docreader stub

libcorpus2/CMakeLists.txt

@@ -41,6 +41,7 @@ SET(libcorpus2_STAT_SRC
 	ann/iob.cpp
 	ann/view.cpp
 	chunk.cpp
+	document.cpp
 	exception.cpp
 	lexeme.cpp
 	sentence.cpp
@@ -53,6 +54,7 @@ SET(libcorpus2_STAT_SRC
 	tokenmetadata.cpp
 	io/cclreader.cpp
 	io/cclwriter.cpp
+	io/docreader.cpp
 	io/helpers.cpp
 	io/fastxces.cpp
 	io/iob-chan.cpp

libcorpus2/document.cpp

+/*
+	Copyright (C) 2010 Tomasz Śniatowski, Adam Radziszewski
+	Part of the libcorpus2 project
+
+	This program is free software; you can redistribute it and/or modify it
+under the terms of the GNU General Public License as published by the Free
+Software Foundation; either version 3 of the License, or (at your option)
+any later version.
+
+	This program is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+or FITNESS FOR A PARTICULAR PURPOSE.
+
+	See the LICENSE and COPYING files for more details.
+*/
+
+#include <libcorpus2/document.h>
+#include <boost/make_shared.hpp>
+
+namespace Corpus2 {
+
+Document::Document()
+	: paragraphs_()
+{
+}
+
+Document::~Document()
+{
+}
+
+} /* end ns Corpus2 */

libcorpus2/document.h

+/*
+	Copyright (C) 2010 Tomasz Śniatowski, Adam Radziszewski
+	Part of the libcorpus2 project
+
+	This program is free software; you can redistribute it and/or modify it
+under the terms of the GNU General Public License as published by the Free
+Software Foundation; either version 3 of the License, or (at your option)
+any later version.
+
+	This program is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+or FITNESS FOR A PARTICULAR PURPOSE.
+
+	See the LICENSE and COPYING files for more details.
+*/
+
+#ifndef LIBCORPUS2_DOCUMENT_H
+#define LIBCORPUS2_DOCUMENT_H
+
+#include <libcorpus2/chunk.h>
+#include <boost/shared_ptr.hpp>
+
+namespace Corpus2 {
+
+/**
+ * A whole document, consisting of consecutive paragraphs ("chunks"), being
+ * sequences of sentences.
+ * Usage of this class assumes that a whole document is read into memory before
+ * any further takes place.
+ */
+class Document
+{
+public:
+	Document();
+	~Document();
+
+	void add_paragraph(const boost::shared_ptr<Chunk> para) {
+		paragraphs_.push_back(para);
+	}
+
+	const std::vector< boost::shared_ptr<Chunk> >& paragraphs() const {
+		return paragraphs_;
+	}
+
+protected:
+	std::vector< boost::shared_ptr<Chunk> > paragraphs_;
+};
+
+} /* end ns Corpus2 */
+
+#endif // LIBCORPUS2_DOCUMENT_H

libcorpus2/io/docreader.cpp

+/*
+	Copyright (C) 2010 Tomasz Śniatowski, Adam Radziszewski
+	Part of the libcorpus2 project
+
+	This program is free software; you can redistribute it and/or modify it
+under the terms of the GNU General Public License as published by the Free
+Software Foundation; either version 3 of the License, or (at your option)
+any later version.
+
+	This program is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+or FITNESS FOR A PARTICULAR PURPOSE.
+
+	See the LICENSE and COPYING files for more details.
+*/
+
+#include <libcorpus2/io/docreader.h>
+#include <boost/make_shared.hpp>
+
+namespace Corpus2 {
+
+} /* end ns Corpus2 */

libcorpus2/io/docreader.h

+/*
+	Copyright (C) 2010 Tomasz Śniatowski, Adam Radziszewski
+	Part of the libcorpus2 project
+
+	This program is free software; you can redistribute it and/or modify it
+under the terms of the GNU General Public License as published by the Free
+Software Foundation; either version 3 of the License, or (at your option)
+any later version.
+
+	This program is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+or FITNESS FOR A PARTICULAR PURPOSE.
+
+	See the LICENSE and COPYING files for more details.
+*/
+
+#ifndef LIBCORPUS2_DOCREADER_H
+#define LIBCORPUS2_DOCREADER_H
+
+#include <libcorpus2/io/reader.h>
+
+namespace Corpus2 {
+
+/**
+ * A reader for whole documents. Note that a whole document is read into memory
+ * before any processing may take place.
+ */
+class DocumentReader {
+public:
+	/**
+	 * Reads a whole document, using the two given path: the morphosyntax and
+	 * chunk-style annotations are read from annot_path, while relations
+	 * between chunk-style annotations are read from rela_path.
+	 * Both path may in particular point to the same path.
+	 * TODO!
+	 */
+	DocumentReader(const std::string &annot_path,
+				   const std::string &rela_path,
+				   const std::string &rdr_class_id = "ccl");
+};
+} /* end ns Corpus2 */
+
+#endif // LIBCORPUS2_DOCREADER_H