public static String getCleanedUpYield(Tree inputTree) {
    Tree copyTree = inputTree.deepCopy();

    if (DEBUG) System.err.println(copyTree.toString());

    String res = copyTree.yield().toString();
    if (res.length() > 1) {
      res = res.substring(0, 1).toUpperCase() + res.substring(1);
    }

    // (ROOT (S (NP (NNP Jaguar) (NNS shares)) (VP (VBD skyrocketed) (NP (NN yesterday)) (PP (IN
    // after) (NP (NP (NNP Mr.) (NNP Ridley) (POS 's)) (NN announcement)))) (. .)))

    res = res.replaceAll("\\s([\\.,!\\?\\-;:])", "$1");
    res = res.replaceAll("(\\$)\\s", "$1");
    res = res.replaceAll("can not", "cannot");
    res = res.replaceAll("\\s*-LRB-\\s*", " (");
    res = res.replaceAll("\\s*-RRB-\\s*", ") ");
    res = res.replaceAll("\\s*([\\.,?!])\\s*", "$1 ");
    res = res.replaceAll("\\s+''", "''");
    // res = res.replaceAll("\"", "");
    res = res.replaceAll("``\\s+", "``");
    res = res.replaceAll("\\-[LR]CB\\-", ""); // brackets, e.g., [sic]

    // remove extra spaces
    res = res.replaceAll("\\s\\s+", " ");
    res = res.trim();

    return res;
  }
 /** Outputs a tree using the output style requested */
 static void outputTree(PrintStream out, CoreMap sentence, List<Output> outputFormats) {
   Tree tree = sentence.get(SentimentCoreAnnotations.SentimentAnnotatedTree.class);
   for (Output output : outputFormats) {
     switch (output) {
       case PENNTREES:
         {
           Tree copy = tree.deepCopy();
           setSentimentLabels(copy);
           out.println(copy);
           break;
         }
       case VECTORS:
         {
           Tree copy = tree.deepCopy();
           setIndexLabels(copy, 0);
           out.println(copy);
           outputTreeVectors(out, tree, 0);
           break;
         }
       case ROOT:
         {
           out.println("  " + sentence.get(SentimentCoreAnnotations.SentimentClass.class));
           break;
         }
       case PROBABILITIES:
         {
           Tree copy = tree.deepCopy();
           setIndexLabels(copy, 0);
           out.println(copy);
           outputTreeScores(out, tree, 0);
           break;
         }
       default:
         throw new IllegalArgumentException("Unknown output format " + output);
     }
   }
 }
Beispiel #3
0
  /**
   * Usage: java edu.stanford.nlp.trees.tregex.tsurgeon.Tsurgeon [-s] -treeFile file-with-trees [-po
   * matching-pattern operation] operation-file-1 operation-file-2 ... operation-file-n
   *
   * <h4>Arguments:</h4>
   *
   * Each argument should be the name of a transformation file that contains a list of pattern and
   * transformation operation list pairs. That is, it is a sequence of pairs of a {@link
   * TregexPattern} pattern on one or more lines, then a blank line (empty or whitespace), then a
   * list of transformation operations one per line (as specified by <b>Legal operation syntax</b>
   * below) to apply when the pattern is matched, and then another blank line (empty or whitespace).
   * Note the need for blank lines: The code crashes if they are not present as separators (although
   * the blank line at the end of the file can be omitted). The script file can include comment
   * lines, either whole comment lines or trailing comments introduced by %, which extend to the end
   * of line. A needed percent mark can be escaped by a preceding backslash.
   *
   * <p>For example, if you want to excise an SBARQ node whenever it is the parent of an SQ node,
   * and relabel the SQ node to S, your transformation file would look like this:
   *
   * <blockquote>
   *
   * <code>
   *    SBARQ=n1 &lt; SQ=n2<br>
   *    <br>
   *    excise n1 n1<br>
   *    relabel n2 S
   * </code>
   *
   * </blockquote>
   *
   * <p>
   *
   * <h4>Options:</h4>
   *
   * <ul>
   *   <li><code>-treeFile &#60;filename&#62;</code> specify the name of the file that has the trees
   *       you want to transform.
   *   <li><code>-po &#60;matchPattern&#62; &#60;operation&#62;</code> Apply a single operation to
   *       every tree using the specified match pattern and the specified operation. Use this option
   *       when you want to quickly try the effect of one pattern/surgery combination, and are too
   *       lazy to write a transformation file.
   *   <li><code>-s</code> Print each output tree on one line (default is pretty-printing).
   *   <li><code>-m</code> For every tree that had a matching pattern, print "before" (prepended as
   *       "Operated on:") and "after" (prepended as "Result:"). Unoperated trees just pass through
   *       the transducer as usual.
   *   <li><code>-encoding X</code> Uses character set X for input and output of trees.
   *   <li><code>-macros &#60;filename&#62;</code> A file of macros to use on the tregex pattern.
   *       Macros should be one per line, with original and replacement separated by tabs.
   *   <li><code>-hf &lt;headfinder-class-name&gt;</code> use the specified {@link HeadFinder} class
   *       to determine headship relations.
   *   <li><code>-hfArg &lt;string&gt;</code> pass a string argument in to the {@link HeadFinder}
   *       class's constructor. <code>-hfArg</code> can be used multiple times to pass in multiple
   *       arguments.
   *   <li><code>-trf &lt;TreeReaderFactory-class-name&gt;</code> use the specified {@link
   *       TreeReaderFactory} class to read trees from files.
   * </ul>
   *
   * <h4>Legal operation syntax:</h4>
   *
   * <ul>
   *   <li><code>delete &#60;name&#62;</code> deletes the node and everything below it.
   *   <li><code>prune &#60;name&#62;</code> Like delete, but if, after the pruning, the parent has
   *       no children anymore, the parent is pruned too. Pruning continues to affect all ancestors
   *       until one is found with remaining children. This may result in a null tree.
   *   <li><code>excise &#60;name1&#62; &#60;name2&#62;</code> The name1 node should either dominate
   *       or be the same as the name2 node. This excises out everything from name1 to name2. All
   *       the children of name2 go into the parent of name1, where name1 was.
   *   <li><code>relabel &#60;name&#62; &#60;new-label&#62;</code> Relabels the node to have the new
   *       label. <br>
   *       There are three possible forms: <br>
   *       <code>relabel nodeX VP</code> - for changing a node label to an alphanumeric string <br>
   *       <code>relabel nodeX /''/</code> - for relabeling a node to something that isn't a valid
   *       identifier without quoting <br>
   *       <code>relabel nodeX /^VB(.*)$/verb\\/$1/</code> - for regular expression based
   *       relabeling. In this case, all matches of the regular expression against the node label
   *       are replaced with the replacement String. This has the semantics of Java/Perl's
   *       replaceAll: you may use capturing groups and put them in replacements with $n. For
   *       example, if the pattern is /foo/bar/ and the node matched is "foo", the replaceAll
   *       semantics result in "barbar". If the pattern is /^foo(.*)$/bar$1/ and node matched is
   *       "foofoo", relabel will result in "barfoo". <br>
   *       When using the regex replacement method, you can also use the sequences ={node} and
   *       %{var} in the replacement string to use captured nodes or variable strings in the
   *       replacement string. For example, if the Tregex pattern was "duck=bar" and the relabel is
   *       /foo/={bar}/, "foofoo" will be replaced with "duckduck". <br>
   *       To concatenate two nodes named in the tregex pattern, for example, you can use the
   *       pattern /^.*$/={foo}={bar}/. Note that the ^.*$ is necessary to make sure the regex
   *       pattern only matches and replaces once on the entire node name. <br>
   *       To get an "=" or a "%" in the replacement, using \ escaping. Also, as in the example you
   *       can escape a slash in the middle of the second and third forms with \\/ and \\\\. <br>
   *   <li><code>insert &#60;name&#62; &#60;position&#62;</code> or <code>
   *       insert &lt;tree&gt; &#60;position&#62;</code> inserts the named node or tree into the
   *       position specified.
   *   <li><code>move &#60;name&#62; &#60;position&#62;</code> moves the named node into the
   *       specified position.
   *       <p>Right now the only ways to specify position are:
   *       <p><code>$+ &#60;name&#62;</code> the left sister of the named node<br>
   *       <code>$- &#60;name&#62;</code> the right sister of the named node<br>
   *       <code>&gt;i &#60;name&#62;</code> the i_th daughter of the named node<br>
   *       <code>&gt;-i &#60;name&#62;</code> the i_th daughter, counting from the right, of the
   *       named node.
   *   <li><code>replace &#60;name1&#62; &#60;name2&#62;</code> deletes name1 and inserts a copy of
   *       name2 in its place.
   *   <li><code>replace &#60;name&#62; &#60;tree&#62; &#60;tree2&#62;...</code> deletes name and
   *       inserts the new tree(s) in its place. If more than one replacement tree is given, each of
   *       the new subtrees will be added in order where the old tree was. Multiple subtrees at the
   *       root is an illegal operation and will throw an exception.
   *   <li>{@code createSubtree <new-label> <name1> [<name2>]} Create a subtree out of all the nodes
   *       from {@code <name1>} through {@code <name2>} and puts the new subtree where that span
   *       used to be. To limit the operation to just one node, elide {@code <name2>}.
   *   <li><code>adjoin &#60;auxiliary_tree&#62; &lt;name&gt;</code> Adjoins the specified auxiliary
   *       tree into the named node. The daughters of the target node will become the daughters of
   *       the foot of the auxiliary tree.
   *   <li><code>adjoinH &#60;auxiliary_tree&#62; &lt;name&gt;</code> Similar to adjoin, but
   *       preserves the target node and makes it the root of &lt;tree&gt;. (It is still accessible
   *       as <code>name</code>. The root of the auxiliary tree is ignored.)
   *   <li><code>adjoinF &#60;auxiliary_tree&#62; &lt;name&gt;</code> Similar to adjoin, but
   *       preserves the target node and makes it the foot of &lt;tree&gt;. (It is still accessible
   *       as <code>name</code>, and retains its status as parent of its children. The root of the
   *       auxiliary tree is ignored.)
   *   <li>
   *   <dt><code>coindex &#60;name1&#62; &#60;name2&#62; ... &#60;nameM&#62; </code> Puts a (Penn
   *       Treebank style) coindexation suffix of the form "-N" on each of nodes name_1 through
   *       name_m. The value of N will be automatically generated in reference to the existing
   *       coindexations in the tree, so that there is never an accidental clash of indices across
   *       things that are not meant to be coindexed.
   * </ul>
   *
   * <p>In the context of <code>adjoin</code>, <code>adjoinH</code>, and <code>adjoinF</code>, an
   * auxiliary tree is a tree in Penn Treebank format with <code>@</code> on exactly one of the
   * leaves denoting the foot of the tree. The operations which use the foot use the labeled node.
   * For example: <br>
   * Tsurgeon: <code>adjoin (FOO (BAR@)) foo</code> <br>
   * Tregex: <code>B=foo</code> <br>
   * Input: <code>(A (B 1 2))</code> Output: <code>(A (FOO (BAR 1 2)))</code>
   *
   * <p>Tsurgeon applies the same operation to the same tree for as long as the given tregex
   * operation matches. This means that infinite loops are very easy to cause. One common situation
   * where this comes up is with an insert operation will repeats infinitely many times unless you
   * add an expression to the tregex that matches against the inserted pattern. For example, this
   * pattern will infinite loop:
   *
   * <blockquote>
   *
   * <code>
   *   TregexPattern tregex = TregexPattern.compile("S=node &lt;&lt; NP"); <br>
   *   TsurgeonPattern tsurgeon = Tsurgeon.parseOperation("insert (NP foo) &gt;-1 node");
   * </code>
   *
   * </blockquote>
   *
   * This pattern, though, will terminate:
   *
   * <blockquote>
   *
   * <code>
   *   TregexPattern tregex = TregexPattern.compile("S=node &lt;&lt; NP !&lt;&lt; foo"); <br>
   *   TsurgeonPattern tsurgeon = Tsurgeon.parseOperation("insert (NP foo) &gt;-1 node");
   * </code>
   *
   * </blockquote>
   *
   * <p>Tsurgeon has (very) limited support for conditional statements. If a pattern is prefaced
   * with <code>if exists &lt;name&gt;</code>, the rest of the pattern will only execute if the
   * named node was found in the corresponding TregexMatcher.
   *
   * @param args a list of names of files each of which contains a single tregex matching pattern
   *     plus a list, one per line, of transformation operations to apply to the matched pattern.
   * @throws Exception If an I/O or pattern syntax error
   */
  public static void main(String[] args) throws Exception {
    String headFinderClassName = null;
    String headFinderOption = "-hf";
    String[] headFinderArgs = null;
    String headFinderArgOption = "-hfArg";
    String encoding = "UTF-8";
    String encodingOption = "-encoding";
    if (args.length == 0) {
      System.err.println(
          "Usage: java edu.stanford.nlp.trees.tregex.tsurgeon.Tsurgeon [-s] -treeFile <file-with-trees> [-po <matching-pattern> <operation>] <operation-file-1> <operation-file-2> ... <operation-file-n>");
      System.exit(0);
    }
    String treePrintFormats;
    String singleLineOption = "-s";
    String verboseOption = "-v";
    String matchedOption =
        "-m"; // if set, then print original form of trees that are matched & thus operated on
    String patternOperationOption = "-po";
    String treeFileOption = "-treeFile";
    String trfOption = "-trf";
    String macroOption = "-macros";
    String macroFilename = "";
    Map<String, Integer> flagMap = Generics.newHashMap();
    flagMap.put(patternOperationOption, 2);
    flagMap.put(treeFileOption, 1);
    flagMap.put(trfOption, 1);
    flagMap.put(singleLineOption, 0);
    flagMap.put(encodingOption, 1);
    flagMap.put(headFinderOption, 1);
    flagMap.put(macroOption, 1);
    Map<String, String[]> argsMap = StringUtils.argsToMap(args, flagMap);
    args = argsMap.get(null);

    if (argsMap.containsKey(headFinderOption))
      headFinderClassName = argsMap.get(headFinderOption)[0];
    if (argsMap.containsKey(headFinderArgOption)) headFinderArgs = argsMap.get(headFinderArgOption);
    if (argsMap.containsKey(verboseOption)) verbose = true;
    if (argsMap.containsKey(singleLineOption)) treePrintFormats = "oneline,";
    else treePrintFormats = "penn,";
    if (argsMap.containsKey(encodingOption)) encoding = argsMap.get(encodingOption)[0];
    if (argsMap.containsKey(macroOption)) macroFilename = argsMap.get(macroOption)[0];

    TreePrint tp = new TreePrint(treePrintFormats, new PennTreebankLanguagePack());
    PrintWriter pwOut = new PrintWriter(new OutputStreamWriter(System.out, encoding), true);

    TreeReaderFactory trf;
    if (argsMap.containsKey(trfOption)) {
      String trfClass = argsMap.get(trfOption)[0];
      trf = ReflectionLoading.loadByReflection(trfClass);
    } else {
      trf = new TregexPattern.TRegexTreeReaderFactory();
    }

    Treebank trees = new DiskTreebank(trf, encoding);
    if (argsMap.containsKey(treeFileOption)) {
      trees.loadPath(argsMap.get(treeFileOption)[0]);
    }
    List<Pair<TregexPattern, TsurgeonPattern>> ops =
        new ArrayList<Pair<TregexPattern, TsurgeonPattern>>();

    TregexPatternCompiler compiler;
    if (headFinderClassName == null) {
      compiler = new TregexPatternCompiler();
    } else {
      HeadFinder hf;
      if (headFinderArgs == null) {
        hf = ReflectionLoading.loadByReflection(headFinderClassName);
      } else {
        hf = ReflectionLoading.loadByReflection(headFinderClassName, (Object[]) headFinderArgs);
      }
      compiler = new TregexPatternCompiler(hf);
    }
    Macros.addAllMacros(compiler, macroFilename, encoding);
    if (argsMap.containsKey(patternOperationOption)) {
      TregexPattern matchPattern = compiler.compile(argsMap.get(patternOperationOption)[0]);
      TsurgeonPattern p = parseOperation(argsMap.get(patternOperationOption)[1]);
      ops.add(new Pair<TregexPattern, TsurgeonPattern>(matchPattern, p));
    } else {
      for (String arg : args) {
        List<Pair<TregexPattern, TsurgeonPattern>> pairs =
            getOperationsFromFile(arg, encoding, compiler);
        for (Pair<TregexPattern, TsurgeonPattern> pair : pairs) {
          if (verbose) {
            System.err.println(pair.second());
          }
          ops.add(pair);
        }
      }
    }

    for (Tree t : trees) {
      Tree original = t.deepCopy();
      Tree result = processPatternsOnTree(ops, t);
      if (argsMap.containsKey(matchedOption) && matchedOnTree) {
        pwOut.println("Operated on: ");
        displayTree(original, tp, pwOut);
        pwOut.println("Result: ");
      }
      displayTree(result, tp, pwOut);
    }
  }