How can I use this query with the query method (SQLiteDatabase.query)? - java

The task I've been given is to write a query to return the number of orphaned rows. I have achieved this but another task is to then not use the rawQuery method to achieve the same result using the query method.
The issue is that I get java.lang.IllegalStateException: Invalid tables
The tables, there are 3 are
the parent table which has an _id column and a name column
the child table which has an _id column, a name column and a childtoparentlink column that is an integer that links to the parent table.
the friend table which has an _id column, a name column and a friendtochildlink column.
The SQL to create and to put rows into the tables, including some orphans is like
CREATE TABLE parent(_id INTEGER PRIMARY KEY,parentname TEXT);
CREATE TABLE child(_id INTEGER PRIMARY KEY,childname TEXT, childtoparentlink INTEGER);
CREATE TABLE friend(_id INTEGER PRIMARY KEY,friendname TEXT, friendtochildlink INTEGER);
INSERT INTO parent VALUES(null,'Parent A');
INSERT INTO parent VALUES(null,'Parent B');
INSERT INTO child VALUES(null,'Child A',1);
INSERT INTO child VALUES(null,'Child B',2);
INSERT INTO child VALUES(null,'Child X',10); -- orphan
INSERT INTO friend VALUES(null,'Friend A',1);
INSERT INTO friend VALUES(null,'Friend B',2);
INSERT INTO friend VALUES(null,'Friend X',100); -- orphan
The query that works and gives the right values when using rawQuery is
SELECT
(
SELECT count() FROM child
LEFT JOIN parent ON child.childtoparentlink = parent._id
WHERE parent.parentname IS NULL
) AS child_mismatches,
(
SELECT count() FROM friend
LEFT JOIN child ON friend.friendtochildlink = child._id
WHERE child.childname IS NULL
) AS friend_mismatches
I get two columns each with a value of 1 (as wanted).
My actual code is :-
public ArrayList<String> checkLinkIntegrity() {
ArrayList<String> return_value = new ArrayList<>();
String suffix = "_mismatches";
String child_result_cl = TB_CHILD + suffix;
String sq_child_mismatches = "(SELECT count() FROM " +
TB_CHILD +
" LEFT JOIN " + TB_PARENT +
" ON " + TB_CHILD + "." + CL_CHILDTOPARENTLINK + " = " +
TB_PARENT + "." + CL_PARENTID +
" WHERE " + TB_PARENT + "." + CL_PARENTNAME + " IS NULL)" +
" AS " + child_result_cl;
String friend_result_cl = TB_FRIEND + suffix;
String sq_friend_mismatches = "(SELECT count() FROM " +
TB_FRIEND +
" LEFT JOIN " + TB_CHILD +
" ON " + TB_FRIEND + "." + CL_FRIENDTOCHILDLINK + " = " +
TB_CHILD + "." + CL_CHILD_ID +
" WHERE " + TB_CHILD + "." + CL_CHILDNAME + " IS NULL)" +
" AS " + friend_result_cl;
String full_query = "SELECT " + sq_child_mismatches + "," + sq_friend_mismatches;
SQLiteDatabase db = this.getWritableDatabase();
Cursor csr;
Log.d("RAWQUERYSQL",full_query);
csr = db.rawQuery(full_query,null);
return_value.addAll(dumpCursorToStringArrayList(csr,"RAWQUERY"));
// Fails invalid table
csr = db.query(null,new String[]{sq_child_mismatches,sq_friend_mismatches},null,null,null,null,null);
return_value.addAll(dumpCursorToStringArrayList(csr,"SECONDTRY"));
csr.close();
return return_value;
}
and the dumpCursortoStringArrayList method is :-
private ArrayList<String> dumpCursorToStringArrayList(Cursor csr, String tablename) {
ArrayList<String> rv = new ArrayList<>();
int original_position = csr.getPosition();
csr.moveToPosition(-1);
rv.add("Table: " + tablename);
while (csr.moveToNext()) {
rv.add("\tRow # " + String.valueOf(csr.getPosition() + 1));
for (String column: csr.getColumnNames()) {
rv.add("\t\tColumn: " + column + "\tvalue is: \t" + csr.getString(csr.getColumnIndex(column)));
}
}
csr.moveToPosition(original_position);
return rv;
}
I get the same error if I try "" instead of null e.g.
If I use only the rawQuery I get
04-22 07:07:33.914 6271-6271/s.q001 I/RESULTS: Table: RAWQUERY
04-22 07:07:33.914 6271-6271/s.q001 I/RESULTS: Row # 1
04-22 07:07:33.914 6271-6271/s.q001 I/RESULTS: Column: child_mismatches value is: 1
04-22 07:07:33.914 6271-6271/s.q001 I/RESULTS: Column: friend_mismatches value is: 1
This is from using
ArrayList<String> results = DBOpenHelper.checkLinkIntegrity();
for (String s : results) {
Log.i("RESULTS",s);
}
How can I run the query with the query method instead of the rawQuery method, to get the better marks?


Your issue is that the query method expects a table as it then generates SQL as per
SELECT your_columns FROM the_table;
As there is no table it issues the Invalid Table exception.
You have to provide something that will satisfy the FROM clause, it cannot be nothing.
You can get around this in a few ways, which I guess is what the homework is trying to get you to ascertain/explore.
Fix 1
You could supply one of the tables that exist e.g. use
csr = db.query(null,new String[]{sq_child_mismatches,sq_friend_mismatches},null,null,null,null,null,"1");
Note the 8th parameter, this LIMITs the number of rows generated to one as there would be a row generated for each row in the fake table.
Fix 2
or as FROM can be a subquery (see diagram) you could use a subquery e.g. one that you have
So you could use :-
csr = db.query(
sq_child_mismatches, //<<<<<<<<<< the fake subquery
new String[]{
sq_child_mismatches,
sq_friend_mismatches
},
null ,null,null,null,null
);
In this case, as the fake subquery returns a single row/result, there is no need for the LIMIT clause.

Related

How to retrieve Real type data from specific cell in android sqlite?

In this table, I want to retrieve the value 55.2 in a variable.
This value is at row 5 (ID=5) and in the column 'Weight' of type REAL.
I can already get the desired row number which is stored in 'lastID' and I know that my data is in the column 'Weight'. So I have my X and my Y in the table.
I also know the sqlite command to retrieve the 55.2 in my cursor:
Cursor cursor2 = db.rawQuery("SELECT Weight FROM <MYTABLE> WHERE ID=" + lastID, null);
Double lastWeight = cursor2.getDouble(0); //This line is wrong, I need the help here!
But I can't get the 55.2 value I am looking for in my variable lastWeight from cursor2.
Any idea?
Addendum
Here the create table:
String CREATE_TABLE2 = "CREATE TABLE " + <MYTABLE> + " (" + UID2 + " INTEGER PRIMARY KEY AUTOINCREMENT, " +
COL_2 + " TEXT," + COL_3 + " TEXT," + COL_4 + " REAL," + COL_5 + " REAL);";
db.execSQL(CREATE_TABLE2);

After the execution of this line:
Cursor cursor2 = db.rawQuery("SELECT Weight FROM <MYTABLE> WHERE ID=" + lastID, null);
you get the results in cursor2.
A Cursor instance like cursor2 is used to loop through its rows and to do so you must first place its index at the 1st row by moveToFirst():
if (cursor2.moveToFirst()) {
Double lastWeight = cursor2.getDouble(0);
........................................
}
The if statement is necessary just in case the cursor does not contain any rows.

Android : Calculate sum and group by month (SQLite)

Is there are any option to select amount, group them by month and calculate sum. I tried to get total sum of each month and pass it to ArrayList.
Example of data:
Amount Date
230 04/03/19
500 05/03/19
400 04/04/19
600 06/04/19
100 04/03/19
... ...
My code structure
private String CREATE_BILLS_TABLE = "CREATE TABLE " + TABLE_BILLS + "("
+ COLUMN_BILL_ID + " INTEGER PRIMARY KEY AUTOINCREMENT,"
+ COLUMN_BILL_USER_ID + " INTEGER,"
+ COLUMN_DESCRIPTION + " TEXT,"
+ COLUMN_AMOUNT + " INTEGER,"
+ COLUMN_DATE_STRING + " TEXT,"
+ COLUMN_COMPANY_NAME + " TEXT,"
+ COLUMN_CATEGORY + " TEXT,"
+ " FOREIGN KEY ("+COLUMN_BILL_USER_ID+") REFERENCES "+TABLE_USER+"("+COLUMN_USER_ID+"));";
public ArrayList<Bills> getDateByUserID(int userID){
SQLiteDatabase db = this.getReadableDatabase();
// sorting orders
ArrayList<Bills> listBillsDates = new ArrayList<Bills>();
Cursor cursor = db.query(TABLE_BILLS, new String[] { COLUMN_BILL_ID,
COLUMN_BILL_USER_ID, COLUMN_DESCRIPTION, COLUMN_AMOUNT, COLUMN_DATE_STRING, COLUMN_COMPANY_NAME, COLUMN_CATEGORY}, COLUMN_BILL_USER_ID + "=?",
new String[] { String.valueOf(userID) }, COLUMN_DATE_STRING, null, null, null);
if (cursor.moveToFirst()) {
do {
Bills bills = new Bills();
bills.setAmount(cursor.getInt(cursor.getColumnIndex(COLUMN_AMOUNT)));
bills.setDateString(cursor.getString(cursor.getColumnIndex(COLUMN_DATE_STRING)));
// Adding record to list
listBillsDates.add(bills);
} while (cursor.moveToNext());
}
cursor.close();
db.close();
// return category list
return listBillsDates;
}

I believe that a query based upon :-
SELECT sum(COLUMN_AMOUNT) AS Monthly_Total,substr(COLUMN_DATE_STRING,4) AS Month_and_Year
FROM TABLE_BILLS
WHERE COLUMN_BILL_USER_ID = 1
GROUP BY substr(COLUMN_DATE_STRING,4)
ORDER BY substr(COLUMN_DATE_STRING,7,2)||substr(COLUMN_DATE_STRING,4,2)
;
Note that other columns values would be arbritary results and as such cannot really be relied upon (fine if the data is always the same). Hence they have not been included.
Will produce the results that you want :-
e.g.
Using the following, to test the SQL :-
DROP TABLE IF EXISTS TABLE_BILLS;
CREATE TABLE IF NOT EXISTS TABLE_BILLS (
COLUMN_BILL_ID INTEGER PRIMARY KEY AUTOINCREMENT,
COLUMN_BILL_USER_ID INTEGER,
COLUMN_DESCRIPTION TEXT,
COLUMN_AMOUNT INTEGER,
COLUMN_DATE_STRING TEXT,
COLUMN_COMPANY_NAME TEXT,
COLUMN_CATEGORY TEXT)
;
-- Add the Testing data
INSERT INTO TABLE_BILLS (
COLUMN_BILL_USER_ID, COLUMN_DESCRIPTION, COLUMN_AMOUNT, COLUMN_DATE_STRING, COLUMN_COMPANY_NAME,COLUMN_CATEGORY)
VALUES
(1,'blah',230,'04/03/19','cmpny','category')
,(1,'blah',500,'05/03/19','cmpny','category')
,(1,'blah',400,'04/04/19','cmpny','category')
,(1,'blah',600,'06/04/19','cmpny','category')
,(1,'blah',100,'04/03/19','cmpny','category')
-- Extra data for another id to check exclusion
,(2,'blah',230,'04/03/19','cmpny','category')
,(2,'blah',500,'05/03/19','cmpny','category')
,(2,'blah',400,'04/04/19','cmpny','category')
,(2,'blah',600,'06/04/19','cmpny','category')
,(2,'blah',100,'04/03/19','cmpny','category')
;
SELECT sum(COLUMN_AMOUNT) AS Monthly_Total,substr(COLUMN_DATE_STRING,4) AS Month_and_Year
FROM TABLE_BILLS
WHERE COLUMN_BILL_USER_ID = 1
GROUP BY substr(COLUMN_DATE_STRING,4)
ORDER BY substr(COLUMN_DATE_STRING,7,2)||substr(COLUMN_DATE_STRING,4,2)
;
Results id :-
The above can then be converted for use by the SQLiteDatabase query method. So your method could be something like :-
public ArrayList<Bills> getDateByUserID(int userID) {
SQLiteDatabase db = this.getReadableDatabase();
String tmpcol_monthly_total = "Monthly_Total";
String tmpcol_month_year = "Month_and_Year";
String[] columns = new String[]{
"sum(" + COLUMN_AMOUNT + ") AS " + tmpcol_monthly_total,
"substr(" + COLUMN_DATE_STRING + ",4) AS " + tmpcol_month_year
};
String whereclause = COLUMN_BILL_USER_ID + "=?";
String[] whereargs = new String[]{String.valueOf(userID)};
String groupbyclause = "substr(" + COLUMN_DATE_STRING + ",4)";
String orderbyclause = "substr(" + COLUMN_DATE_STRING + ",7,2)||substr(" + COLUMN_DATE_STRING + ",4,2)";
ArrayList<Bills> listBillsDates = new ArrayList<Bills>();
Cursor cursor = db.query(TABLE_BILLS, columns, whereclause,
whereargs, groupbyclause, null, orderbyclause, null);
if (cursor.moveToFirst()) {
do {
Bills bills = new Bills();
bills.setAmount(cursor.getInt(cursor.getColumnIndex(tmpcol_monthly_total)));
bills.setDateString(cursor.getString(cursor.getColumnIndex(tmpcol_month_year))); //<<<<<<<<<< NOTE data is MM/YY (otherwise which date to use? considering result will be arbrirtaryy)
// Adding record to list
listBillsDates.add(bills);
} while (cursor.moveToNext());
}
cursor.close();
db.close();
// return category list
return listBillsDates;
}
The above has been tested and run and using the following code :-
ArrayList<Bills> myMonthlyTotals = mDBHelper.getDateByUserID(1);
Log.d("BILLSCOUNT","The number of bills extracted was " + String.valueOf(myMonthlyTotals.size()));
for (Bills b: myMonthlyTotals) {
Log.d("MONTHYLTOTAL","Monthly total for " + b.getDateString() + " was " + String.valueOf(b.getAmount()));
}
In an activity, resulted in the following in the log
:-
04-14 11:58:25.876 16653-16653/? D/BILLSCOUNT: The number of bills extracted was 2
04-14 11:58:25.877 16653-16653/? D/MONTHYLTOTAL: Monthly total for 03/19 was 830
04-14 11:58:25.877 16653-16653/? D/MONTHYLTOTAL: Monthly total for 04/19 was 1000
Please consider the comments in regard to values from non-aggreagted columns be arbitrary values. As per :-
Each non-aggregate expression in the result-set is evaluated once for an arbitrarily selected row of the dataset. The same arbitrarily selected row is used for each non-aggregate expression. Or, if the dataset contains zero rows, then each non-aggregate expression is evaluated against a row consisting entirely of NULL values. SELECT - 3. Generation of the set of result rows.
As per the comments, using recognised date formats can make the underlying SQL simpler and likely more efficient.

Single Row Sub query return more than one row

I wrote a SQL query which updates the record, in most cases it runs fine, but from yesterday it fails to update the row.I am currently working on Spring MVC based Web application, in which I need to perform the DML operation by calling update()method.
I am using JDBC template and in my update method i wrote this query.
It fetches the 947 records for January month 2018 and I already checked all records are unique.
I am here stuck, i am not able to find the duplication of record.I thought this exception occurred because of multiple record, but I think i am wrong, there is something which i am not able to identify.
Here is the query:
UPDATE SALARY_DETAIL_REPORT_012018 sd
SET sd.EMP_ID =
(SELECT e.emp_id from EMPLOYEE e
WHERE e.VERIFY_STATUS = 1
AND e.RELIEF_TYPE IS NULL
AND e.emp_code = to_char(sd.emp_code)
AND e.EMP_TYPE_ID!=03)
WHERE EXISTS
(SELECT e.emp_id from EMPLOYEE e
WHERE e.VERIFY_STATUS = 1
AND e.emp_code = to_char(sd.emp_code)
AND e.EMP_TYPE_ID!=03 AND e.RELIEF_TYPE IS NULL)
AND sd.EMP_ID IS NULL
AND sd.REFERENCE_ID LIKE '203-%'
and HERE is Java Code in my DAOImpl class
public void update(String tableSuffix, String regionId, String circleId) {
String tabName = "SALARY_DETAIL_REPORT_" + tableSuffix;
String q = "UPDATE " + tabName + " sd"
+ " SET sd.EMP_ID = (SELECT e.emp_id "
+ " from EMPLOYEE e WHERE e.VERIFY_STATUS = 1 AND e.RELIEF_TYPE IS NULL "
+ " AND e.emp_code = to_char(sd.emp_code) AND e.EMP_TYPE_ID!=03) "
+ " WHERE "
+ " EXISTS (SELECT e.emp_id from EMPLOYEE e WHERE e.VERIFY_STATUS = 1 "
+ " AND e.emp_code = to_char(sd.emp_code) AND e.EMP_TYPE_ID!=03 AND e.RELIEF_TYPE IS NULL) "
+ " AND sd.EMP_ID IS NULL";
if (circleId != null && !circleId.trim().equals("")) {
q += " AND sd.REFERENCE_ID LIKE '" + circleId + "-%' ";
} else {
q += " AND sd.REFERENCE_ID LIKE '" + regionId + "-%' ";
}
// System.out.println("q " + q);
MapSqlParameterSource param = new MapSqlParameterSource();
getNamedParameterJdbcTemplate().update(q, param);
}
Please suggest me the best solution

You need to find the rows that prevent your query from running.
Run this query:
SELECT sd.emp_code, COUNT(*) as cnt
FROM EMPLOYEE e
WHERE e.VERIFY_STATUS = 1 AND
e.RELIEF_TYPE IS NULL AND
e.emp_code = to_char(sd.emp_code) AND
e.EMP_TYPE_ID <> 03
GROUP BY sd.emp_code
HAVING COUNT(*) > 1;
This has the candidate problems. What can you do? The simplest thing is one of the problems:
Change the SELECT to SELECT MAX(sd.emp_code)
Change the WHERE with AND rownum = 1

Does the prepared-statement work this way?

I am trying to populate one table in my database with pretty complex data. For this, I am using a generator API (which gives me random data).
public void populateCrackers(){
PreparedStatement psm;
String queryJoke = "(SELECT jid FROM Jokes WHERE jid=?)";
String queryHat = "(SELECT hid FROM Hats WHERE hid=?)";
String queryGift = "(SELECT gid FROM Gifts WHERE gid=?)";
String query = "INSERT INTO Crackers(cid, name, jid, hid, gid, quantity) VALUES(" +
"?, " +
"?, " +
queryJoke + ", " +
queryHat + ", " +
queryGift + ", " +
"?)";
System.out.println(query);
String cracker_String = utils.JSONUtils.getJSON(crackerAPI, client);
JSONObject crackerJSON = new JSONObject(cracker_String);
JSONArray crackers = crackerJSON.getJSONArray("results");
for(int j=0; j<crackers.length(); j++){
try{
psm = connection.prepareStatement(query);
psm.setInt(1,crackers.getJSONObject(j).getInt("cid"));
psm.setString(2, crackers.getJSONObject(j).getString("cname"));
psm.setInt(3, crackers.getJSONObject(j).getInt("rjoke"));
psm.setInt(4, crackers.getJSONObject(j).getInt("rhat"));
psm.setInt(5, crackers.getJSONObject(j).getInt("rgift"));
psm.setInt(6, crackers.getJSONObject(j).getInt("cquantity"));
psm.execute();
System.out.println(crackers.getJSONObject(j).get("cid") + " "
+ crackers.getJSONObject(j).get("cname") + " "
+ crackers.getJSONObject(j).get("cquantity") + " "
+ crackers.getJSONObject(j).get("rjoke") + " "
+ crackers.getJSONObject(j).get("rhat") + " "
+ crackers.getJSONObject(j).get("rgift"));
}catch (Exception e){
e.printStackTrace();
}
}
}
This is the method that populates my "Crackers" tab. I am wondering if this be accepted as a prepared statement. When I run it in psql interactive command line tool, exactly that statement with some chosen ids (e.g INSERT INTO Crackers (cid, name, hid, jid, gid, quantity) VALUES('cid', 'name', (SELECT hid FROM Hats WHERE hid=11), (SELECT jid FROM Jokes where jid=99), (SELECT gid FROM Gifts WHERE gid=13), 5) it works flawlessly.
Does my preparedstatement break the Constraint?
Any ideas?
LATER EDIT: The inconsistency is the form of that null values can reach my Crackers table (e.g. Cracker(1, "hello", null, null, 3, 123) appears in the table.

There is nothing about Prepared statement. Constraint can be broken by parameters you set to it. And you can run your PLSQL statement as anonimous block in PreparedStatement as well.
Just surround it with BEGIN ... END. only one thing is different - for JDBC parameters are ? mark not :parameter as for PLSQL and there is no way to use named parameter.
That means if you need to use parameter more than once for JDBC you have to have that many ? marks and set all of them.
So, focus on parameters you pass to and their sequence.

The code is correct, though the prepared statement must be closed, and it would be better to create the statement once, before the for loop.
Now there is crackers.length() times a statement created but not closed. That might give problems.
Use the try-with-resouce syntax for automatic closing, irrespective of any exception or return.
try (PreparedStatement psm = connection.prepareStatement(query)) {
for (int j = 0; j < crackers.length(); j++) {
...
psm.executeUpdate();
And call executeUpdate instead of the more general execute. The resulting update count might be of interest (1/0).

I realised I had the wrong constraints on my table. I was letting null values in. There was nothing wrong with the prepared statement.
The right query to create the table is this one:
String createCrackersQuery = "CREATE TABLE Crackers(" +
" cid INTEGER," +
" name VARCHAR NOT NULL," +
" jid INTEGER NOT NULL," +
" hid INTEGER NOT NULL," +
" gid INTEGER NOT NULL," +
" quantity INTEGER NOT NULL," +
" CONSTRAINT Cracker_Primary PRIMARY KEY (cid)," +
" CONSTRAINT Cracker_Name_Unique UNIQUE(name)," +
" CONSTRAINT Joke_Foreign FOREIGN KEY (jid) REFERENCES Jokes(jid)," +
" CONSTRAINT Hat_Foreign FOREIGN KEY (hid) REFERENCES Hats(hid), " +
" CONSTRAINT Gift_Foreign FOREIGN KEY (gid) REFERENCES Gifts(gid)" +
")";

Multiple tables join via Spring JdbcTemplate returns an empty result

Trying to join 3 tables within a query returns an empty result. Strange enough, having one table removed (two tables join) returns some set. Here is what I do:
String sql = "SELECT\n" +
" tc.constraint_name, tc.table_name, kcu.column_name, \n" +
" ccu.table_name AS foreign_table_name,\n" +
" ccu.column_name AS foreign_column_name, constraint_type \n" +
"FROM \n" +
" information_schema.table_constraints AS tc \n" +
" JOIN information_schema.key_column_usage AS kcu\n" +
" ON tc.constraint_name = kcu.constraint_name\n" +
" JOIN information_schema.constraint_column_usage AS ccu\n" +
" ON ccu.constraint_name = tc.constraint_name\n" +
"WHERE constraint_type = 'FOREIGN KEY'";
List<Map<String, Object>> foreignTable1 = jdbcTemplate(getShardId(sku)).queryForList(sql);
Would always return an empty set.

Try using outer joins and check whether there are rows which don't have corresponding IDs so that the join removes the non-matching rows. Especially that you write, that two tables result in a non-empty result set seems to indicate, that the join with the third table does not result in matching rows of the result set of the first two.

Categories