Examples with DefaultRow org.knime.core.data.def.DefaultRow used on opensource projects

Example 86 with DefaultRow

use of org.knime.core.data.def.DefaultRow in project knime-core by knime.

the class ColumnToGrid2NodeModel method execute.

/**
 * {@inheritDoc}
 */
@Override
protected BufferedDataTable[] execute(final BufferedDataTable[] inData, final ExecutionContext exec) throws Exception {
    String groupColumn = m_configuration.getGroupColumn();
    final ExecutionMonitor mainExec;
    final BufferedDataTable inputTable;
    if (groupColumn != null) {
        exec.setMessage("Sorting input table");
        BufferedDataTable in = inData[0];
        ExecutionContext sortExec = exec.createSubExecutionContext(0.5);
        ColumnRearranger sortFilterRearranger = new ColumnRearranger(in.getDataTableSpec());
        String[] relevantCols = new String[m_included.length + 1];
        System.arraycopy(m_included, 0, relevantCols, 0, m_included.length);
        relevantCols[relevantCols.length - 1] = groupColumn;
        sortFilterRearranger.keepOnly(relevantCols);
        BufferedDataTable toBeSortedTable = exec.createColumnRearrangeTable(in, sortFilterRearranger, exec.createSubProgress(0.0));
        SortedTable sorter = new SortedTable(toBeSortedTable, Collections.singletonList(groupColumn), new boolean[] { true }, sortExec);
        inputTable = sorter.getBufferedDataTable();
        mainExec = exec.createSubProgress(0.5);
    } else {
        inputTable = inData[0];
        mainExec = exec;
    }
    exec.setMessage("Assembling output");
    DataTableSpec spec = inputTable.getDataTableSpec();
    DataTableSpec outSpec = createOutputSpec(spec);
    BufferedDataContainer cont = exec.createDataContainer(outSpec);
    int[] includeIndices = new int[m_included.length];
    for (int i = 0; i < m_included.length; i++) {
        int index = spec.findColumnIndex(m_included[i]);
        includeIndices[i] = index;
    }
    int gridCount = m_configuration.getColCount();
    final int cellCount;
    final int groupColIndex;
    if (groupColumn != null) {
        cellCount = includeIndices.length * gridCount + 1;
        groupColIndex = spec.findColumnIndex(groupColumn);
    } else {
        cellCount = includeIndices.length * gridCount;
        groupColIndex = -1;
    }
    final DataCell[] cells = new DataCell[cellCount];
    PushBackRowIterator it = new PushBackRowIterator(inputTable.iterator());
    long currentRow = 0;
    long totalRows = inputTable.size();
    long currentOutRow = 0;
    DataCell curGroupValue = null;
    while (it.hasNext()) {
        Arrays.fill(cells, DataType.getMissingCell());
        // assign group column (if enabled)
        if (groupColIndex >= 0) {
            DataRow row = it.next();
            curGroupValue = row.getCell(groupColIndex);
            cells[cells.length - 1] = curGroupValue;
            it.pushBack(row);
        }
        for (int grid = 0; grid < gridCount; grid++) {
            if (!it.hasNext()) {
                break;
            }
            DataRow inRow = it.next();
            DataCell groupValue = groupColIndex < 0 ? null : inRow.getCell(groupColIndex);
            if (ConvenienceMethods.areEqual(curGroupValue, groupValue)) {
                mainExec.setProgress(currentRow / (double) totalRows, "Processing row " + currentRow + "/" + totalRows + ": " + inRow.getKey());
                currentRow += 1;
                mainExec.checkCanceled();
                for (int i = 0; i < includeIndices.length; i++) {
                    cells[grid * includeIndices.length + i] = inRow.getCell(includeIndices[i]);
                }
            } else {
                // start new group, i.e. new row
                it.pushBack(inRow);
                break;
            }
        }
        RowKey key = RowKey.createRowKey(currentOutRow++);
        cont.addRowToTable(new DefaultRow(key, cells));
    }
    cont.close();
    return new BufferedDataTable[] { cont.getTable() };
}

Example 87 with DefaultRow

use of org.knime.core.data.def.DefaultRow in project knime-core by knime.

the class FileRowIterator method next.

/**
 * {@inheritDoc}
 */
@Override
public DataRow next() {
    int rowLength = m_tableSpec.getNumColumns();
    int colsToRead = m_skipColumns.length;
    assert rowLength <= colsToRead;
    String token = null;
    boolean isMissingCell;
    String rowHeader;
    DataCell[] row = new DataCell[rowLength];
    // lines (if we are supposed to).
    if (!hasNext()) {
        throw new NoSuchElementException("The row iterator proceeded beyond the last line of '" + m_frSettings.getDataFileLocation().toString() + "'.");
    }
    // counts the columns (tokens) read from the file
    int readCols = 0;
    // counts the number of columns we've created (excl. skipped columns)
    int createdCols = 0;
    // This will also read it from file, if supposed to.
    try {
        rowHeader = createRowHeader(m_rowNumber - 1);
    } catch (TokenizerException fte) {
        throw prepareForException(fte.getMessage() + " (line: " + m_tokenizer.getLineNumber() + " source: '" + m_frSettings.getDataFileLocation() + "')", m_tokenizer.getLineNumber(), "ERR", row);
    }
    // we made sure before that there is at least one token in the stream
    assert rowHeader != null;
    // if the last token ended with the delimiter (and not a LF)
    boolean lastTokenWasDelimited = false;
    // Now, read the columns until we have enough or see a row delimiter
    while (readCols < colsToRead) {
        try {
            token = m_tokenizer.nextToken();
        } catch (TokenizerException fte) {
            throw prepareForException(fte.getMessage() + " (line: " + m_tokenizer.getLineNumber() + " (" + rowHeader + ") source: '" + m_frSettings.getDataFileLocation() + "')", m_tokenizer.getLineNumber(), rowHeader, row);
        }
        if (token != null) {
            // remember the delimiter of the last token before the EOF
            lastTokenWasDelimited = m_tokenizer.lastTokenWasDelimited();
        }
        // row delims are returned as token
        if ((token == null) || m_frSettings.isRowDelimiter(token, m_tokenizer.lastTokenWasQuoted())) {
            // line ended early.
            m_tokenizer.pushBack();
            // we need the row delim in the file, for after the loop
            break;
        }
        // column delimiters).
        if (token.equals("") && (!m_tokenizer.lastTokenWasQuoted())) {
            isMissingCell = true;
        } else if (token.equals(m_frSettings.getMissingValueOfColumn(readCols))) {
            // equals(null) if it was not specified - which is fine.
            isMissingCell = true;
        } else {
            isMissingCell = false;
        }
        if (!m_skipColumns[readCols]) {
            DataColumnSpec cSpec = m_tableSpec.getColumnSpec(createdCols);
            // now get that new cell
            // (it throws an exception at us if it couldn't)
            row[createdCols] = createNewDataCellOfType(cSpec.getType(), token, isMissingCell, m_frSettings.getFormatParameterForColumn(readCols).orElse(null), rowHeader, row);
            createdCols++;
        }
        readCols++;
    }
    // but only if the last token was actually delimited (with a swallowed delimiter - not LF)
    if (token == null && readCols == colsToRead - 1 && lastTokenWasDelimited) {
        if (!m_skipColumns[readCols]) {
            row[createdCols++] = DataType.getMissingCell();
        }
        // we consumed this last delimiter:
        lastTokenWasDelimited = false;
    }
    int lineNr = m_tokenizer.getLineNumber();
    if ((lineNr > 0) && (token != null) && (token.equals("\n"))) {
        lineNr--;
    }
    // puke and die - unless we are told otherwise
    if (m_frSettings.getSupportShortLines()) {
        // pad the row with missing values
        while (createdCols < rowLength) {
            row[createdCols++] = DataType.getMissingCell();
        }
    } else {
        if (createdCols < rowLength) {
            FileReaderException ex = prepareForException("Too few data elements " + "(line: " + lineNr + " (" + rowHeader + "), source: '" + m_frSettings.getDataFileLocation() + "')", lineNr, rowHeader, row);
            if (m_frSettings.getColumnNumDeterminingLineNumber() >= 0) {
                ex.setDetailsMessage("The number of columns was " + "determined by the entries above line no." + m_frSettings.getColumnNumDeterminingLineNumber());
            }
            throw ex;
        }
    }
    token = m_tokenizer.nextToken();
    if (!m_frSettings.isRowDelimiter(token, m_tokenizer.lastTokenWasQuoted())) {
        // flag for real data tokens
        lastTokenWasDelimited = m_tokenizer.lastTokenWasDelimited();
    }
    // eat all empty tokens til the end of the row, if we're supposed to
    if (m_frSettings.ignoreEmptyTokensAtEndOfRow()) {
        lastTokenWasDelimited = false;
        while (!m_frSettings.isRowDelimiter(token, m_tokenizer.lastTokenWasQuoted()) && token.equals("") && (!m_tokenizer.lastTokenWasQuoted())) {
            try {
                token = m_tokenizer.nextToken();
            } catch (TokenizerException fte) {
                throw prepareForException(fte.getMessage() + "(line: " + lineNr + " (" + rowHeader + "), source: '" + m_frSettings.getDataFileLocation() + "')", lineNr, rowHeader, row);
            }
        }
    }
    // data items in the file than we needed for one row: barf and die.
    if (!m_frSettings.isRowDelimiter(token, m_tokenizer.lastTokenWasQuoted()) || lastTokenWasDelimited) {
        FileReaderException ex = prepareForException("Too many data elements " + "(line: " + lineNr + " (" + rowHeader + "), source: '" + m_frSettings.getDataFileLocation() + "')", lineNr, rowHeader, row);
        if (m_frSettings.getColumnNumDeterminingLineNumber() >= 0) {
            ex.setDetailsMessage("The number of columns was " + "determined by line no." + m_frSettings.getColumnNumDeterminingLineNumber());
        }
        throw ex;
    }
    m_rowNumber++;
    // report progress
    // only if an execution context exists an if the underlying
    // URL is a file whose size can be determined
    double readBytes = m_source.getNumberOfBytesRead();
    if (m_exec != null && m_source.getFileSize() > 0 && readBytes / PROGRESS_JUNK_SIZE > m_lastReport) {
        // assert readBytes <= m_frSettings.getDataFileSize();
        m_exec.setProgress(readBytes / m_source.getFileSize());
        m_lastReport++;
    }
    return new DefaultRow(rowHeader, row);
}

Example 88 with DefaultRow

use of org.knime.core.data.def.DefaultRow in project knime-core by knime.

the class FileRowIterator method prepareForException.

/*
     * !!!!!!!!!! Creates the exception object (storing the last read items in
     * the row of the exception), sets the global "exception thrown" flag, and
     * closes the input stream. !!!!!!!!!!
     */
private FileReaderException prepareForException(final String msg, final int lineNumber, final String rowHeader, final DataCell[] cellsRead) {
    /*
         * indicate we have thrown (actually will throw...) an exception, and
         * close the stream as we will not read anymore from the stream after
         * the exception.
         */
    m_exceptionThrown = true;
    m_tokenizer.closeSourceStream();
    DataCell[] errCells = new DataCell[cellsRead.length];
    System.arraycopy(cellsRead, 0, errCells, 0, errCells.length);
    for (int c = 0; c < errCells.length; c++) {
        if (errCells[c] == null) {
            errCells[c] = DataType.getMissingCell();
        }
    }
    String errRowHeader = "ERROR_ROW (" + rowHeader.toString() + ")";
    DataRow errRow = new DefaultRow(errRowHeader, errCells);
    return new FileReaderException(msg, errRow, lineNumber);
}

Example 89 with DefaultRow

use of org.knime.core.data.def.DefaultRow in project knime-core by knime.

the class ListFiles method addLocationToContainer.

/**
 * Adds a File to the table.
 *
 * @param file
 */
private void addLocationToContainer(final URL url) throws UnsupportedEncodingException, URISyntaxException {
    DataCell[] row = new DataCell[2];
    if ("file".equalsIgnoreCase(url.getProtocol())) {
        row[0] = new StringCell(Paths.get(url.toURI()).toString());
    } else {
        row[0] = new MissingCell("URL is remote and does not have a local location");
    }
    row[1] = new StringCell(url.toString());
    m_dc.addRowToTable(new DefaultRow(RowKey.createRowKey(m_currentRowID), row));
    m_currentRowID++;
}

Example 90 with DefaultRow

use of org.knime.core.data.def.DefaultRow in project knime-core by knime.

the class FixedWidthRowIterator method next.

/**
 * {@inheritDoc}
 */
@Override
public DataRow next() {
    int rowLength = m_tableSpec.getNumColumns();
    int createdCols = 0;
    String token = null;
    String rowHeader;
    if (!hasNext()) {
        throw new NoSuchElementException("The row iterator proceeded beyond the last line of '" + m_nodeSettings.getFileLocation().toString() + "'.");
    }
    DataCell[] row;
    if (!m_nodeSettings.getHasRowHeader()) {
        rowHeader = "Row" + m_lineNumber++;
        row = new DataCell[rowLength];
    } else {
        rowHeader = m_tokenizer.nextToken();
        row = new DataCell[rowLength];
    }
    DataColumnSpec cSpec = null;
    while (createdCols < rowLength) {
        m_dataCellFactory.setMissingValuePattern(m_missingValuePatterns[createdCols]);
        m_dataCellFactory.setFormatParameter(m_formatParameters[createdCols]);
        token = m_tokenizer.nextToken();
        if (!m_tokenizer.getReachedEndOfLine()) {
            cSpec = m_tableSpec.getColumnSpec(createdCols);
            DataCell result = m_dataCellFactory.createDataCellOfType(cSpec.getType(), token);
            if (result != null) {
                row[createdCols] = result;
            } else {
                // something went wrong during cell creation.
                // figure out which column we were trying to read
                int errCol = 0;
                while (errCol < row.length && row[errCol] != null) {
                    errCol++;
                }
                // create an error message
                String errorMsg = m_dataCellFactory.getErrorMessage();
                errorMsg += " In line " + m_tokenizer.getLineNumber() + " (" + rowHeader + ") at column #" + errCol + " ('" + m_tableSpec.getColumnSpec(errCol).getName() + "').";
                assert rowHeader != null;
                // wrong, and close the stream
                throw prepareForException(errorMsg, m_tokenizer.getLineNumber(), rowHeader, row);
            }
        } else {
            // no more characters in this line but we need more columns
            // just add missing cells
            row[createdCols] = new MissingCell(null);
        }
        createdCols++;
    }
    double readBytes = m_inputStream.getNumberOfBytesRead();
    if (m_exec != null && m_inputStream.getFileSize() > 0 && readBytes / PROGRESS_JUNK_SIZE > m_lastReport) {
        // assert readBytes <= m_frSettings.getDataFileSize();
        m_exec.setProgress(readBytes / m_inputStream.getFileSize());
        m_lastReport++;
    }
    return new DefaultRow(rowHeader, row);
}