Question

I am trying to use JSP on server-side to perform a variable number of queries and output the result of all of them as a single block of JSON data for an ExtJS line chart.

The reason the number of queries is variable is because each one represent a different series (a different line) on the line chart, and the number of series is different depending on the line chart that the user selects.

I am using hibernate and my persistence class returns each query data as a: List<Map<String, Object>> (each Map represents one row).

There will always be at least one series (one line on the graph, one query to execute), so the way I was thinking of setting this up is as follows:

1) Have the initial query run and get the first series

2) Run another query to check for any other series that should be on the graph

3) For each “other” series found in the second query run a query that gets the data for that series (same number of rows) and then merge that data into the first List<Map<String, Object>> that was returned in #1 as another column. The query is set-up to order it properly it just needs to be merged at the same index level.

4) Output that List as JSON.

My problem is with #3, I am not sure how to go about the merging the data.

Here’s what I have so far:

GenericSelectCommand graphData = new GenericSelectCommand(graphDataQuery);
GenericSelectCommand backSeriesData = new GenericSelectCommand(backSeriesQuery);

List<Map<String, Object>> graphDataList;
List<Map<String, Object>> backSeriesList;

try
{
    Persistor myPersistor = new Persistor();

    // 1) GET THE INITIAL LINE CHART SERIES
    myPersistor.executeTransact(graphData);
    graphDataList = graphData.getRows();

    // 2) LOOK FOR ANY ADDITIONAL SERIES THAT SHOULD BE ON THE LINE CHART
    myPersistor.executeTransact(backSeriesData);
    backSeriesList = backSeriesData.getRows();

    // 3) FOR EACH ADDITIONAL SERIES FOUND, RUN A QUERY AND APPEND THE DATA TO THE INITIAL LINE CHART SERIES (graphDataList)
    for (int i = 0; i < backSeriesList.size(); i++)
    {
        Map<String, Object> backSeriesBean = backSeriesList.get(i);

        // THIS QUERY RETURNS ONE COLUMN OF INT VALUES (THE LINE CHART DATA) WITH THE EXACT SAME NUMBER OF ROWS AS THE INITIAL LINE CHART SERIES (graphDataList)
        String backDataQuery = "exec runQuery 'getBackData', '" + backSeriesBean.get("series_id") + "'";

        GenericSelectCommand backData = new GenericSelectCommand(backDataQuery);
        myPersistor.executeTransact(backData);
        List<Map<String, Object>> backDataList = backData.getRows();

        // FOR EACH RECORD IN THE BACK DATA (Map<String, Object>)
        for (int i = 0; i < backDataList.size(); i++)
        {
            Map<String, Object> backDataBean = backDataList.get(i);
            // HOW DO I ADD IT TO THE RECORD AT THE SAME INDEX LEVEL IN graphDataList (List<Map<String, Object>>)
        }

    }


}
catch (Throwable e)
{
    System.err.println("Error: ");
    System.err.println(e.getCause());
}
finally
{
    myPersistor.closeSession();
}

// 4) RETURN THE DATA AS JSON NOW THAT IT IS MERGED
for (int i = 0; i < graphDataList.size(); i++)
{
    Map<String, Object> graphDataBean = graphDataList.get(i);
    out.println(/*JSON FORMAT + graphDataBean.get('data') + JSON FORMAT*/)
}

SOLUTION:

GenericSelectCommand graphData = new GenericSelectCommand(graphDataQuery);
GenericSelectCommand backSeries = new GenericSelectCommand(backSeriesQuery);

List<Map<String, Object>> graphDataList = Collections.emptyList();
List<Map<String, Object>> backSeriesList = Collections.emptyList();
List backDataListArray = new ArrayList();

try
{

    // GET THE INITIAL LINE CHART SERIES
    Persistor.instance().executeTransact(graphData);
    graphDataList = graphData.getRows();

    // LOOK FOR ANY ADDITIONAL SERIES THAT SHOULD BE ON THE LINE CHART
    Persistor.instance().executeTransact(backSeries);
    backSeriesList = backSeries.getRows();

    // FOR EACH ADDITIONAL SERIES FOUND, RUN THE QUERY AND ADD IT TO backDataListArray
    for (int i = 0; i < backSeriesList.size(); i++)
    {
        Map<String, Object> backSeriesBean = backSeriesList.get(i);

        String backDataQuery = "exec runQuery 'getBackData', " + backSeriesBean.get("series_id");
        GenericSelectCommand backData = new GenericSelectCommand(backDataQuery);
        Persistor.instance().executeTransact(backData);
        List<Map<String, Object>> backDataList = backData.getRows();
        backDataListArray.add(backDataList);
    }
}
catch (Throwable e)
{
    System.err.println("Error: ");
    System.err.println(e.getCause());
}
finally
{
    Persistor.instance().closeSession();
}

// FOR EACH RECORD IN THE ORIGINAL QUERY, WRITE THE JSON STRING
for (int i = 0; i < graphDataList.size(); i++)
{

    StringBuilder backDataString = new StringBuilder();

    // BUILD THE BACK DATA STRING (IF THERE IS ANY)
    for (int j = 0; j < backDataListArray.size(); j++)
    {
        List<Map<String, Object>> backDataList = (List<Map<String, Object>>) backDataListArray.get(j);
        Map<String, Object> backDataBean = backDataList.get(i);
        Map<String, Object> backSeriesBean = backSeriesList.get(j);

        backDataString.append(backSeriesBean.get("the_series") + ": " + backDataBean.get("the_count") + ", ");
    }

    Map<String, Object> graphDataBean = graphDataList.get(i);

    out.println("{the_quota: " + graphDataBean.get("the_quota") + ", " + "count_pt_year: " + graphDataBean.get("count_pt_year") + ", " + backDataString + "date_string: '" + graphDataBean.get("date_string") + "'}" + (i + 1 == graphDataList.size() ? "" : "," ));
}

Answer 1

I would not merge the lists. I would just create an outer list for each query and then go through the outer list and return each series list. You can just create the outer list as:

List outerList = new ArrayList();

I would not worry about specifying the types for the outer list as it just makes it more complicated for little benefit.