cnt
NTERM: statement stmt print NTERM: statement stmt print
NTERM: declaration assignment NTERM: declaration assignment
NTERM: par_expr expr expr_2 expr_f NTERM: par_expr expr expr_2 expr_f
@ -64,7 +61,7 @@ compare_op[promote] -> LESS | LESS_EQUAL | GREATER | GREATER_EQUAL | EQUAL | NOT
// ------------------------------------------------------------------------------------------------- // -------------------------------------------------------------------------------------------------
// START -> class IDENTIFIER { class_cnt } // START -> class IDENTIFIER { class_cnt }
s[promote] -> CLASS IDENTIFIER L_BRACE class_cnt R_BRACE | eps S[promote] -> CLASS IDENTIFIER L_BRACE class_cnt R_BRACE | eps
// class_cnt -> public static void main(String[] args) { block_cnt } // class_cnt -> public static void main(String[] args) { block_cnt }
class_cnt[promote delifempty] -> PUBLIC STATIC VOID_TYPE IDENTIFIER_MAIN L_PAREN STRING_TYPE L_BRACKET R_BRACKET IDENTIFIER R_PAREN L_BRACE block_cnt R_BRACE | eps class_cnt[promote delifempty] -> PUBLIC STATIC VOID_TYPE IDENTIFIER_MAIN L_PAREN STRING_TYPE L_BRACKET R_BRACKET IDENTIFIER R_PAREN L_BRACE block_cnt R_BRACE | eps