Important Common List of MySQL Queries

November 1, 2016, 1:49 am

≫ Next: Leonids - Android Particle Animation Library

≪ Previous: Encode String security using MD5 algorithm in Java

*Aggregates*	Pivot table using math tricks	Scope to the week of a given date	Parents with and without children
Basic aggregation	Pivot table with CONCAT	Sum accumulated time by date	*Ordering resultsets*
Aggregate across columns	Pivot table without GROUP_CONCAT	Sum booking days for a given month	Emulate Row_Number()
Aggregates across multiple joins	Unpivot a table	Sum time values	Next row
Aggregates excluding leaders	*Data comparison*	The date of next Thursday	Order by leading digits
Aggregates of specified size	Backslashes in data	Track when a value changed	Order by month name
All X for which all Y are Z	Compare data in two tables	What month does a week fall in?	Order by numerics then alphas
All X meeting multi-row conditions	Show rows where column value changed	YearMonth()	Pagination
Avoiding repeat aggregation	Using bit operators	*Period arithmetic*	Suppress repeating ordering values
Break out sum on condition	*Database metadata*	Audit trails and point-in-time architecture	*Relational division*
Cascading aggregates	Compare structures of two tables	Find overlapping periods	All possible recipes with given ingredients
Combine Group_Concat() with counts	Compare two databases	Find sequenced duplicates	All X for which all Y are Z (relational division)
Correlated aggregating subqueries	Database size	In or out at a given date and time?	Who makes all the parts for a given assembly?
Cross-aggregates	Find the size of all databases on the server	Peak visit counts by datetime period	*Sequences*
Every paying customer wins a prize	List databases, tables, columns	Sum for time periods	Make a table of sequential ints
Group data by datetime periods	List differences between two databases	*Scheduling*	Find adjacent unbooked theatre seats
League table	List users of a database	Appointments available	Find average row-to-row time interval
Nested, banded aggregation	Rename Database	Find available reservation periods	Find blocks of unused numbers
Pairwise matchmaking	Replace text in all string columns of a database	Game schedule	Find missing numbers in a sequence
Show only one child row per parent row	*Foreign keys*	Is a given booking period available?	Find missing values in a range
Skip repeating values	Foreign key basics	Pivot table schedule	Find previous and next values in a sequence
Totals and subtotals--simply	Change or drop a foreign key	*Frequencies*	Find row with next value of specified column
Track state changes	Find child tables	Display column values which occur N times	Find sequence starts and ends
Values linked with all values of another column	Find cross-database foreign keys	Display every Nth row	Find shortest & longest per-user event time intervals
Within-group aggregates	Find parent tables	*Graphs and Hierarchies*	Find specific sequences
Within-group aggregates with a wrinkle	*Primary keys*	Trees, networks and parts explosions in MySQL	Find the next value after a sequence
Within-group quotas (Top N per group)	Add auto-incrementing primary key to a table	Dijkstra’s shortest path algorithm	Gaps in a time series
*Aggregates and Statistics*	Auto-increment: reset next value	Multiple trees in one table	Make values of a column sequential
Average the top 50% values per group	Find primary key of a table	Tree query performance	Track stepwise project completion
Correlation	*Show*	Trees of known depth	Winning streaks
Count unique values of one column	Show Create Trigger	*JOIN*	*Spherical geometry*
Median	Show Table Status equivalent from information_schema	Approximate joins	Great circle distance
Mode	Show Tables	Cascading JOINs	*Statistics without aggregates*
Moving average	*Date and time*	Classroom scheduling	Top ten
Multiple sums across a join	Age in years	Data-driven joins	*Stored procedures*
Percentiles	Compute date from year, week number and weekday	Full Outer Join	A cursor if necessary, but not necessarily a cursor
Rank order	Count business days between two dates	Intersection and difference	Emulate sp_exec
Running sums, chequebooks	Count Tuesdays between two dates	Many-to-many joins	Variable-length argument for query IN() clause
Sum across categories	Date of first day of next month	What else did buyers of X buy?	*Strings*
*Aggregates from bands of values*	Date of first Friday of next month	*Join or subquery?*	Count delimited substrings
History of world records	Date of Monday in a given week of the year	Parents without children	Count substrings
*Pivot tables (crosstabs)*	Date of Monday of this week	Parties who have contracts with one another	Levenshtein distance
Pivot table basics: rows to columns	Datetime difference	The unbearable slowness of IN()	Proper case
Automate pivot table queries	Duration in years, months, days and time	The [Not] Exists query pattern	Retrieve octets from IP addresses
Group column statistics in rows	Julian date	What exams did a student not register for?	Return digits or alphas from a string
Monthly expenses	Last business day before a reference date	*NULLs*	Split a string
Monthly sales	Make a calendar table	List NULLs at end of query output	Strip HTML tags

Basic aggregation

Aggregates are more popular than any other lookup topic here, by a ratio of more than 2 to 1.

Basic aggregation is the simplest grouping query pattern: for column foo, display the smallest, largest, sum, average or some other statistic of column bar values:
SELECT foo, MIN(bar) AS bar FROM tbl GROUP BY foo
Return the highest bar value for each foo, ordering top to bottom by that value:
SELECT foo, MAX(bar) AS Count FROM tbl GROUP BY foo ORDER BY Count DESC;
Ditto for AVG(), COUNT() etc. The pattern easily extends to multiple grouping column expressions.

For aggregating functions like MIN() and MAX() that return a single value, there may be multiple instances of the result. If you wish to see them, put the aggregating query in a subquery and join to it from a direct query on the table:
SELECT a.foo, a.bar FROM tbl a JOIN ( SELECT foo, MAX(bar) AS Count FROM tbl GROUP BY foo ) b ON a.foo=b.foo AND a.bar=b.count ORDER BY a.foo, a.bar;
MySQL introduced the SQL extension GROUP_CONCAT(), which makes short work of listing items in groups. For example, given a table of suppliers and the parts they make ...
CREATE TABLE supparts(supID char(2),partID char(2)); INSERT INTO supparts VALUES ('s1','p1'),('s1','p2'),('s1','p3'),('s1','p4'),('s1','p5'),('s1','p6'), ('s2','p1'),('s2','p2'),('s3','p2'),('s4','p2'),('s4','p4'),('s4','p5');
List suppliers for each part:
SELECT partID,GROUP_CONCAT(supID ORDER BY supID) AS Suppliers FROM supparts GROUP BY partID; +--------+-------------+ | partID | Suppliers | +--------+-------------+ | p1 | s1,s2 | | p2 | s1,s2,s3,s4 | | p3 | s1 | | p4 | s1,s4 | | p5 | s1,s4 | | p6 | s1 | +--------+-------------+
When there are several tables to be joined, the beginner may feel overwhelmed by the complexity of the problem. Suppose you're asked to retrieve the top computer desk salesperson for this schema:
drop table if exists salespersons, orders, orderlines, products; create table salespersons(salespersonid int,name char(8)); insert into salespersons values(1,'Sam'),(2,'Xavier'); create table orders(orderid int,salespersonid int); insert into orders values(1,1),(2,1),(3,1),(4,2),(5,2); create table orderlines(lineid int,orderid int,productid int,qty int); insert into orderlines values(1,1,1,1),(2,1,1,2),(3,2,2,1),(4,3,1,1),(5,4,1,1),(6,5,2,2); create table products(productid int,name char(32)); insert into products values(1,'computer desk'),(2,'lamp'),(3,'desk chair');
Two rules of thumb help with problems like this: solve one step at a time, and work from inside out. Here "inside out" means start by building the join list needed to retrieve sales data:
from salespersons s join orders o using(salespersonid) join orderlines l using(orderid) join products p using(productid)
Test those joins with a query that just lists sales data:
select s.name, p.name, l.qty from salespersons s join orders o using(salespersonid) join orderlines l using(orderid) join products p using(productid) +--------+---------------+------+ | name | name | qty | +--------+---------------+------+ | Sam | computer desk | 1 | | Sam | computer desk | 2 | | Sam | lamp | 1 | | Sam | computer desk | 1 | | Xavier | computer desk | 1 | | Xavier | lamp | 2 | +--------+---------------+------+
Now we just need to filter for 'computer desk' sales, add aggregation, and pick off the top seller:
select s.name, sum(l.qty) as n -- sum quantities from salespersons s join orders o using(salespersonid) join orderlines l using(orderid) join products p using(productid) where p.name='computer desk' group by s.name -- aggregate by salesperson order by n desc limit 1; -- order by descending sum, pick off top value +------+------+ | name | n | +------+------+ | Sam | 4 | +------+------+
If columns other than the GROUP BY columns must be retrieved, and if the grouping expression does not have a strictly 1:1 relationship with (ie., are not "functionally dependent" on) those columns, then to avoid returning arbitrary values for those non-grouping columns, you need to either add those columns to the Group By clause, or put the GROUP BY query in a subquery and join that result to the other columns, for example:
SELECT s.partID, s, thiscol, s.thatcol, anothercol, x.Suppliers FROM supparts s JOIN ( SELECT partID,GROUP_CONCAT(supID ORDER BY supID) AS Suppliers FROM supparts GROUP BY partID ) x USING(partID)
If sql_mode includes ONLY_FULL_GROUP_BY, MySQL returns an error for a query that Selects non-aggregated columns that aren't functionally dependent on the grouping columns.

Before version 5.7.5 MySQL, with ONLY_FULL_GROUP_BY set, blindly rejected queries like this with the message that those columns are missing from the GROUP BY expression. Starting in 5.7.5 though, MySQL is better at detecting functional dependencies, and it won't generate such an error if it concludes that the ungrouped columns are functionally dependent on grouoed columns. If it decides they aren't functionally dependent, the message says that. For full discussion of this, see Roland Bouman's excellent discussion at http://rpbouman.blogspot.nl/2014/09/mysql-575-group-by-respects-functional.html

Last updated 01 Oct 2016

Back to the top

Subscribe for Us

Feedback

Aggregate across columns

You track squash court bookings and fees. A court booking has one fee, but it has two member references, and those can be the same if one member has brought a guest. For each booking row, the fee is to be split between the two members. What query correctly aggregates fees including cases where the two members of record are the same?
DROP TABLE IF EXISTS bookings; CREATE TABLE bookings ( court_id int(11) NOT NULL, member1 int(11) NOT NULL, member2 int(11) NOT NULL, time timestamp NOT NULL, fee decimal(5,2) NOT NULL ); INSERT INTO bookings ( court_id , member1 , member2 , time , fee ) VALUES (1, 1000, 1001, '2009-09-09 15:49:38', 3.00), (2, 1000, 1000, '2009-09-08 15:50:04', 3.00);
For this data the correct result is
member fees 1000 4.50 1001 1.50
An efficient solution, posted by 'laptop alias' on a MySQL forum:
SELECT member, ROUND(SUM(fee/2),2) AS total FROM ( SELECT member1 AS member, fee FROM bookings UNION ALL SELECT member2, fee FROM bookings ) AS tmp GROUP BY member;

Last updated 09 Sep 2009

Back to the top

Subscribe for Us

Feedback

Aggregates across multiple joins

Given a parent table and two child tables, a query which sums values in both child tables, grouping on a parent table column, returns sums that are exactly twice as large as they should be. In this example from the MySQL General Discussion list:


DROP TABLE IF EXISTS packageItem,packageCredit,packageItemTax; 
CREATE TABLE packageItem ( 
 packageItemID INT,  
 packageItemName CHAR(20),  
 packageItemPrice DECIMAL(10,2) 
); 
INSERT INTO packageItem VALUES(1,'Delta Hotel',100.00); 

CREATE TABLE packageCredit ( 
 packageCreditID INT,  
 packageCreditItemID INT,

Last updated 22 Feb 2013

Back to the top

Subscribe for Us

Feedback

Aggregates excluding leaders

You have a table of grouped ranks ...


DROP TABLE IF EXISTS grps,ranks; 
CREATE TABLE grps (grp int); 
INSERT INTO grps VALUES(1),(2),(3),(4); 
CREATE TABLE ranks(grp int,rank int); 
INSERT INTO ranks VALUES(1, 4 ),(1, 7 ),(1, 9 ),(2, 2 ),(2, 3 ),(2, 5 ),(2, 6 ),(2, 8 ),(3, 1 ),(4,11 ),(4,12 ),(4,13 );

and you wish to list ranks by group omitting the leading rank in each group. The simplest query for group leaders is ...

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Aggregates of specified size

Find the values of a table column c1 for which there are a specified number of listed values in another column c2.

To get an overview of the values of c2 for each value of c1:


SELECT 
  c1,  
  GROUP_CONCAT(c2 ORDER BY c2) AS 'C2 values'
FROM table 
GROUP BY c1;

To retrieve a list of c1 values for which there exist specific values in another column c2, you need an IN clause specifying the c2 values and a HAVING clause specifying the required number of different items in the list ...


SELECT c1  
FROM table 
WHERE c2 IN (1,2,3,4) 
GROUP BY c1 
HAVING COUNT(DISTINCT c2)=4;

This is easy to generalise to multiple column expressions, and a HAVING clause specifying any number of items from the IN list.

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

All X for which all Y are Z

You have an election database with tables for candidates, parties and districts. A candidate belongs to one party; a district may have any number of candidates:


DROP TABLE IF EXISTS parties,districts,candidates; 
CREATE TABLE parties ( 
  party char(12) NOT NULL, 
  PRIMARY KEY (party) 
) ENGINE=MyISAM DEFAULT CHARSET=latin1; 
INSERT INTO parties VALUES ('Conservative'),('Liberal'),('Socialist'),('Green'),('Libertarian'); 

CREATE TABLE districts ( 
  district char(10) DEFAULT NULL 
) ENGINE=MyISAM DEFAULT CHARSET=latin1; 
INSERT INTO districts VALUES ('Essex'),('Malton'),('Riverdale'),('Guelph'),('Halton');

Last updated 19 Mar 2016

Back to the top

Subscribe for Us

Feedback

All X meeting multi-row conditions

Given this table ...


drop table if exists mpu; 
create table mpu( id int, id_marketplace int, label int, property int); 
insert into mpu values 
(1 , 10 , 3 , 0 ),(2 , 10 , 6 , 35 ),(4 , 10 , 10 , 22 ),(5 , 10 , 9 , 0 ), 
(6 , 11 , 3 , 0 ),(7 , 11 , 6 , 5 ),(8 , 11 , 7 , 7 ),(9 , 11 , 7 , 10 ), 
(10 , 11 , 10 , 21),(11 , 12 , 3 , 0 ),(12 , 12 , 6 , 5 ),(13 , 12 , 7 , 8 ), 
(14 , 12 , 7 , 9 ),(15 , 12 , 10 , 21 ),(16 , 13 , 3 , 0  ),(17 , 13 , 6 , 35 ), 
(18 , 13 , 7 , 7),(19 , 13 , 7 , 8 ),(20 , 13 , 10 , 20 );select * from mpu; 
+------+----------------+-------+----------+ 
| id   | id_marketplace | label | property | 
+------+----------------+-------+----------+

Last updated 19 Jan 2014

Back to the top

Subscribe for Us

Feedback

Avoiding repeat aggregation

In a good introductory tutorial on MySQL subqueries, Jeremy Cole developed a triply nested query to retrieve the name, population, head of state, and number of official languages in countries with the most official languages. The query uses two tables from the MySQL world database:


CREATE TABLE country ( 
  Code char(3) NOT NULL DEFAULT '' PRIMARY KEY, 
  Name char(52) NOT NULL DEFAULT '', 
  Population int(11) NOT NULL DEFAULT '0', 
  HeadOfState char(60) DEFAULT NULL, 
  ... other columns ... 
); 
CREATE TABLE countrylanguage ( 
  CountryCode char(3) NOT NULL DEFAULT '' PRIMARY KEY, 
  Language char(30) NOT NULL DEFAULT '', 
  IsOfficial enum('T','F') NOT NULL DEFAULT 'F',

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Break out sum on condition

To obtain part sums of a column based on specific conditions, use If() or CASE, eg ...


select  
  sum( if(condition1, col, 0) ) as cond1, 
  sum( if(condition2, col, 0) ) as cond2, 
  ...etc...  
from tbl;

Last updated 16 Aug 2012

Back to the top

Subscribe for Us

Feedback

Cascading aggregates

When you have parent-child-grandchild tables, eg companies, users, actions, and your query requirement is for per-parent aggregates from the child table and per-child aggregates from the grandchild table, then cascading joins yield spuriously multiplied counts, and correlated subqueries fail because the second correlated subquery cannot find a visible joining column.

One solution is to use derived tables. Assuming ...


DROP TABLE IF EXISTS companies,users,actions; 
CREATE TABLE companies (id int, name char(10)); 
INSERT INTO COMPANIES VALUES(1,'abc ltd'),(2,'xyz inc'); 
CREATE TABLE users (id INT,companyid INT); 
INSERT INTO users VALUES(1,1),(2,1),(3,1),(4,2),(5,2); 
CREATE TABLE actions (id INT, userid INT, date DATE);  
INSERT INTO actions VALUES 
( 1, 1, '2009-1-2'),( 2, 1, '2009-1-3'),( 3, 2, '2009-1-4'),( 4, 2, '2009-1-5'),( 5, 3, '2009-1-6'),

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Combine Group_Concat() with counts

Rather than list instances including repeats on one line, you want to list distinct instances and their counts. One way is to do a simple GROUP BY query and in your application layer remove the newlines from the result. Or you can do it in one step:


drop table if exists t; 
create table t ( 
  type int(10) , 
&nbspnbsp; instance int(10)  
) ; 
insert into t values  
(1,4),(1,7),(1,9),(1,10),(2,2),(2,3),(2,5),(2,6),(2,8),(3,1),(4,11); 

select group_concat( concat( type,'(',qty,')') separator ', ') list  
from ( 
  select type, count(*) qty  
  from t  
  group by type 
) n 
+------------------------+ 
| list                   | 
+------------------------+ 
| 1(4), 2(5), 3(1), 4(1) | 
+------------------------+

Last updated 28 Jun 2012

Back to the top

Subscribe for Us

Feedback

Correlated aggregating subqueries

If you need a child table aggregate result for each row in a query, what's faster, putting the correlated aggregating subquery in the SELECT list, or in the FROM clause?

Thus, given tables stocks(id, user, ticker,lastvalue) and transactions(id, sid, action, qty, price, commission), and the requirement to retrieve stock values and per-stock transaction counts, which of these queries will run faster?


select ticker, id, lastvalue, (select count(*) from transactions where sid=stocks.id) N  
from stocks;

... or ...


select s.ticker, s.id, s.lastvalue, t.N 
from stocks s  
join ( 
  select sid, count(*) N  
  from transactions 
  group by sid 
) t on s.id=t.sid;

The first query's syntax is simpler, so it's often the first choice. EXPLAIN reports that the second query requires examination of more rows than the first query does.

But benchmarks with caching turned off show that the second query is much faster--mainly because it executes one subquery rather than a subquery per row.

If you have a choice, put the aggregating subquery in the FROM clause.

Last updated 10 Oct 2014

Back to the top

Subscribe for Us

Feedback

Cross-aggregates

Given the table authorbook(authid INT, bookid INT), what query finds the books who have authors with more than one book in the table?

Even one level of recursion can induce a mild trance. Escape the trance by taking the problem one step at a time. First write the query that finds the authors with multiple books. Then join an outer query to that on authorid, and have the outer query select bookid:


SELECT a1.bookid 
FROM authorbook a1 
INNER JOIN ( 
  SELECT authid,count(bookid) 
  FROM authorbook a2 
  GROUP BY authid 
  HAVING COUNT(bookid)>1 
) AS a3 ON a1.authid=a3.authid;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Every paying customer wins a prize

A superstore is running a promotion: each day, every fifth a customer wins a prize. Each day you're given a text file with data columns for customerID, timestamp and order amount. How do you find every fifth customer?

Load the data into a table, then write the query inside out; with this kind of problem, nested queries are a boon.

1. Using Load Data Infile, load the text file into a table (indexless for speed) named tbl with columns for customerID, timestamp and amount, and index the table on (customerID, timestamp).

2. Write a query that tracks per-customer order by timestamp with user variables:


set @id='', @ord=1; 
select

Last updated 12 Sep 2013

Back to the top

Subscribe for Us

Feedback

Group data by datetime periods

To group rows by a time period whose length in minutes divides evenly into 60, use this formula:


GROUP BY ((60/periodMinutes) * HOUR( thistime ) + FLOOR( MINUTE( thistime ) / periodMinutes ))

where thistime is the TIME column and periodMinutes is the period length in minutes. So to group by 15-min periods, write ...


SELECT ... 
GROUP BY ( 4 * HOUR( thistime ) + FLOOR( MINUTE( thistime ) / 15 )) 
...

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

League table

Here is a simple soccer league table setup that was developed in the MySQL Forum by J Williams and a contributor named "Laptop Alias". The teams table tracks team ID and name, the games table tracks home and away team IDs and goal totals for each game. The query for standings is built by aggregating a UNION of home team and away team game results:


DROP TABLE IF EXISTS teams, games; 
CREATE TABLE teams(id int primary key auto_increment,tname char(32)); 
CREATE TABLE games(id int primary key auto_increment, date datetime,  
                   hteam int, ateam int, hscore tinyint,ascore tinyint); 

INSERT INTO teams VALUES(1,'Wanderers'),(2,'Spurs'),(3,'Celtics'),(4,'Saxons'); 
INSERT INTO games VALUES 
(1,'2008-1-1 20:00:00',1,2,1,0),(2,'2008-1-1 20:00:00',3,4,0,2), 
(3,'2008-1-8 20:00:00',1,3,1,1),(4,'2008-1-8 20:00:00',2,4,2,1); 
SELECT * FROM teams; 
+----+-----------+

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Nested, banded aggregation

Employee sales commission rates increase with sales totals according to specified bands of sales total amounts—like a graduated income tax in reverse. To compute total commissions due each employee, we need to aggregate twice: first to get sales per employee, then to get commissions per employee:


DROP TABLE IF EXISTS sales, commissions; 
CREATE TABLE sales(employeeID int,sales int); 
INSERT INTO sales VALUES(1,2),(1,5),(1,7),(2,9),(2,15),(2,12); 
SELECT * FROM sales; 
+------------+-------+ 
| employeeID | sales | 
+------------+-------+ 
|          1 |     2 | 
|          1 |     5 | 
|          1 |     7 | 
|          2 |     9 |

Last updated 15 Nov 2009

Back to the top

Subscribe for Us

Feedback

Pairwise matchmaking

Given tables tracking users and their hobbies, how do we write a query that ranks pairs of users on hobby similarity?


DROP TABLE IF EXISTS users,hobbies,users_hobbies; 
CREATE TABLE users( id int, name char(16) ) ; 
INSERT INTO users VALUES (1,'John'),(2,'Lewis'),(3,'Muhammad'); 
CREATE TABLE hobbies( id int, title char(16) ) ; 
INSERT INTO hobbies  
VALUES (1,'Sports'),(2,'Computing'),(3,'Drinking'),(4,'Racing'),(5,'Swimming'),(6,'Photography'); 
CREATE TABLE users_hobbies( user_id int, hobby_id int ) ; 
INSERT INTO users_hobbies  
VALUES (1,2),(1,3),(1,6),(2,1),(2,5),(2,6),(3,2),(3,5),(3,6),(1,2),(1,3),(1,6),(2,1), 
(2,5),(2,6),(3,2),(3,5),(3,6),(1,2),(1,3),(1,6),(2,1),(2,5),(2,6),(3,2),(3,5),(3,6);

Last updated 17 Sep 2010

Back to the top

Subscribe for Us

Feedback

Show only one child row per parent row

Given tables parent(id int not null primary key, etc...) and child (id int not null primary key, pid int not null references parent (id), etc...), how do we write a query that retrieves only one child row per pid even when the child table has multiple matching rows? MySQL permits use of GROUP BY even when the SELECT list specifies no aggregate function, so this will work:


select p.id, c.id  
from parent p  
join child c on p.id=c.pid 
group by p.id;

But is it accurate? No, because it displays only the first c.pid value it happens to find. For further discussion see Within-group aggregates.

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Skip repeating values

You want to report all unique values of a column and skip all rows repeating any of these values.


SELECT col 
FROM foo 
GROUP BY col

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Totals and subtotals--simply

You have a table that tracks attendance hours ...


drop table if exists t;  
create table t(d date, id int, hrs int);  
insert into t values  
('2013-10-1',1,5), ('2013-10-1',2,6), ('2013-10-1',3,2), ('2013-10-1',3,5),  
('2013-10-2',1,1), ('2013-10-2',1,2), ('2013-10-2',2,3), ('2013-10-2',2,4),  
('2013-10-3',1,2), ('2013-10-3',1,2), ('2013-10-3',1,2);

... and you need an attendance summary by date and person. It turns out that this aggregating query is easy top write—first write an inner query to Group By date and person With Rollup, then write an outer query that renames the Rollup Null labels.

Last updated 31 Oct 2013

Back to the top

Subscribe for Us

Feedback

Track state changes

You have a table containing a sequence of IDs, states and dates. The query requirement is to list state changes and their dates ordered by date, and count instances of each sequenced state. This data from a blog entry by a Russian MySQLer calling himself Quassnoi ...


+----+-------+------------------+ 
|ID  | State |  Datetime        | 
+----+-------+------------------+ 
| 12 |   1   | 2009-07-16 10:00 | 
| 45 |   2   | 2009-07-16 13:00 | 
| 67 |   2   | 2009-07-16 14:40 | 
| 77 |   1   | 2009-07-16 15:00 | 
| 89 |   1   | 2009-07-16 15:30 | 
| 99 |   1   | 2009-07-16 16:00 | 
+----+-------+------------------+

Last updated 20 Oct 2014

Back to the top

Subscribe for Us

Feedback

Values linked with all values of another column

You have a table in which each row references one text and one keyword in the text ...


DROP TABLE IF EXISTS keywords; 
CREATE TABLE keywords (txtID int, keyword char(8)); 
INSERT INTO keywords VALUES(1 , 'foo'),(2 , 'bar'),(1 , 'foo'),(2 , 'foo');

... and you want a list of texts which include every keyword.

You might think you have to join and match. You don't. All you need to do is count the distinct keywords which occur for each text, then for each text compare that number with the entire list of distinct keywords:


SELECT txtID, COUNT(DISTINCT keyword) AS N 
FROM keywords 
GROUP BY txtID 
HAVING N = (SELECT COUNT(DISTINCT keyword) FROM keywords); 
+-------+---+ 
| txtID | N | 
+-------+---+ 
|     2 | 2 | 
+-------+---+

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Within-group aggregates

You have a products table with columns item, supplier, price. Multiple suppliers offer various prices for the same item. You need to find the supplier with the lowest price for each item.
DROP TABLE IF EXISTS products; CREATE TABLE products(item int,supplier int,price decimal(6,2)); INSERT INTO products VALUES(1,1,10),(1,2,15),(2,2,20),(2,1,21),(2,2,18); SELECT * FROM products; +------+----------+-------+ | item | supplier | price | +------+----------+-------+ | 1 | 1 | 10.00 | | 1 | 2 | 15.00 | | 2 | 2 | 20.00 | | 2 | 1 | 21.00 | | 2 | 2 | 18.00 | +------+----------+-------+
Your first thought may be to GROUP BY item, but that is not guaranteed to return the correct supplier value for each minimum item price. Grouping by both item and supplier will return more information than you want. Nor can you write WHERE price=MIN(...) because the query engine will evaluate the WHERE clause before it knows the MIN value.

This is the problem of aggregating within aggregates. It is sometimes called the 'groupwise aggregates' problem, but the term 'groupwise' is ambiguous. We think better names for it are subaggregates, inner aggregates, or within-group aggregates.

It's easy to show that the within-group aggregates problem is a form of the problem of returning values from non-grouping columns in a GROUP BY query. Suppose you write ...
SELECT item, supplier, Min(price) FROM products GROUP BY item;
Will this reliably return the correct supplier per item? No. Unless there is exactly one supplier per item, the supplier value returned will be arbitrary. To retrieve the correct supplier for each item, you need more query logic.

Here are six solutions. As you'll see, some are very slow, a couple are quite fast, and some work for only some versions of the problem:

1. Self-exclusion join solution: One way to model within-group minima or maxima is via an left self exclusion join...
SELECT p1.item, p1.supplier, p1.price FROM products AS p1 LEFT JOIN products AS p2 ON p1.item = p2.item AND p1.price > p2.price WHERE p2.item IS NULL;
...because in the resultset built by joining on left item = right item and left price > right price, the left-sided rows for which there is no greater right-sided price are precisely the per-item rows with the smallest prices. The query runs more than an order of magnitude faster with an index on (item, supplier).

2. Intermediate table solution: Another solution is to derive an intermediate table of aggregated minimum prices, then query that table. Before MySQL 4.1, the intermediate table has to be a temporary table:
CREATE TEMPORARY TABLE tmp ( item INT, minprice DECIMAL DEFAULT 0.0 ); LOCK TABLES products READ; INSERT INTO tmp SELECT item, MIN(price) FROM products GROUP BY item;
to which you then join the products table to find the matching suppliers:
SELECT products.item, supplier, products.price FROM products JOIN tmp ON products.item = tmp.item WHERE products.price=tmp.minprice; UNLOCK TABLES; DROP TABLE tmp;
3. Correlated subquery solution: From MySQL 4.1 on, the temporary table can be a correlated subquery. This is the most intuitively obvious syntax for the problem. It's also the slowest—up to a hundred times slower than the exclusion join, whether the queries are compared with or without indexes:
SELECT item, supplier, price FROM products AS p1 WHERE price = ( SELECT MIN(p2.price) FROM products AS p2 WHERE p1.item = p2.item );
4. FROM clause subquery solution: If we move the aggregating subquery from the WHERE clause to the FROM clause, the query improves from a hundred times slower than the self-exclusion join to twenty times faster:
SELECT p.item, p.supplier, p.price FROM products AS p JOIN ( SELECT item, MIN(price) AS minprice FROM products GROUP BY item ) AS pm ON p.item = pm.item AND p.price = pm.minprice;
Some users have trouble mapping elements of this model to their instance of the problem. The model has five elements (or sets of them):

(i) a table, which might be a view, a single physical table, or a table derived from joins
(ii) one or more grouping columns,
(iii) one or more columns to aggregate,
(iv) one or more columns not mentioned in the GROUP BY clause,
(v) an aggregating job to do, typically MIN() or MAX().

In the product/minimum price solution above:

(i) table tbl = product
(ii) grouping column grouping_col = item
(iii) column to aggregate = col_to_aggregate = price
(iv) non-aggregated columns other_detail, ...etc... = supplier
(v) aggregating function = MIN().

Here's an interesting variation on the problem. A simple table tracks company branches and the kinds of stock they carry:
DROP TABLE IF EXISTS branch_stock; CREATE TABLE branch_stock( stock_id INT PRIMARY KEY AUTO_INCREMENT KEY, branch_id INT NOT NULL, stock_item VARCHAR(12) NOT NULL ); INSERT INTO branch_stock (branch_id,stock_item) VALUES (1,'trainers'),(1,'trainers'),(1,'jumper'),(2,'tie'),(2,'shoes');
How do we find the most frequent product for each branch, including ties? Join the intermediate table of grouped counts to itself on matching branches, non-matching stock and not-greater counts in one than in the other:
SELECT DISTINCT a.* FROM ( SELECT branch_id, stock_item, COUNT(*) qty FROM branch_stock GROUP BY branch_id,&nbspnbsp;stock_item ) a JOIN ( SELECT branch_id, stock_item, COUNT(*) qty FROM branch_stock GROUP BY branch_id, stock_item ) b ON b.branch_id=a.branch_id AND b.stock_item<>a.stock_item AND b.qty<=a.qty; +-----------+------------+-----+ | branch_id | stock_item | qty | +-----------+------------+-----+ | 1 | trainers | 2 | | 2 | shoes | 1 | | 2 | tie | 1 | +-----------+------------+-----+
5. Ordered subquery solution: A "trick" solution, made possible by MySQL's non-standard tolerance of GROUP BY when there is no aggregating SELECT expression, uses ORDER BY in a subquery to find the lowest prices, and GROUP BY in the outer query to pick them off:
SELECT * FROM ( SELECT * FROM products ORDER BY price ASC ) AS s GROUP BY item;
It's succinct, it's fast if the query engine can find an index to order on, and it retrieves the prices and associated values in one step, but this query is seldom much faster than the FROM clause subquery described above.

The self-exclusion join and WHERE clause subquery methods scale badly because they're O(N²). If ordering and grouping columns are indexed, they become O(N * log N), but are still substantially slower than the FROM clause subquery method.

6. Min-Concat-trick solution: Finally, here is a radically different model of the problem. It can find both within-group minima and within-group maxima in a single query. The model aggregates the concatenated within-group grouped column value and the within-group grouping column name in a single string, then uses SUBSTR() to break them apart in the result:
SELECT item, SUBSTR( MIN( CONCAT( LPAD(price,6,0),supplier) ), 7) AS MinSupplier, LEFT( MIN( CONCAT( LPAD(price,6,0),supplier) ), 6)+0 AS MinPrice, SUBSTR( MAX( CONCAT( LPAD(price,6,0),supplier) ), 7) AS MaxSupplier, LEFT( MAX( CONCAT( LPAD(price,6,0),supplier) ), 6)+0 AS MaxPrice FROM products GROUP BY item; +------+-------------+----------+-------------+----------+ | item | MinSupplier | MinPrice | MaxSupplier | MaxPrice | +------+-------------+----------+-------------+----------+ | 1 | 1 | 10 | 2 | 15 | | 2 | 2 | 18 | 1 | 21 | +------+-------------+----------+-------------+----------+
Try all solutions to find which is fastest for your version of the problem.

To find the top or bottom N per group, you might think the LIMIT clause would work, but LIMIT is limited in subqueries. See Within-group quotas.

Last updated 03 Dec 2012

Back to the top

Subscribe for Us

Feedback

Within-group aggregates with a wrinkle

We have a wages table holding wage rates by waiter and startdate, and a tips table which tracks hours worked and tips received per waiter per day. The requirement is to report wages and concurrent tips per waiter per day.


DROP TABLE IF EXISTS wages,tips; 
CREATE TABLE wages( id int, waiter int, start date, rate decimal(6,2)); 
INSERT INTO wages VALUES 
( 1, 4, '2005-01-01', 5.00 ), 
( 2, 4, '2005-03-01', 6.00 ), 
( 3, 5, '2007-01-05', 7.00 ), 
( 4, 5, '2008-03-20', 8.00 ), 
( 5, 5, '2008-04-01', 9.00 ); 
CREATE TABLE tips( 
  id int,  
  date date,

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Within-group quotas (Top N per group)

A table has multiple rows per key value, and you need to retrieve, say, the first or earliest two rows per key. For example:


DROP TABLE IF EXISTS test; 
CREATE TABLE test( id INT, entrydate DATE ); 
INSERT INTO test VALUES 
( 1, '2007-5-01' ),( 1, '2007-5-02' ),( 1, '2007-5-03' ),( 1, '2007-5-04' ), 
( 1, '2007-5-05' ),( 1, '2007-5-06' ),( 2, '2007-6-01' ),( 2, '2007-6-02' ), 
( 2, '2007-6-03' ),( 2, '2007-6-04' ),( 3, '2007-7-01' ),( 3, '2007-7-02' ), 
( 3, '2007-7-03' );

One approach is to rank rows with user variables and pick off the top two for each key in the WHERE clause:

Last updated 20 Apr 2016

Back to the top

Subscribe for Us

Feedback

Average the top 50% values per group

Each row of a games table records one game score for a team:


DROP TABLE IF EXISTS games; 
CREATE TABLE games(id INT, teamID INT, score INT); 
INSERT INTO games VALUES  
  (1,1,3),(2,1,4),(3,1,5),(4,1,6),(5,2,6), 
  (6,2,7),(7,2,8),(8,2,7),(9,2,6),(10,2,7);

How would we write a query that returns the average of the top 50% of scores per team?

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Correlation

Probably more bunk has been written about correlation than about any other statistic. We'll keep this short and straight. At its simplest, correlation is a statistical measure of non-random, linear association between pairs of values in a dataset. It's denoted by r, and varies from -1 through +1, where -1 indicates perfect inverse correlation (the regression line goes down left to right), 0 indicates no correlation (there is no regression line; it's just a scatterplot), and +1 indicates perfect direct correlation (the regression line goes up left to right).

For an example we'll use a bit of imaginary data:


drop table if exists t; 
create table t (id int, x int, y float); 
insert into t values 
(1 , 68, 4.1),(2 , 71, 4.6),(3 , 62, 3.8),(4 , 75, 4.4),(5 , 58, 3.2), 
(6 , 60, 3.1),(7 , 67, 3.8),(8 , 68, 4.1),(9 , 71, 4.3),(10, 69, 3.7), 
(11, 68, 3.5),(12, 67, 3.2),(13, 63, 3.7),(14, 62, 3.3),(15, 60, 3.4), 
(16, 63, 4.0),(17, 65, 4.1),(18, 67, 3.8),(19, 63, 3.4),(20, 61, 3.6);

Last updated 17 Jun 2012

Back to the top

Subscribe for Us

Feedback

Count unique values of one column


SELECT col_name, COUNT(*) AS frequency  
FROM tbl_name 
GROUP by col_name 
ORDER BY frequency DESC;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Median

Statistically, the median is the middle value--the value that is smaller than that found in half of all remaining rows, and larger than that found in the other half:


SELECT a.hours As Median 
FROM BulbLife As a, bulbLife AS b 
GROUP BY a.Hours 
Having Sum( Sign( 1-Sign(b.Hours-a.Hours ))) = Floor(( Count(*)+1)/2)  ;

This formula works, but it doesn't scale—it's O(n*n)—and it's awkward to use.

So we posted a MySQL implementation of Torben Mogenson's algorithm for calculating the median (http://ndevilla.free.fr/median/median/node20.html). It's said to be amongst the fastest, but it proved too slow. Then Joe Wynne offered an algorithm which looks correct and does scale. Here it is as a MySQL stored procedure:

Last updated 12 Dec 2012

Back to the top

Subscribe for Us

Feedback

Mode

Statistically, the mode is the most frequently occurring value. Given tables parent(id int) and child(pid int, cid int), where child.pid references parent.id as a foreign key, what query finds the parent.id most often represented in the child id, that is, the modal count of child.pid?


SELECT pid, COUNT(*) AS frequency 
FROM child 
GROUP BY pid 
ORDER BY frequency DESC 
LIMIT 1;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Moving average

Given a table of dates and daily values, retrieve their moving 5-day average:


DROP TABLE IF EXISTS t; 
CREATE TABLE t (dt DATE, qty INT); 
INSERT INTO t VALUES  
('2007-1-1',5),('2007-1-2',6),('2007-1-3',7),('2007-1-4',8),('2007-1-5',9), 
('2007-1-6',10),('2007-1-7',11),('2007-1-8',12),('2007-1-9',13);  

SELECT  
  a.dt,  
  a.qty, 
  Round( ( SELECT SUM(b.qty) / COUNT(b.qty) 
           FROM t AS b

Last updated 24 Oct 2016

Back to the top

Subscribe for Us

Feedback

Multiple sums across a join

You have a parties table that holds info on peoples' names etc, and a contracts table in which each row defines one contract, identifying a client as clientpartyID and a contractor as contractorpartyID, each of these a foreign key referencing parties.partyID. You want a list of parties showing how many contracts they have participated in as client, and how many they've participated in as contractor.


SELECT  
  p.partyID, 
  p.name, 
  (SELECT COUNT(*) FROM contractor_client c1 WHERE c1.clientpartyID = p.partyID ) 
  AS ClientDeals, 
  (SELECT COUNT(*) FROM contractor_client c2 WHERE c2.contractorpartyID = p.partyID) 
  AS ContractorDeals 
FROM parties p 
ORDER BY partyID;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Percentiles

In the Sakila table film, retrieve a top-down percentile ranking of film lengths:


SELECT  
  a.film_id , 
  ROUND( 100.0 * ( SELECT COUNT(*) FROM film AS b WHERE b.length <= a.length ) / total.cnt, 1 )  
  AS percentile 
FROM film a  
CROSS JOIN (  
  SELECT COUNT(*) AS cnt  
  FROM film  
) AS total 
ORDER BY percentile DESC;

Last updated 07 Oct 2009

Back to the top

Subscribe for Us

Feedback

Rank order

Without MSSQL's RANK() aggregate function, how do we display rank order in a MySQL query, for example from a table like this?


CREATE TABLE votes( name CHAR(10), votes INT ); 
INSERT INTO votes VALUES 
  ('Smith',10),('Jones',15),('White',20),('Black',40),('Green',50),('Brown',20);

The query is a two-step:
1. Join the table to itself on the value to be ranked, handling ties
2. Group and order the result of the self-join on rank:

Last updated 24 Apr 2014

Back to the top

Subscribe for Us

Feedback

Running sums, chequebooks

A user variable can maintain a per-row cumulative sum of column values. Initialise it, then adjust its value as desired in the appropriate SELECT expression:


SET @total=0; 
SELECT id, value, @total:=@total+value AS RunningSum 
FROM tbl;

Chequebook balancing programs often use this pattern. This one tracks the running balance and how much money from the most recent deposit remains:


DROP TABLE IF EXISTS chequebook;

Last updated 06 Jan 2011

Back to the top

Subscribe for Us

Feedback

Sum across categories

You often need to sum across several categories to total customer purchase amounts, salesperson sales amounts, political party election spending, etc.

For this example assume three tables: candidates, parties and ridings. You want to get the total amount spent in all ridings by every party in one output row. Here is the schema:


CREATE TABLE candidates ( 
  id int(11) NOT NULL default '0', 
  `name` char(10) , 
  riding char(12) , 
  party char(12) , 
  amt_spent decimal(10,0) NOT NULL default '0', 
  PRIMARY KEY  (id) 
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Aggregates from bands of values

Aggregating by bands of values can clarify data patterns. The banding trick is a simple transformation on the banding column:


<band width> * Floor( <banding column> / <band width> )

so to count and average scores in bands of 10, ie 0-9,10-19 and so on ...


create table scores(score int); 
insert into scores values(5),(15),(25),(35); 
SELECT 10 * FLOOR( score / 10  ) AS  Bottom,

Last updated 20 Oct 2014

Back to the top

Subscribe for Us

Feedback

History of world records

In a table of track event results, rows marked as 'WR' in the `Note` column represent world record times. How would we retrieve the history of those world records?


drop table if exists results; 
CREATE TABLE results ( 
  ID int UNSIGNED PRIMARY KEY AUTO_INCREMENT, 
  Name varchar(50) COLLATE utf8_unicode_ci NOT NULL, 
  Date date NOT NULL, 
  Time int NOT NULL, 
  Note varchar(50) COLLATE utf8_unicode_ci NOT NULL 
) ENGINE=InnoDB  DEFAULT CHARSET=utf8 COLLATE=utf8_unicode_ci AUTO_INCREMENT=14 ; 

INSERT INTO results (ID, Name, Date, Time, Note) VALUES 
(1, 'Bill', '2012-01-01', 58, 'WR'),(2, 'John', '2012-01-01', 59, ''),

Last updated 23 Oct 2012

Back to the top

Subscribe for Us

Feedback

Pivot table basics: rows to columns

From table tbl( class, member ), you want to cross-tabulate all classes with their members. In SQL terms, you aggregate members over classes. In MySQL:


SELECT class,GROUP_CONCAT(member) 
FROM tbl 
GROUP BY class;

With that simple query, you're halfway toward cross-tabulation, halfway to implementing a simple CUBE, and halfway to basic entity-attribute-value (EAV) logic. This is easier to see if we have two columns, rather than just one, to tabulate against the grouping column:


DROP TABLE IF EXISTS tbl;

Last updated 16 Feb 2011

Back to the top

Subscribe for Us

Feedback

Automate pivot table queries

You have a sales table listing product, salesperson and amount:


DROP TABLE IF EXISTS sales; 
CREATE TABLE sales ( 
  id int(11) default NULL, 
  product char(5) default NULL, 
  salesperson char(5) default NULL, 
  amount decimal(10,2) default NULL 
) ENGINE=MyISAM DEFAULT CHARSET=latin1; 
INSERT INTO sales VALUES  
  (1,'radio','bob','100.00'), 
  (2,'radio','sam','100.00'), 
  (3,'radio','sam','100.00'),

Last updated 17 Aug 2015

Back to the top

Subscribe for Us

Feedback

Group column statistics in rows

A pivot (or crosstab, or contingency) table aggregates sets of column values into rows of statistics, and pivots target value statistics on partitioning criteria defined by any available data.

Spreadsheet applications have intuitive point-and-click interfaces for generating pivot tables. RDBMSs generally do not. The task looks difficult in SQL, though, only until you have coded a few.

If you ported the Microsoft sample database Northwind to your MySQL database (as described in chapter 11 of Get It Done with MySQL), you can execute this example step by step. Even if you haven't ported Northwind, the example is easy to follow.

Amongst the tables in the Northwind database are:


employees(employeeID, lastname, firstname, ...) 
orders(orderID, customerId, employeeID, orderdate, ...)

Last updated 15 Apr 2011

Back to the top

Subscribe for Us

Feedback

Monthly expenses

You have four tables to track revenue and expenses—bankaccount, cash, accountitem, accountcategory:


drop table if exists accountitem,accountcategory,bankaccount,cash; 
create table accountitem( 
  itemid int primary key auto_increment,itemname char(32),itemcatid int 
); 
create table accountcategory( 
  categoryid int primary key auto_increment,categoryname char(32),isexpense bool 
); 
create table bankaccount( 
  id int auto_increment primary key,amount decimal(12,2),itemid int,entrydate date 
); 
create table cash(

Last updated 16 Feb 2010

Back to the top

Subscribe for Us

Feedback

Monthly sales

The pivot logic that reports expenses by month and year works just as well for sales. In the usual order entry system with customers, orders, products, orders, order items and payments (leaving out irrelevant details) ...


drop table if exists items, payments, products, orders, customers; 
create table customers(cust_id int primary key) engine=innodb; 
create table orders( 
  order_id int primary key,  
  cust_id int, 
  order_date date, 
  foreign key(cust_id) references customers(cust_id)  -- ORDER->CUST FK 
) engine=innodb; 
create table products(prod_id int primary key) engine=innodb; 
create table items( 
  item_id int primary key,

Last updated 29 Feb 2012

Back to the top

Subscribe for Us

Feedback

Pivot table using math tricks

http://en.wikibooks.org/wiki/Programming:MySQL/Pivot_table

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Pivot table with CONCAT

Here is a MySQL pivot table query for room bookings by weekday:


SELECT slot 
  , max(if(day=1, concat(subject,' ',room), '')) as day1 
  , max(if(day=2, concat(subject,' ',room), '')) as day2 
  , max(if(day=3, concat(subject,' ',room), '')) as day3 
  , max(if(day=4, concat(subject,' ',room), '')) as day4 
  , max(if(day=5, concat(subject,' ',room), '')) as day5 
from schedule 
group by slot

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Pivot table without GROUP_CONCAT

Data designs often require flexibility in numbers and names of data points per instance row: instead of saving all the data points belonging to a key value in a single row, you save each data point as a name-value pair in its own row.

Thus given table user_class(user_id INT, class_id CHAR(20), class_value CHAR(20)) with these rows:


user_id  class_id   class_value 
1        firstname  Rogier 
1        lastname   Marat 
2        firstname  Jean 
2        lastname   Smith

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Unpivot a table

http://stackoverflow.com/questions/15184381/mysql-turn-table-into-different-table
http://stackoverflow.com/questions/25319348/unpivot-multiple-columns-in-mysql
http://stackoverflow.com/questions/17361150/sql-unpivot-table

Last updated 01 Sep 2016

Back to the top

Subscribe for Us

Feedback

Backslashes in data

Backslashes multiply weirdly:


SELECT 'a\b' RLIKE 'a\b';

returns 1, as does...


SELECT 'a\\b' RLIKE 'a\\\\b';

because in a pair of backslashes, the second is not escaped by the first, so to compare two literals you double each backslash in the RLIKE argument. But if you are querying a table for such a string from the MySQL client, this doubling happens twice--once in the client, and once in the database--so to find a column value matching 'a\\b', you need to write...


SELECT desc FROM xxx WHERE desc RLIKE 'aa\\\\\\\\bb';

That's eight backslashes to match two!

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Compare data in two tables

This query UNIONs queries for matching column names from two tables, and keeps just those rows which occur once in the union. Those are the rows with unmatched data. Customise your column list { id, col1, col2, col3 ...} as desired, but usually you'll want it to start with the primary key:


SELECT MIN(TableName) as TableName, id, col1, col2, col3, ... 
FROM ( 
  SELECT 'Table a' as TableName, a.id, a.col1, a.col2, a.col3, ... 
  FROM a 
  UNION ALL 
  SELECT 'Table b' as TableName, b.id, b.col1, b.col2, b.col3, ... 
  FROM b 
) AS tmp 
GROUP BY id, col1, col2, col3, ... 
HAVING COUNT(*) = 1 
ORDER BY 1;

Last updated 26 Nov 2010

Back to the top

Subscribe for Us

Feedback

Show rows where column value changed

SQL is set-oriented, but it can solve row-by-row problems. Suppose you need to retrieve only the rows that differ from immediately previous rows given some ordering spec:


drop table if exists t; 
create table t ( 
  p char(3), 
  d date 
); 
insert into t values 
('50%','2008-05-01'), 
('30%','2008-05-02'), 
('30%','2008-05-03'), 
('50%','2008-05-04'), 
('50%','2008-05-05'),

Last updated 11 Aug 2010

Back to the top

Subscribe for Us

Feedback

Using bit operators

If you store ipaddress values in a 32-bit unsigned int column, you can retrieve octets with bit-shifting:


select  
  ipAddress,  
  (ipAddress >> 24) as firstOctet, 
  (ipAddress>>16) & 255 as secondOctet, 
  (ipAddress>>8) & 255 as thirdOctet, 
  ipAddress & 255 as fourthOctet 
from ...

Last updated 17 Feb 2011

Back to the top

Subscribe for Us

Feedback

Compare structures of two tables

To compare columns by name and ordinal position in tables test.t1 and test.t2:


SELECT 
  MIN(TableName) AS 'Table', 
  column_name AS 'Column', 
  ordinal_position AS 'Position'
FROM ( 
  SELECT 
    't1' as TableName, 
    column_name, 
    ordinal_position 
  FROM information_schema.columns AS i1 
  WHERE table_schema='test' AND table_name='t1'

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Compare two databases

One of EF Codd's rules for relational databases is the no-back-door rule: all info about tables should be accessible only by a query on tables. Since version 5, the MySQL implementation of information_schema (I_S) helps meet Codd's requirement. I_S supplies metadata in tables, so it's the first place to look for how to compare the structures of two databases.

Elsewhere on this page there is a simple query template for comparing data in two structurally similar tables:


SELECT MIN(TableName) as TableName, id, col1, col2, col3, ... 
FROM ( 
  SELECT 'Table a' as TableName, a.id, a.col1, a.col2, a.col3, ... 
  FROM a 
  UNION ALL 
  SELECT 'Table b' as TableName, b.id, b.col1, b.col2, b.col3, ... 
  FROM b 
) AS tmp

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Database size


SELECT  
  table_schema AS 'Db Name', 
  Round( Sum( data_length + index_length ) / 1024 / 1024, 3 ) AS 'Db Size (MB)', 
  Round( Sum( data_free ) / 1024 / 1024, 3 ) AS 'Free Space (MB)'
FROM information_schema.tables 
GROUP BY table_schema ;

Last updated 18 Jun 2010

Back to the top

Subscribe for Us

Feedback

Find the size of all databases on the server

This is based on a query Mark Leith posted to the MySQL General Discussion list.


DROP VIEW IF EXISTS dbsize; 
CREATE VIEW dbsize AS 
SELECT  
  s.schema_name AS 'Schema', 
  SUM(t.data_length) AS Data, 
  SUM( t.index_length ) AS Indexes, 
  SUM(t.data_length) + SUM(t.index_length) AS 'Mb Used', 
  IF(SUM(t.data_free)=0,'',SUM(t.data_free)) As 'Mb Free', 
  IF( SUM(t.data_free)=0, 
      '', 
      100 * (SUM(t.data_length) + SUM(t.index_length)) / ((SUM(t.data_length)+SUM(t.index_length) + SUM(IFNULL(t.data_free,0))) ) 
    ) AS 'Pct Used', 
  COUNT(table_name) AS Tables 
FROM information_schema.schemata s 
LEFT JOIN information_schema.tables t ON s.schema_name = t.table_schema 
GROUP BY s.schema_name  
WITH ROLLUP

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

List databases, tables, columns


SELECT 
  t.table_schema AS 'Database'
  ,t.table_name AS 'Table'
  ,t.table_type AS 'Table Type'
  ,c.column_name AS 'Column'
  ,c.data_type AS 'Data Type'
FROM information_schema.tables t 
JOIN information_schema.columns c ON t.table_schema = c.table_schema AND t.table_name = c.table_name 
WHERE t.table_schema NOT IN( 'mysql','information_schema') 
ORDER BY t.table_schema,t.table_type,t.table_name,c.ordinal_position;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

List differences between two databases

Given two databases named @db1 and @db2:


SELECT  
  MIN(table_name) as TableName,  
  table_catalog,table_schema,table_name,column_name, 
  ordinal_position,column_default,is_nullable, 
  data_type,character_maximum_length,character_octet_length, 
  numeric_precision,numeric_scale,character_set_name, 
  collation_name,column_type,column_key, 
  extra,privileges,column_comment 
FROM ( 
  SELECT 'Table a' as TableName,  
  table_catalog,table_schema,table_name,column_name,

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

List users of a database


DROP PROCEDURE IF EXISTS ListDbUsers; 
DELIMITER | 
CREATE PROCEDURE ListDbUsers( dbname CHAR(64) ) 
  SELECT host,user 
  FROM mysql.user 
  WHERE Select_priv = 'Y'  
       OR Insert_priv = 'Y'  
     OR Update_priv = 'Y'  
     OR Delete_priv = 'Y'  
     OR Create_priv = 'Y'  
     OR Drop_priv = 'Y'  
     OR Reload_priv = 'Y'  
     OR Shutdown_priv = 'Y'

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Rename Database

Sometimes it's necessary to rename a database. There are two crucial reasons there's no such command in MySQL: there's no way to lock a whole database, and information_schema isn't transactional.

MySQL 5.1.7 introduced a RENAME DATABASE command, but the command left several unchanged database objects behind, and was found to lose data, so it was dropped in 5.1.23.

The operating system's view of a database is that it's just another folder on the disk. That may tempt you to think you could just rename it. Don't.

Renaming a database is perfectly safe if and only if:

it has no tables with foreign keys that reference, or are referenced by, a table in another database;
it has no procedures, functions or triggers that reference objects in another database;
no other database has tables, functions, procedures or triggers that reference this database.

So there is one safe way to rename a database: dump it completely ...

Last updated 29 Nov 2011

Back to the top

Subscribe for Us

Feedback

Replace text in all string columns of a database

Generate SQL to replace 'old_text' with 'new_text' in all character columns in 'dbname':


Select Concat( 'update ', table_schema, '.', table_name,  
               ' set ', column_name,  
               '=replace(', column_name, ',''old_text'',''new_text'');'
             ) 
From information_schema.columns 
Where (data_type Like '%char%' or data_type like '%text' or data_type like '%binary') 
  And table_schema = 'dbname';

Last updated 23 May 2010

Back to the top

Subscribe for Us

Feedback

Foreign key basics

Referential integrity (RI) is data consistency between related tables. ACID-compliant database engines like InnoDB provide foreign keys to automate RI maintenance. Simple example: customers have orders, which have items; by defining a customers foreign key in the orders table and an orders foreign key in the items table, you ensure that no orders or items row gets orphaned.

SQL implements FKs as key objects defined by CREATE TABLE and ALTER TABLE. MySQL syntax is:


[CONSTRAINT symbol] FOREIGN KEY [constraintID] (keycolumnname, ...) 
REFERENCES tbl_name (keycolumnname, ...) 
[ ON DELETE { RESTRICT | CASCADE | SET NULL | NO ACTION } ] 
[ ON UPDATE { RESTRICT | CASCADE | SET NULL | NO ACTION } ]

To drop a FK in MySQL you need to know the CONSTRAINT symbol, so it's good practice to adopt a standard naming convention and use it here.

Last updated 09 Oct 2013

Back to the top

Subscribe for Us

Feedback

Change or drop a foreign key

To change a foreign key, first drop it, then declare the new, revised foreign key. The syntax for declaring a foreign key is ...


[CONSTRAINT [constraint_name]]  
FOREIGN KEY [key_name] (keycol_name,...) reference_definition

and the syntax for dropping one is ...


DROP FOREIGN KEY constraint_name

Notice that you can omit the CONSTRAINT when you declare a foreign key, but the only way to DROP a foreign key is to reference it by the constraint_name which you probably never specified!

There should be a circle of hell reserved for designers who build inconsistencies like this into their tools. The only way round this one is to run SHOW CREATE TABLE to find out what the foreign key's constraint_name is, so you can write the DROP statement. Here is a wee test case:

drop table if exists a,b; create table a(i int primary key)engine=innodb; create table b(i int,foreign key(i) references a(i)) engine=innodb; show create table\G  CREATE TABLE `b` (   `i` int(11) DEFAULT NULL,   KEY `i` (`i`),   CONSTRAINT `b_ibfk_1` FOREIGN KEY (`i`) REFERENCES `a` (`i`) ) ENGINE=InnoDB DEFAULT CHARSET=latin1  -- drop and recreate the FK: alter table b drop foreign key b_ibfk_1; alter table b add foreign key(i) references a(i) on update cascade; show create table b\G  Create Table: CREATE TABLE `b` (   `i` int(11) DEFAULT NULL,   KEY `i` (`i`),   CONSTRAINT `b_ibfk_1` FOREIGN KEY (`i`) REFERENCES `a` (`i`) ON UPDATE CASCADE ) ENGINE=InnoDB DEFAULT CHARSET=latin1  drop table a,b;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Find child tables

Starting with MySQL 5, you can find all tables which have foreign key references to a given table with an information_schema query, here encapsulated in a stored procedure which takes two arguments, a database name and a table name. When the table argument is blank or NULL, the procedure returns all parent-child links where the parent table is in the specified database; otherwise it returns parent-child links for the specified parent table:


DROP PROCEDURE IF EXISTS ListChildren; 
DELIMITER | 
CREATE PROCEDURE ListChildren( pdb CHAR(64), ptbl CHAR(64) ) 
BEGIN 
  IF ptbl = '' OR ptbl IS NULL THEN 
    SELECT  
      c.table_schema as 'Parent Schema', 
      u.referenced_table_name as 'Parent Table', 
      u.referenced_column_name as 'Parent Column', 
      u.table_schema as 'Child Schema', 
      u.table_name as 'Child Table',

Last updated 28 Jan 2011

Back to the top

Subscribe for Us

Feedback

Find cross-database foreign keys


SELECT  
  u.table_schema as 'RefSchema', 
  u.table_name as 'RefTable', 
  u.column_name as 'RefColumn', 
  u.referenced_table_schema as 'Schema', 
  u.referenced_table_name as 'Table', 
  u.referenced_column_name as 'Column'
FROM information_schema.table_constraints AS c 
JOIN information_schema.key_column_usage AS u 
USING( constraint_schema, constraint_name ) 
WHERE c.constraint_type='FOREIGN KEY' AND u.table_schema <> u.referenced_table_schema;

To find them for a particular database, add the WHERE condition:


AND 'dbname' IN(u.table_schema, u.referenced_table_schema)

Last updated 27 Nov 2011

Back to the top

Subscribe for Us

Feedback

Find parent tables

List tables which are referenced by foreign key constraints in a given table.This is a simple query on two information_schema tables: table_constraints and key_column_usage. It is easy to parameterise, so we show it in stored procedures. The first sproc lists all foreign key references in a database. The second lists all foreign key references for a table.


CREATE PROCEDURE ListParentsForDb( pdb CHAR(64) ) 
  SELECT  
   u.table_schema AS 'Schema', 
   u.table_name AS 'Table', 
   u.column_name AS 'Key', 
   u.referenced_table_schema AS 'Parent Schema', 
   u.referenced_table_name AS 'Parent table', 
   u.referenced_column_name AS 'Parent key'
  FROM information_schema.table_constraints AS c 
  JOIN information_schema.key_column_usage AS u 
  USING( constraint_schema, constraint_name )

Last updated 16 Mar 2013

Back to the top

Subscribe for Us

Feedback

Add auto-incrementing primary key to a table

The steps are: (i) recreate the table, populating a new column from an incrementing user variable, then (ii) alter the table to add auto_increment and primary key properties to the new column. So given table t with columns named `dt` and `observed`...


DROP TABLE IF EXISTS t2; 
SET @id=0; 
CREATE TABLE t2 
  SELECT @id:=@id+1 AS id, dt, observed FROM t ORDER BY dt; 
ALTER TABLE t2  
  MODIFY id INT AUTO_INCREMENT PRIMARY KEY; 
DROP TABLE t; 
RENAME TABLE t2 TO t;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Auto-increment: reset next value

ALTER TABLE tbl SET AUTO_INCREMENT=val;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Find primary key of a table

To retrieve primary keys of db.tbl...


SELECT k.column_name  
FROM information_schema.table_constraints t  
JOIN information_schema.key_column_usage k  
USING (constraint_name,table_schema,table_name) 
WHERE t.constraint_type='PRIMARY KEY'  
  AND t.table_schema='db'
  AND t.table_name='tbl'

For pre-5 versions of MySQL:


SHOW INDEX FROM tbl 
WHERE key_name='primary';

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Show Create Trigger

MySQL added a Show Create Trigger command in 5.1.21. If you use an earlier MySQL version, here is a stored procedure which behaves like Show Create Trigger:


DROP PROCEDURE IF EXISTS ShowCreateTrigger;  
DELIMITER go 
CREATE PROCEDURE ShowCreateTrigger( IN db CHAR(64), IN tbl CHAR(64) )  
  BEGIN  
    SELECT   
      CONCAT(  
        'CREATE TRIGGER ',trigger_name, CHAR(10),   
        action_timing,' ', event_manipulation, CHAR(10),  
        'ON ',event_object_schema,'.',event_object_table, CHAR(10),  
        'FOR EACH ROW', CHAR(10),  
        action_statement, CHAR(10)

Last updated 08 Jun 2015

Back to the top

Subscribe for Us

Feedback

Show Table Status equivalent from information_schema

Fill in schema and table names in ...


SELECT 
  table_name, 
  engine, 
  version, 
  row_format, 
  table_rows, 
  avg_row_length, 
  data_length, 
  max_data_length, 
  index_length, 
  data_free, 
  auto_increment, 
  create_time, 
  update_time, 
  check_time, 
  table_collation, 
  checksum, 
  create_options, 
  table_comment 
FROM information_schema.tables 
where table_schema='???' AND table_name='???';

Last updated 09 Oct 2009

Back to the top

Subscribe for Us

Feedback

Show Tables

The MySQL SHOW TABLES command is fine, but sometimes we want a little more information.

This simple stored procedure lists the table name, engine type, version, collation and rowcount for every table in a database. (Individual databases come and go, so we keep all such database-wide stored routines in a system database.)


DROP PROCEDURE IF EXISTS showtables; 
CREATE PROCEDURE showtables() 
  SELECT 
    table_name AS 'Table', 
    IFNULL(engine, 'VIEW') AS Engine, 
    version AS Version, 
    table_collation AS Collation, 
    table_rows AS Rows

Last updated 22 Nov 2009

Back to the top

Subscribe for Us

Feedback

Age in years

Given a birthdate in @dob, here are two simple formulae for age in years:


Date_format( From_Days( To_Days(Curdate()) - To_Days(@dob) ), '%Y' ) + 0 

Year(Curdate()) - Year(@dob) - ( Right(Curdate(),5) < Right(@dob,5) )

and here is one for age in years to two decimal places, ignoring day of month:


Round((((Year(now()) - Year(@dob)))*12 + (((Month(now()) - Month(@dob)))))/12, 2)

Last updated 27 Dec 2012

Back to the top

Subscribe for Us

Feedback

Compute date from year, week number and weekday


SET @yr=2012, @wk=26, @day=0; 
SELECT Str_To_Date( Concat(@yr,'-',@wk,'-',If(@day=7,0,@day) ), '%Y-%U-%w' ) AS Date; 
+------------+ 
| Date       | 
+------------+ 
| 2012-06-24 | 
+------------+

Last updated 01 Jul 2012

Back to the top

Subscribe for Us

Feedback

Count business days between two dates

The simplest support for counting business days between any two dates is a calendar table calendar(d date,isholiday bool) populated for all days in all possibly relevant years. Then the following query gives the inclusive number of business days between dates dStart and dStop:


SELECT COUNT(*) 
FROM calendar 
WHERE d BETWEEN dStart AND dStop  
  AND DAYOFWEEK(d) NOT IN(1,7) 
  AND isholiday=0;

If that solution is not available, you have to do with a weekday count, which this function (corrected 6 Jul 2009) computes:

Last updated 10 Jul 2009

Back to the top

Subscribe for Us

Feedback

Count Tuesdays between two dates

Date arithmetic is deceptively hard. One way to appreciate the difficulties is to read Chapter 21 in our book. Another is to try to calculate the number of Tuesdays (or another weekday) between two dates. It's not a back-of-the-napkin problem.

An earlier formula we had for this problem sometimes gave incorrect results. As a debugging aid, we wrote a brute force calculator for the problem:


SET GLOBAL log_bin_trust_function_creators=1; 
DROP FUNCTION IF EXISTS DayCount; 
DELIMITER | 

CREATE FUNCTION DayCount( d1 DATE, d2 DATE, daynum SMALLINT ) RETURNS INT 
BEGIN 
  DECLARE days INT DEFAULT 0; 
  IF D1 IS NOT NULL AND D2 IS NOT NULL THEN

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Date of first day of next month

Date of first day of next month:


concat(left(curdate() + interval 1 month, 8), '-01');

Date of first day of previous month:


concat(left(curdate() - interval 1 month, 8), '-01');

Last updated 29 Mar 2012

Back to the top

Subscribe for Us

Feedback

Date of first Friday of next month

Assuming a calendar table calendar(date DATE) with one row per date through the relevant period...


SET @d = NOW(); 
SELECT MIN(date) AS 1stFridayOfMonth 
FROM calendar  
WHERE YEAR(date) = IF( MONTH(@d) = 12, 1+YEAR(@d), YEAR(@d) ) 
  AND MONTH(date) = IF( MONTH(@d) = 12, 1, MONTH(@d) + 1 ) 
  AND WEEKDAY(date)=4;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Date of Monday in a given week of the year

Given a valid year value @y and a week value @w between 1 and 53, this formula gives the date of Monday in week @w of year @y:


adddate( makedate(@y,@w*7), interval 2-dayofweek(makedate(@y,7)) day )

Last updated 29 Nov 2011

Back to the top

Subscribe for Us

Feedback

Date of Monday of this week


set @d = '2011-11-29'; 
select adddate( date(@d), interval 2-dayofweek(@d) day ) as 1stdayofweek; 
+---------------+ 
| 1stydayofweek | 
+---------------+ 
| 2011-11-28    | 
+---------------+

Last updated 29 Nov 2011

Back to the top

Subscribe for Us

Feedback

Datetime difference

Find the difference between two datetime values in seconds, minutes, hours or days. If dt1 and dt2 are datetime values of the form 'yyyy-mm-dd hh:mm:ss', the number of seconds between dt1 and dt2 is


UNIX_TIMESTAMP( dt2 ) - UNIX_TIMESTAMP( dt1 )

To get the number of minutes divide by 60, for the number of hours divide by 3600, and for the number of days, divide by 3600 * 24.

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Duration in years, months, days and time


DROP FUNCTION IF EXISTS PeriodLen; 
DROP FUNCTION IF EXISTS NumLabel; 
DELIMITER go 
CREATE FUNCTION PeriodLen( dt1 datetime, dt2 datetime ) RETURNS CHAR(128) 
BEGIN 
  DECLARE yy,m0,mm,d0,dd,hh,mi,ss,t1 BIGINT; 
  DECLARE dtmp DATETIME; 
  DECLARE t0 TIMESTAMP; 
  SET yy = TIMESTAMPDIFF(YEAR,dt1,dt2); 
  SET m0 = TIMESTAMPDIFF(MONTH,dt1,dt2); 
  SET mm = m0 MOD 12; 
  SET dtmp = ADDDATE(dt1, interval m0 MONTH); 
  SET d0 = TIMESTAMPDIFF(DAY,dt1,dt2);

Last updated 20 Jun 2012

Back to the top

Subscribe for Us

Feedback

Julian date

Unix_Timestamp( datetimevalue ) / (60*60*24) ) + 2440587.5

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Last business day before a reference date

Given a date value in datetimecol ...


SELECT  
  @refday := datetimecol, 
  @dow := DAYOFWEEK(@refday) AS DOW, 
  @subtract := IF( @dow = 1, 2, IF( @dow = 2, 3, 1 )) AS MINUS,  
  @refday - INTERVAL @subtract DAY AS LastBizDay 
FROM ... etc

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Make a calendar table

You need a calendar table for joins to datetime data in other tables:


create table calendar ( dt datetime primary key );

An elegant method of generating any desired number of sequential values, posted by Giuseppe Maxia on his blog, is ...

Create three dummy rows in a View.
Cross join them to make 10 dummy rows.
Cross join those to make 100, 1,000 or however many you need.

So to give the calendar table a million rows at one-hour intervals starting on 1 Jan 1970:

Last updated 13 Jun 2013

Back to the top

Subscribe for Us

Feedback

Scope to the week of a given date

To scope a query to the calendar week (Sunday through Saturday) of a date value @d, write ...


... WHERE d BETWEEN AddDate(@d,-DayOfWeek(@d)+1) and AddDate(@d,7-DayOfWeek(@d)) ...

Last updated 29 Dec 2009

Back to the top

Subscribe for Us

Feedback

Sum accumulated time by date

You track resource booking periods. You need a query to report daily usage for a given resource.

First the problem of calculating per-diem usage. Call the starting datetime of a booked period pStart, and its ending datetime pEnd. Then for a given date pDate, if the period began before pDate. then pDate usage begins at 00:00:00, otherwise it starts at pStart; likewise if the period extends past pDate, then pDate usage ends at midnight on pDate, otherwise it ends at pEnd. Therefore the period begins at...


IF( pStart < pDate, CAST(pDate AS DATETIME ), pStart )

and ends at...

Last updated 20 Oct 2014

Back to the top

Subscribe for Us

Feedback

Sum booking days for a given month

A table tracks beginning and ending booking dates dfrom and dto. How to sum booking durations in days for a given month, say July?

A booking may begin and end in the target month, or begin, or end, or neither:


select  
  sum( 
    case  
       when month(dfrom)=7 and month(dfrom)=month(dto) then datediff(dto,dfrom) 
       when month(dfrom)=7                             then datediff(last_day(dfrom),dfrom) 
       when month(dto)=7                               then datediff(dto,date_format(dto,'%Y-%m-01')) 
       else 0 
     end 
  ) as DaysInJuly 
  from ...

Last updated 10 Nov 2014

Back to the top

Subscribe for Us

Feedback

Sum time values


SELECT SEC_TO_TIME( SUM( TIME_TO_SEC( time_col ))) AS total_time  
FROM tbl;

Summing values like '12:65:23' produces meaningless results.

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

The date of next Thursday

Given a date and its weekday number (1=Sunday, ..., 7=Saturday), there are three possibilities:
1. Today is Thursday: then next Thursday is 7 days from now.
2. Today is before Thursday: then next Thursday is (5 minus today's weekday number) from now.
3. Today is after Thursday: then next Thursday is 7 + (5 minus today's weekday number).


set @d=curdate(); 
set @n = dayofweek(curdate()); 
select  
  @d:=adddate(curdate(),0) as date,  
  @n:=dayofweek(adddate(curdate(),0)) as weekday,

Last updated 02 Dec 2009

Back to the top

Subscribe for Us

Feedback

Track when a value changed

You have a table that tracks a value and the time when the value was measured ...


drop table if exists changes; 
create table changes(time time,value int); 
insert into changes values 
('00:00', 0 ), 
('01:05', 1 ), 
('01:09', 1 ), 
('01:45', 1 ), 
('02:24', 0 ), 
('12:20', 1 ), 
('12:40', 0 ), 
('14:32', 0 ),

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

What month does a week fall in?

"The" month of a week is ambiguous if the week straddles months. If we adopt the convention that the month of a week is the month of its beginning Sunday, then on 29 Nov 2009 ...


SET @weekno = Month CurDate() ); 
SET @date = AddDate('2009-01-01', 7*@weekno ); 
SET @day = DayOfWeek( @date ); 
SET @datecomp = IF( @day = 1, @date, AddDate( @date, 1-@day )); 
SELECT @date,@day,@datecomp,Month(@datecomp) AS month; 
+------------+------+------------+-------+ 
| @date      | @day | @datecomp  | month | 
+------------+------+------------+-------+ 
| 2009-12-03 |    5 | 2009-11-29 |    11 | 
+------------+------+------------+-------+

Last updated 30 Nov 2009

Back to the top

Subscribe for Us

Feedback

YearMonth()

We often need to compute datetimes based on year and month. This tiny function simplifies such queries:


set global log_bin_trust_function_creators=1; 
create function yearmonth(d date) returns int 
return 100*year(d)+month(d);

Then to find date values within the three-month period bounded by the first day of last month and the last day of next month, write ...


select d 
from tbl 
where yearmonth(d) between yearmonth(curdate()-interval 1 month) and yearmonth(curdate()+interval 1 month);

Last updated 23 Aug 2010

Back to the top

Subscribe for Us

Feedback

Audit trails and point-in-time architecture

http://www.artfulsoftware.com/infotree/Transaction time validity in MySQL.pdf

Last updated 31 Jul 2012

Back to the top

Subscribe for Us

Feedback

Find overlapping periods

You have a table of visits, and you would like to display the time periods during which there are visit time overlaps.


drop table if exists visits; 
create table visits(id int primary key,start datetime,end datetime); 
insert into visits values 
(1, '2008-09-01 15:01', '2008-09-01 15:04'), 
(2, '2008-09-01 15:02', '2008-09-01 15:09'), 
(3, '2008-09-01 15:12', '2008-09-01 15:15'), 
(4, '2008-09-01 16:11', '2008-09-01 16:23'), 
(5, '2008-09-01 16:19', '2008-09-01 16:25'), 
(6, '2008-09-01 17:52', '2008-09-01 17:59'), 
(7, '2008-09-01 18:18', '2008-09-01 18:22'), 
(8, '2008-09-01 16:20', '2008-09-01 16:22'),

Last updated 05 Sep 2009

Back to the top

Subscribe for Us

Feedback

Find sequenced duplicates

A table that tracks time periods may require period uniqueness. That means it has no sequenced duplicates.

If a table has columns processID, start_date and end_date, those three columns are period unique if there exists no pair of rows with the same processID and overlapping start_date and end_date values. If there is such a pair of rows, the table exhibits sequenced duplication.

Another way of saying it: if an instant is the smallest datetime unit of start_date and end_date columns, then if there are no sequenced duplicates, there is exactly one processID value at any instant.

Here is a query to find sequenced duplicates for those columns:


SELECT t.processid  
FROM tbl t  
WHERE EXISTS (  
  SELECT * FROM tbl AS t3  
  WHERE t3.processid IS NULL  
)  
OR EXISTS (  
  SELECT * FROM tbl AS t1  
  WHERE 1 < (  
    SELECT COUNT(processid)  
    FROM tbl AS t2  
    WHERE t1.processid&nnbsp;= t2.processid  
      AND t1.start_date < t2.end_date  
      AND t2.start_date < t1.end_date  
  )  
);

Last updated 26 Dec 2012

Back to the top

Subscribe for Us

Feedback

In or out at a given date and time?

Employees punch in and out. You track these events with ...


drop table if exists tbl; 
create table tbl( empno smallint, clockdate date, clocktime time, clocktype char(1) ); 
insert into tbl values  
(1,  '2014-05-15', '09:00:00', 'I' ), 
(1,  '2014-05-15', '11:00:00', 'O' ), 
(1,  '2014-05-15', '12:30:00', 'I' ), 
(1,  '2014-05-15', '19:00:00', 'O' );

Was employee 1 in or out at 12:30pm on 15 May 2014? Use a self-join to play the in and out events against the given datetime ...


select if(count(1),'Yes','No' ) as InStore  
from tbl a 
join tbl b using(empno, clockdate) 
where empno=1 
  and a.clockdate='2014-5-15'
  and a.clocktime<='12:30:00' and a.clocktype='I'
  and b.clocktime>'12:30:00' and b.clocktype='O'; 
+---------+ 
| InStore | 
+---------+ 
| Yes     | 
+---------+

Last updated 10 Jun 2014

Back to the top

Subscribe for Us

Feedback

Peak visit counts by datetime period

You have a visits table (id int, start datetime, end datetime), and you wish to track peak visit counts. A simple solution is to self-join on non-matching IDs and overlapping visit times, group by ID, then order by the resulting counts:


SELECT a.id,group_concat(b.id) as Overlaps, count(b.id)+1 as OverlapCount 
FROM visits a 
JOIN visits b on a.id < b.id and a.start < b.end and b.start < a.end 
GROUP BY a.id 
ORDER BY OverlapCount DESC;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Sum for time periods

A table tracks attendance at some location:


drop table if exists t; 
create table t(interval_id int,start datetime,end datetime, att int); 
insert into t values 
(1,'2007-01-01 08:00:00','2007-01-01 12:00:00',5 ), 
(2,'2007-01-01 13:00:00','2007-01-01 17:00:00',10), 
(3,'2007-01-01 10:00:00','2007-01-01 15:00:00',15), 
(4,'2007-01-01 14:00:00','2007-03-07 19:00:00',20); 
select * from t; 
+-------------+---------------------+---------------------+------+ 
| interval_id | start               | end                 | att  | 
+-------------+---------------------+---------------------+------+

Last updated 03 Mar 2015

Back to the top

Subscribe for Us

Feedback

Appointments available

Given a clinic of physicians, patients and appointments, how to find an available appointment time for a given physician?

This is a variant of the [Not] Exists query pattern. Though we can write it with subqueries, performance will be crisper with a join. But finding data that is not there requires a join to data which is there. So in addition to tables for appointments, doctors and patients, we need a table of all possible appointment datetimes. Here's a schema illustrating the idea ...


CREATE TABLE a_dt (         -- POSSIBLE APPOINTMENT DATES AND TIMES 
  d DATE, 
  t TIME 
); 
CREATE TABLE a_drs (        -- DOCTORS 
  did INT                   -- doctor id 
); 
CREATE TABLE a_pts (        -- PATIENTS

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Find available reservation periods

Given a bookings table where each row specifies one reservation period for one property, find the unbooked periods for a given property:


CREATE TABLE bookings( ID int, propertyID int, startDate date, endDate date ); 
INSERT INTO bookings VALUES  
  (1,1,'2007-1-1','2007-1.15'), 
  (2,1,'2007-1-20','2007-1.31'), 
  (3,1,'2007-2-10','2007-2-17'); 
SELECT * FROM bookings;  
+------+------------+------------+------------+ 
| ID   | propertyID | startDate  | endDate    | 
+------+------------+------------+------------+ 
|    1 |          1 | 2007-01-01 | 2007-01-15 | 
|    2 |          1 | 2007-01-20 | 2007-01-31 |

Last updated 05 Mar 2015

Back to the top

Subscribe for Us

Feedback

Game schedule

List all possible home-away encounters of teams listed in a table.


SELECT t1.name AS Visiting, 
       t2.name AS Home 
FROM teams AS t1 
STRAIGHT_JOIN teams AS t2 
WHERE t1.ID <> t2.ID;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Is a given booking period available?

You rent vacation properties, tracking bookings with a table like this:


CREATE TABLE bookings( ID int, propertyID int, startDate date, endDate date ); 
INSERT INTO bookings VALUES (1,1,'2007-1-1','2007-1.15'),(2,1,'2007-1-20','2007-1.31'); 
SELECT * FROM bookings; 
+------+------------+------------+------------+ 
| ID   | propertyID | startDate  | endDate    | 
+------+------------+------------+------------+ 
|    1 |          1 | 2007-01-01 | 2007-01-15 | 
|    2 |          1 | 2007-01-20 | 2007-01-31 | 
+------+------------+------------+------------+

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Pivot table schedule

You have a schedule table (period, day, subject, room) with a primary key period,day to avoid duplicate bookings. You wish to display the schedule as periods, subjects and rooms in rows, and days of the week in columns.


SELECT  
  period, 
  MAX(IF(day=1, CONCAT(subject,' ',room), '')) AS Mon, 
  MAX(IF(day=2, CONCAT(subject,' ',room), '')) AS Tue, 
  MAX(IF(day=3, CONCAT(subject,' ',room), '')) AS Wed, 
  MAX(IF(day=4, CONCAT(subject,' ',room), '')) AS Thu, 
  MAX(IF(day=5, CONCAT(subject,' ',room), '')) AS Fri 
FROM schedule 
GROUP BY period

MAX() chooses existing over blank entries, and GROUP BY lines everything up on the same row.

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Display column values which occur N times


SELECT id 
FROM tbl 
GROUP BY id 
HAVING COUNT(*) = N;

Change the HAVING condition to >1 to list duplicate values, etc.

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Display every Nth row

Display every Nth row in tbl where id is sequential in MySQL before version 4.1:


SELECT id 
FROM tbl 
GROUP BY id  
HAVING MOD(id, N) = 0;

Last updated 02 Sep 2015

Back to the top

Subscribe for Us

Feedback

Trees, networks and parts explosions in MySQL

http://www.artfulsoftware.com/mysqlbook/sampler/mysqled1ch20.html

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Dijkstra’s shortest path algorithm

Given a table of source-to-destination paths, each of whose nodes references a row in a nodes table, how do we find the shortest path from one node to another?

One answer is Dijkstra's algorithm (http://en.wikipedia.org/wiki/Dijkstra's_algorithm). Peter Larsson has posted a SQL Server implementation of it on the SQL Team Forum. Here is a MySQL implementation.

The DDL:


DROP TABLE IF EXISTS dijnodes,dijpaths; 
CREATE TABLE dijnodes ( 
  nodeID int PRIMARY KEY AUTO_INCREMENT NOT NULL, 
  nodename varchar (20) NOT NULL, 
  cost int NULL,

Last updated 26 Dec 2012

Back to the top

Subscribe for Us

Feedback

Multiple trees in one table

Multiple trees can bury you in table glut. Can we combine them in one table and still query them conveniently? Yes. What's more, it's useful to combine edge list and nested sets representations of the same trees in one table. Here is a table definition for the job:


CREATE TABLE trees ( 
  root int(11) DEFAULT 0,         -- to which tree does this node belong? 
  nodeID int(11) NOT NULL,        -- edge list representation of this node 
  parentID int(11) DEFAULT NULL, 
  level smallint(6) DEFAULT 0,    -- depth of this node in this tree 
  lft int(11) DEFAULT 0,          -- nested sets representation of this node 
  rgt int(11) DEFAULT 0, 
  PRIMARY KEY (root,nodeID) 
) ENGINE=InnoDB;

Last updated 05 Sep 2011

Back to the top

Subscribe for Us

Feedback

Tree query performance

http://www.artfulsoftware.com/infotree/treequeryperformance.pdf

Last updated 04 Apr 2010

Back to the top

Subscribe for Us

Feedback

Trees of known depth

A tree is a hierarchy where each node except the root has one parent. A parent-child link is an edge. Edges in a tree of known depth can be queried with n-1 self-joins where n is the number of edges from top to bottom. Here is a simple example of depth 2:


drop table if exists t; 
create table t(id int, parent int, ord int, title char(20)); 
insert into t values 
(1, 0, 0, 'Root'), 
(2, 1, 0, 'Home'), 
(3, 1, 1, 'Projects'), 
(5, 1, 2, 'Secret area'), 
(4, 1, 3, 'Tutorials'), 
(8, 1, 4, 'Example'), 
(6, 4, 0, 'Computing'), 
(7, 4, 1, 'Life');

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Approximate joins

There are two main ways to reconcile payments against charges:

Open Item: match payments against individual charges, typically by carrying the charge number in the payments table
Statement: list and sum all charges and all payments, and show the difference as the outstanding balance.

The Open Item method needs a foolproof way to match payments to charges, but what if the customer neglected to return a copy of the invoice, or to write the invoice number on the cheque? Reconciliation staff spend much of their time resolving such problems.

Can we help? Yes! It won't be entirely foolproof, but it will drastically cut down the onerous work of reconciliation.

Here is DDL for a test case:


CREATE SCHEMA approx; 
USE approx;

Last updated 05 May 2010

Back to the top

Subscribe for Us

Feedback

Cascading JOINs

Show parents, children and grandchildren including parents without children


SELECT parent.id AS ParentID, 
       IFNULL(child.parent_id,') AS ChildParentID, 
       IFNULL(child.id,') AS ChildID, 
       IFNULL(grandchild.child_id,') AS GrandchildChildID 
FROM parent 
  LEFT JOIN child ON parent.id=child.parent_id 
  LEFT JOIN grandchild ON child.id=grandchild.child_id;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Classroom scheduling

You have n student classes of known size, and m classrooms of known size, where m>=n. What's the best algorithm for assigning as many classes as possible to rooms of adequate size?

It's a version of the combinatorial knapsack problem. It's known to be NP-complete, which means it's possible to verify any correct solution but there is no known algorithm for quickly finding a correct solution. How then to proceed?

Early in 2010 Joe Celko resurrected the problem in a Simple Talk column, and challenged readers to improve on SQL Server solutions he'd published in the third edition of his "SQL for Smarties". Here's his small version of the problem modified for MySQL:


DROP TABLE IF EXISTS Rooms, Classes; 
CREATE TABLE Rooms( 
  room_nbr CHAR(2) NOT NULL PRIMARY KEY, room_size INTEGER NOT NULL 
) ENGINE=MyISAM; 
CREATE TABLE Classes(

Last updated 28 Jan 2011

Back to the top

Subscribe for Us

Feedback

Data-driven joins

Data-driven table relationships are hard to maintain, but sometimes they cannot be avoided. How do we build joins for them? One way is to use a CASE statement in the SELECT list to handle the joining possibilities. In this example, the parent.linktable column determines the name of the table where a particular parent row's data is. The method is fine when the number of child tables is small:


USE test; 
DROP TABLE IF EXISTS parent, child1, child2; 

CREATE TABLE parent ( 
  id INT UNSIGNED PRIMARY KEY,  
  linktable CHAR(64) NOT NULL 
); 
INSERT INTO parent VALUES (1, 'child1'), (2, 'child2'); 

CREATE TABLE child1 ( 
  id INT UNSIGNED PRIMARY KEY,

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Full Outer Join

A FULL OUTER join between tables a and b retrieves:

all rows from a, with matching rows or nulls from b, and
all rows from b, with matching rows or nulls from a

so for these tables:


DROP TABLE IF EXISTS a,b; 
CREATE TABLE a(id int,name char(1)); 
CREATE TABLE b(id int,name char(1)); 
INSERT INTO a VALUES(1,'a'),(2,'b'); 
INSERT INTO b VALUES(2,'b'),(3,'c'); 
SELECT * FROM a; 
+------+------+ 
| id   | name |

Last updated 12 Oct 2014

Back to the top

Subscribe for Us

Feedback

Intersection and difference

MySQL implements UNION, but does not directly implement INTERSECTION or DIFFERENCE.

INTERSECTION is just an INNER JOIN on all columns:


drop table if exists a,b; 
create table a(i int,j int); 
create table b like a; 
insert into a values(1,1),(2,2); 
insert into b values(1,1),(3,3); 
select * from a join b using(i,j); 
+------+------+ 
| i    | j    |

Last updated 30 Jun 2009

Back to the top

Subscribe for Us

Feedback

Many-to-many joins

To model a many:many relationship between two tables a and b, you need a bridging table where each row represents one instance of an association between a row in a and a row in b, as in this example:


drop table if exists users,actions,useractions; 
create table users(userid int primary key, username char(32)); 
insert into users values(1, 'James'),(2, 'Alex'),(3, 'Justin'); 
create table actions(actionid int primary key, action char(32)); 
insert into actions values(1, 'Login'),(2, 'Logout'),(3, 'Delete'),(4, 'Promote'); 
create table useractions(uaid int primary key, userid int, actionid int); 
insert into useractions values(1,1,1),(2,1,2),(3,3,4); 

select u.username, a.action 
from useractions ua 
join users   u using (userid)

Last updated 16 Mar 2010

Back to the top

Subscribe for Us

Feedback

What else did buyers of X buy?

We often want to know how certain column values associate with other column values, for example "What else did buyers of x buy?", or "What projects did Sean, Ian and Gerard all work on?"

Start with buyers of x. The table that summarises this information might be a View that encapsulates joins from customers to orders to orderitems to products, perhaps scoped on a recent date range. Here we ignore all such detail. We focus only on the SQL patterns that solve this kind of problem:


DROP TABLE IF EXISTS userpurchases; 
CREATE TABLE userpurchases( custID INT UNSIGNED, prodID INT UNSIGNED ); 
INSERT INTO userpurchases  
VALUES (1,1),(1,2),(2,4),(3,1),(3,2),(4,2),(4,3),(5,1),(5,2),(5,3); 
SELECT custID, GROUP_CONCAT(prodID ORDER BY prodID) AS PurchaseList 
FROM userpurchases 
GROUP BY custID; 
+--------+--------------+

Last updated 28 Jan 2011

Back to the top

Subscribe for Us

Feedback

Join or subquery?

Usually, a JOIN is faster than an uncorrelated subquery. For example in the sakila test database, customer is a parent of rental (via customer_id) which in turn is a parent of payment (via rental_id). The subquery version of a query for whether a customer has made payments and rentals...


SELECT DISTINCT c.customer_id 
FROM customer c 
WHERE c.customer_id IN ( 
   SELECT r.customer_id  
   FROM rental r 
   JOIN payment p USING (rental_id)  
   WHERE c.customer_id = 599; 
);

is eight times slower than the join version...

  
SELECT DISTINCT c.customer_id 
FROM customer c 
JOIN rental r USING (customer_id) 
JOIN payment p USING (rental_id)  
WHERE c.customer_id = 599;

Running EXPLAIN on the two queries reveals why: the subquery version has to read most customer rows, while the join version proceeds inside out and discovers it needs to read just one customer row.

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Parents without children

Given tables parent(id INT), child(id INT,parent_id INT), how do we find parents with no children? It's the All X for which there is no Y pattern, which can be written as an exclusion join...


SELECT parent.id 
FROM parent  
LEFT JOIN child ON parent.id = child.parent_id  
WHERE child.parent_id IS NULL;

or with a NOT EXISTS subquery, which is logically equivalent to the exclusion join, but usually performs much slower:


SELECT parent.id AS ParentID 
FROM parent 
WHERE NOT EXISTS ( 
  SELECT parent.id 
  FROM parent  
  JOIN child ON parent.ID = child.parent_id 
);

Last updated 28 Jan 2011

Back to the top

Subscribe for Us

Feedback

Parties who have contracts with one another

You have a parties table that holds info on peoples' names etc, and a contracts table where each row has clientID and contractorID value pointing at a parties.partyID value--that is, each contracts row points at two parties rows. You want to list the names of all contractors and their clients.


SELECT clientpartyID,  
       pCli.name AS Client,  
       contractorpartyID,  
       pCon.name AS Contractor 
FROM contracts 
  INNER JOIN parties AS pCli  
    ON contracts.clientpartyID = pCli.partyID 
  INNER JOIN parties AS pCon  
    ON contracts.contractorpartyID = pCon.partyID;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

The unbearable slowness of IN()

You track orders and their items in orders and orderdetails tables, as in the NorthWind database. How many of your orders have been for multiple items? We can use the standard SQL IN() operator to answer the question:


SELECT orderID 
FROM orders 
WHERE orderID IN ( 
  SELECT orderID  
  FROM orderdetails  
  GROUP BY orderID  
  HAVING COUNT(orderID) > 1 
);

Last updated 05 Nov 2015

Back to the top

Subscribe for Us

Feedback

The [Not] Exists query pattern

Given a table employee( employeeID INT, mgr_employeeID INT, salary DECIMAL(10,2)), find the managers who earn less than one or more of their subordinates.

We can write this query directly from the logic of its spec...


SELECT DISTINCT employeeID 
FROM employee AS e 
WHERE EXISTS (  
  SELECT employeeID 
  FROM employee AS m 
  WHERE m.mgr_employeeID = e.employeeID AND e.salary > m.salary 
);

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

What exams did a student not register for?

We have a students table, an exams table, and a registrations table in which each row registers one student for one exam. How do we find the exams for which a particular student is not registered?


DROP TABLE IF EXISTS students, exams, registrations; 
CREATE TABLE students ( 
  sid int(10) unsigned PRIMARY KEY auto_increment, 
  firstname varchar(45) NOT NULL default '', 
  lastname varchar(45) NOT NULL default ''
); 
INSERT INTO students VALUES  
(1, 'Jack', 'Malone'),(2, 'Hiro', 'Nakamura'),(3, 'Bree', 'Van de Kamp'), 
(4, 'Susan', 'Mayer'),(5, 'Matt', 'Parkman'),(6, 'Claire', 'Bennet'); 

CREATE TABLE exams (

Last updated 28 Jan 2011

Back to the top

Subscribe for Us

Feedback

List NULLs at end of query output

If ordering by col...


... ORDER BY IF(col IS NULL, 0, 1 ), col ...

Last updated 30 Dec 2009

Back to the top

Subscribe for Us

Feedback

Parents with and without children

You have parties and contracts tables. Every contracts row has a contractorpartyID value which references a row in parties, and a clientpartyID value which also references a row in parties. How to list all parties and their contracts, showing blanks as empty strings rather than NULLs?


SELECT parties.partyID,  
       IFNULL(contractorpartyID,'') AS contractor,  
       IFNULL(clientpartyID,'') AS client 
FROM parties  
LEFT JOIN contractor_client ON partyID=contractorpartyID 
ORDER BY partyID; 
+---------+------------+--------+ 
| partyID | contractor | client | 
+---------+------------+--------+ 
|       1 |            |        | 
|       2 | 2          | 1      | 
|       3 |            |        | 
+---------+------------+--------+

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Emulate Row_Number()

ISO SQL defines a ROW_NUMBER() OVER function with an optional PARTITION clause for generating a derived row number column in a resultset. Several RDBMSs—including DB2, Oracle and SQL Server—implement it. Here is the simplest possible example. Given a table with two columns i and j, generate a resultset that has a derived sequential row_number column taking the values 1,2,3,... for a defined ordering of j which resets to 1 when the value of i changes:


DROP TABLE IF EXISTS test; 
CREATE TABLE test(i int,j int); 
INSERT INTO test  
VALUES (3,31),(1,11),(4,14),(1,13),(2,21),(1,12),(2,22),(3,32),(2,23),(3,33);

The result must look like this:

Last updated 29 Jan 2013

Back to the top

Subscribe for Us

Feedback

Next row

You have a table of names, you have retrieved a row with name $name, and you want the row for the next name in name order. MySQL LIMIT syntax makes this very easy:


SELECT * 
FROM tbl 
WHERE name > $name 
ORDER BY name 
LIMIT 1

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Order by leading digits

To have column values 1abc,10abc,8abc appear in the expected order 1abc,8abc,10abc, take advantage of a trick built into MySQL string parsing ...


SELECT '3xyz'+0; 
+----------+ 
| '3xyz'+0 | 
+----------+ 
|        3 | 
+----------+

to write ...


SELECT ... 
... 
ORDER BY colname+0, colname;

Last updated 31 Mar 2012

Back to the top

Subscribe for Us

Feedback

Order by month name

The MySQL FIELD(str,str1,str2,...,strN) function returns 1 if str=str1, 2 if str=str2, etc., so ...


SELECT .  
ORDER BY FIELD(month,'JAN','FEB','MAR',...,'NOV','DEC') .

will order query output from a legacy table in month-number order.

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Order by numerics then alphas

Given a column type with values 1,2,3,a,b,c, how to get the order 3,2,1,c,b,a?


ORDER BY type RLIKE '^[0-9]+$' DESC, `type` DESC

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Pagination

Suppose you have a phone book of names, addresses, etc. You are displaying 20 rows per page, you're on page 100, and you want to display page 99. How do you do this knowing only what page you are on?

Assuming...

1-based page numbers
you are on page P
each page shows N rows

then the general formula for translating a 1-based page number into a first LIMIT argument is ...


 MAX(0,P-1) * N

which for the 99th 20-row page evaluates to 1960, and the second argument to LIMIT is just N, so to see page 99, write...

  
 SELECT ... LIMIT (1960, N);

The trouble with this is scaling. MySQL doesn't optimise LIMIT at all well. SELECT ... LIMIT 1000000,20 will unfortunately retrieve not just the twenty rows starting at the millionth row; it will retrieve a million and twenty rows before it shows you the 20 you asked for! The bigger the result, the longer LIMIT takes.

What's the alternative? Build pagination logic into the WHERE clause, and ensure that there is a covering index for the paginating column. On a table of 100,000 indexed random integers, SELECT ... WHERE ... for the last 20 integers in the table is twice as fast as the comparable LIMIT query. With a million integers, it's more than 500 times faster!

If your interface calls for showing only 20 rows per page on a given order, retrieve the twenty rows, plus the row just before the set if it exists, plus the next row after those twenty if it exists. When the user clicks the Previous Page button, adjust the WHERE clause to specify rows where the key value is <= the row just before the current set and ORDER BY the current index DESC LIMIT 20; likewise when the user clicks the Next Page button, have the WHERE clause specify key values >= that of the row just after the set, and ORDER BY the current index ASC LIMIT 20.

Last updated 22 Oct 2014

Back to the top

Subscribe for Us

Feedback

Suppress repeating ordering values

You have tables tracking authors and their books, for example:


CREATE TABLE author ( 
  id int(4) NOT NULL auto_increment PRIMARY KEY, 
  name text NOT NULL 
); 
INSERT INTO author (id, name)  
VALUES (1,'Brad Phillips'),(2,'Don Charles'),(3,'Kur Silver'); 
CREATE TABLE book ( 
  id int(4) NOT NULL auto_increment PRIMARY KEY, 
  name text NOT NULL 
); 
INSERT INTO book (id, name)

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

All possible recipes with given ingredients

We have tables for recipes (r), ingredients required for recipes (ri), and ingredients now available in the pantry (p). In table p there may be many rows for a given recipe, each specifying one ingredient.


drop table if exists r,p,ri; 
create table r(id int); 
insert into r values(1),(2),(3); 
create table p(id int); 
insert into p values(1),(2),(3); 
create table ri(rid int,pid int); 
insert into ri values (1,1),(1,2),(2,1),(2,4),(3,5),(3,6),(3,7);  
select id as recipes from r; 
+---------+ 
| recipes | 
+---------+

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

All X for which all Y are Z (relational division)

You have an election database with tables listing political parties, election districts, and candidates running for parties in those districts. You want to know which parties have candidates running in all districts. Under Aggregates we show a GROUP BY solution (here).

If there are reasons not to aggregate, relational division can solve the problem. The basic idea in relational division is that, aside from aggregation, SQL has no direct way to express "all Xs for which all Y are Z", but does have a NOT EXISTS operator, so we can express "all Xs for which all Y are Z" in SQL as a double negative: "all Xs for which no Y is not Z". Once you think of formulating the question this way, the query almost writes itself:


SELECT DISTINCT party FROM parties 
WHERE NOT EXISTS ( 
  SELECT * FROM districts  
  WHERE NOT EXISTS ( 
    SELECT * FROM candidates 
    WHERE candidates.party=parties.party AND candidates.district=districts.district 
  ) 
);

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Who makes all the parts for a given assembly?

One way to arrive at the answer is by asking: What are the assembly-supplier pairs such that no part of the assembly is not made by the supplier? That's relational division again, formulated for two tables by Stephen Todd. Given assemblyparts(assembly,part) and partsuppliers(part,supplier) tables, here is a query that Joe Celko credits to Pierre Mullin.


SELECT DISTINCT  
  AP1.assembly,  
  SP1.supplier 
FROM AssemblyParts AS AP1, PartSuppliers AS SP1 
WHERE NOT EXISTS ( 
  SELECT * 
  FROM AssemblyParts AS AP2 
  WHERE AP2.assembly = AP1.assembly 
  AND NOT EXISTS ( 
    SELECT SP2.part 
    FROM PartSuppliers AS SP2 
    WHERE SP2.part = AP2.part AND SP2.supplier = SP1.supplier 
  ) 
);

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Make a table of sequential ints

For example, a table of 1,000 ints starting from 0 with interval=1...


drop table if exists ints; 
create table ints(i tinyint); 
insert into ints values(0),(1),(2),(3),(4),(5),(6),(7),(8),(9); 

drop table if exists temp; 
create table temp 
select 100*c.i+10*b.i+a.i as iseq  
from ints a,ints b, ints c order by iseq;

Last updated 21 Oct 2014

Back to the top

Subscribe for Us

Feedback

Find adjacent unbooked theatre seats

A theatre booking service is often asked to book adjecent seats:


drop table if exists seats; 
create table seats(row char(1),seat int,booked tinyint); 
insert into seats values 
('i',1,0),('i',2,0),('i',3,0),('i',4,0),('i',5,0),('i',6,0),('i',7,0),('i',8,0), 
('i',9,0),('i',10,0),('i',11,0),('i',12,0),('i',13,0),('i',14,0),('i',15,0), 
('j',1,1),('j',2,0),('j',3,1),('j',4,0),('j',5,0),('j',6,0),('j',7,1),('j',8,0), 
('j',9,0),('j',10,0),('j',11,0),('j',12,0),('j',13,1),('j',14,0),('j',15,0);

The simplest method is a self-join:

Last updated 26 Oct 2010

Back to the top

Subscribe for Us

Feedback

Find average row-to-row time interval

You have a table of sequential times, and you're asked to report the average time interval between them in seconds.

Assuming table t with timestamp column ts and no ts duplicate values, the mean time difference between consecutive rows is the difference between the smallest (earliest) and largest (latest) timestamp divided by the rowcount - 1:


Select TimeStampDiff(SECOND, Min(ts), Max(ts) ) / ( Count(DISTINCT(ts)) -1 )  
FROM t

Not sure of the formula?

1. The mean distance between consecutive rows is the sum of distances between consecutive rows, divided by the number of consecutive rows.

2. The sum of differences between consecutive rows is just the distance between the first row and last sorted row (assuming they are sorted by timestamp).

3. The number of consecutive rows is the total number of rows - 1.

Last updated 20 Jun 2014

Back to the top

Subscribe for Us

Feedback

Find blocks of unused numbers

In a table of sequential IDs with no missing values, some are used and some are not. Find the blocks of unused IDs, if any:


DROP TABLE IF EXISTS tbl; 
CREATE TABLE tbl(id INT,used BOOL); 
INSERT INTO tbl VALUES (1,1),(2,1),(3,0),(4,1),(5,0),(6,1),(7,1),(8,1), 
                       (9,0),(10,0),(11,1),(12,1),(13,0),(14,0),(15,0); 
SELECT * FROM tbl; 
+------+------+ 
| id   | used | 
+------+------+ 
|    1 |    1 | 
|    2 |    1 | 
|    3 |    0 |

Last updated 25 Sep 2009

Back to the top

Subscribe for Us

Feedback

Find missing numbers in a sequence

You have a table tbl(id int) with values (1,2,4,18,19,20,21), and you wish to find the first missing number in its sequence of id values:


SELECT t1.id+1 AS Missing 
FROM tbl AS t1 
LEFT JOIN tbl AS t2 ON t1.id+1 = t2.id 
WHERE t2.id IS NULL 
ORDER BY id LIMIT 1; 
+---------+ 
| Missing | 
+---------+ 
|       3 | 
+---------+

Last updated 05 Mar 2015

Back to the top

Subscribe for Us

Feedback

Find missing values in a range

You have a table named tbl with an integer primary key named id, and you need to know what key values between 0 and 999999 are missing in the table.

The simplest solution is an exclusion join from a virtual table of sequential numbers to the target table:


create or replace view v as  
  select 0 i union select 1 union select 2 union select 3 union select 4  
  union select 5 union select 6 union select 7 union select 8 union select 9;  

select x.i 
from ( 
  select a.i*100000 + b.i*10000 + c.i*1000 + d.i*100 + e.i*10 + f.i as i  
  from v a  
  join v b  
  join v c  
  join v d  
  join v e  
  join v f 
) x 
left join tbl on x.i=tbl.id 
where tbl.id is null 
order by x.i;

Last updated 11 Jul 2012

Back to the top

Subscribe for Us

Feedback

Find previous and next values in a sequence

Given a table t(ID int, ...), how would we display each ID and its next highest value?

A simple method ...


SELECT id, (SELECT MIN(id) from t as x WHERE x.id > t.id) AS Next 
FROM t 
ORDER BY id;

...but early versions of MySQL did not optimise correlated qubqueries at all well.

Last updated 16 Dec 2015

Back to the top

Subscribe for Us

Feedback

Find row with next value of specified column

Sometimes we need next values of a column on a given row ordering. Oracle has a LEAD(...) OVER(...) construct to simplify this query. MySQL does not.

A few efficient ways to do this are described under "Find previous and next values in a sequence". Here we look at more ambitious solutions.

The logic is:

1. Form a resultset consisting of all relevant rows, joined with all relevant rows that have greater values in the ordering columns. For example, if the table has these rows:


+------+ 
|    2 |

Last updated 17 Apr 2015

Back to the top

Subscribe for Us

Feedback

Find sequence starts and ends

A traditional way to find the first and last values of column value sequences in a table like this ...


drop table if exists t; 
create table t(id int); 
insert into t values(1),(2),(3),(4),(6),(7),(8);

... uses an exclusion join on the previous sequential value to find the first value of each sequence, and the minimum next value from a left join and an exclusion join on the previous sequential value to find the end of each sequence:


SELECT

Last updated 14 Dec 2014

Back to the top

Subscribe for Us

Feedback

Find shortest & longest per-user event time intervals

You have a table of users and event times, and you need to find the shortest and longest per-user event time intervals:


drop table if exists t; 
create table t( t timestamp, user smallint); 
insert into t values 
('2014-11-28 18:30:02',  1),('2014-11-28 18:30:05',  1),('2014-11-28 18:30:08',  1), 
('2014-11-28 18:30:11',  1),('2014-11-28 18:30:15',  1),('2014-11-28 18:30:18',  1), 
('2014-11-28 18:30:21',  1),('2014-11-28 18:30:23',  1),('2014-11-28 18:30:26',  1), 
('2014-11-28 18:30:29',  2),('2014-11-28 18:30:32',  2),('2014-11-28 18:30:33',  2), 
('2014-11-28 18:30:37',  2),('2014-11-28 18:30:40',  2),('2014-11-28 18:30:42',  2), 
('2014-11-28 18:30:44',  2),('2014-11-28 18:31:01',  2),('2014-11-28 18:31:04',  2), 
('2014-11-28 18:31:07',  2),('2014-11-28 18:31:10',  2);

Last updated 03 Dec 2014

Back to the top

Subscribe for Us

Feedback

Find specific sequences

You have a table which tracks hits on named web pages...


CREATE TABLE hits ( 
  id INT NOT NULL DEFAULT 0, 
  page CHAR(128) DEFAULT '', 
  time TIMESTAMP NOT NULL DEFAULT 0, 
  PRIMARY KEY(id, time) 
)

where id is unique to a session. Here is a bit of sample data:

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Find the next value after a sequence

Given a sequence of values on a given ordering, what is the next value? It's a common requirement (eg in DNA sequencing). A MySQL Forum user posted this example:


drop table if exists data; 
create table data(id smallint unsigned primary key auto_increment, val smallint); 
insert into data (val) values 
(8),(21),(28),(29),(31),(32),(27),(20),(31),(1),(18),(35), 
(18),(30),(22),(9),(2),(8),(33),(8),(19),(31),(6),(31),(14),(5), 
(26),(29),(34),(34),(19),(27),(29),(3),(21),(18),(31),(5),(18), 
(34),(4),(15),(12),(20),(28),(31),(13),(22),(19),(30),(0),(2), 
(30),(28),(2),(10),(27),(9),(23),(28),(29),(16),(16),(31),(35),(18), 
(2),(15),(1),(30),(15),(11),(17),(26),(35),(1),(22),(19),(23),(1), 
(18),(35),(28),(13),(9),(14);

Last updated 31 Mar 2012

Back to the top

Subscribe for Us

Feedback

Gaps in a time series

Advanced time series analysis generally requires custom software, but straightforward SQL queries can answer simple time series questions. You have a jobtimes table with columns ID, job, machine, start_time, and stop_time. You wish to know which machines have had gaps between activity periods. It's a version of "Find available booking periods":


drop table jobtimes; 
create table jobtimes(id int, machine smallint, start_time timestamp, stop_time timestamp); 
insert into jobtimes values(1,1,'2011-7-1 08:00:00', '2011-7-1 10:00:00'); 
insert into jobtimes values(2,1,'2011-7-1 11:00:00', '2011-7-1 14:00:00'); 
insert into jobtimes values(3,2,'2011-7-1 08:00:00', '2011-7-1 09:00:00'); 
insert into jobtimes values(4,2,'2011-7-1 09:00:00', '2011-7-1 10:00:00'); 
insert into jobtimes values(5,3,'2011-7-1 08:00:00', '2011-7-1 08:30:00'); 
insert into jobtimes values(6,3,'2011-7-1 10:00:00', '2011-7-1 12:00:00'); 
select * from jobtimes; 
+------+---------+---------------------+---------------------+ 
| id   | machine | start_time          | stop_time           |

Last updated 15 Jul 2011

Back to the top

Subscribe for Us

Feedback

Make values of a column sequential

You have a table tbl with a column whose values can be freely overwritten and which needs to be filled with perfectly sequential values starting with 1:


SET @i=0; 
UPDATE tbl SET keycol=(@i:=@i+1);

But more often, what you need is a mechanism for guaranteeing an unbroken sequence of integer key values whenever a row is inserted.

Here is the bad news: auto_increment does not provide this logic, for example it does not prevent inserton failures, editing of the column, or subsequent deletion.

The good news is that the logic for guaranteeing an unbroken sequence is straightforward:

(i) use InnoDB

(ii) create a table with the desired sequence in one column and further column(s) to track use (to use such a table for sequential columns in multiple other tables, provide a tracking column for each such table)

(iii) write Triggers to fetch the next available sequential key from that table, mark it used, and assign it on Insert

(iv) put all such logic inside transaction blocks

(v) prohibit deletion in the table.

Last updated 05 Mar 2013

Back to the top

Subscribe for Us

Feedback

Track stepwise project completion

A master table has one row for each project, and the number of sequential steps required to complete each project. A detail table has one row per project per completed step:


DROP TABLE IF EXISTS t1 ; 
CREATE TABLE t1 ( 
  id INT, projectname CHAR(2), projectsteps INT 
); 
INSERT INTO t1 VALUES  
(1, 'xx', 3),  
(2, 'yy', 3), 
(3, 'zz', 5); 

DROP TABLE IF EXISTS t2; 
CREATE TABLE t2 (

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Winning streaks

Given a table of IDs and won-lost results, how do we find the longest winning streak?


drop table if exists results; 
create table results(id int,result char(1)); 
insert into results values 
(1,'w'),(2,'l'),(3,'l'),(4,'w'),(5,'w'),(6,'w'),(7,'l'),(8,'w'),(9,'w'); 
select * from results; 
+------+--------+ 
| id   | result | 
+------+--------+ 
|    1 | w      | 
|    2 | l      | 
|    3 | l      |

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Great circle distance

Find the distance in kilometres between two points on the surface of the earth. This is just the sort of problem stored functions were made for. For a first order approximation, ignore deviations of the earth's surface from the perfectly spherical. Then the distance in radians is given by a number of trigonometric formulas. ACOS and COS behave reasonably:


             COS(lat1-lat2)*(1+COS(lon1-lon2)) - COS(lat1+lat2)*(1-COS(lon1-lon2)) 
rads = ACOS( --------------------------------------------------------------------- ) 
                                              2

We need to convert degrees latitude and longitude to radians, and we need to know the length in km of one radian on the earth's surface, which is 6378.388. The function:


set log_bin_trust_function_creators=TRUE;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Top ten

We often want to know the top 1, 2, 10 or whatever values from a query. This is dead simple in MySQL. However many JOINs and WHEREs the query has, simply ORDER BY the column(s) whose highest values are sought, and LIMIT the resultset:


  SELECT (somecolumn), (othercolumns) ... 
  FROM (some tables) ... 
  ORDER BY somecolumn DESC 
  LIMIT 10;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

A cursor if necessary, but not necessarily a cursor

You have photos (id INT, photo BLOB, tally INT) and votes(id INT, userID INT, photoID INT) tables. You wish to update photos.tally values from counts per photo in the votes table. You can use a cursor to walk the photos table, updating the tally as you go:


DROP TABLE IF EXISTS photos; 
CREATE TABLE photos (id INT, photo BLOB, tally INT);  
INSERT INTO photos VALUES(1,'',0),(2,'',0); 
DROP TABLE IF EXISTS VOTES; 
CREATE TABLE VOTES( userID INT, photoID INT); 
INSERT INTO votes VALUES (1,1),(2,1),(2,2); 

DROP PROCEDURE IF EXISTS updatetallies; 
DELIMITER // 
CREATE PROCEDURE updatetallies() 
BEGIN

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Emulate sp_exec

Sometimes it is desirable to call multiple stored procedures in one command. In SQL Server this can be done with sp_exec. In MySQL we can easily write such an sproc that calls as many sprocs as we please, for example...


USE sys; 
DROP PROCEDURE IF EXISTS sp_exec; 
DELIMITER | 
CREATE PROCEDURE sp_exec( p1 CHAR(64), p2 CHAR(64) ) 
BEGIN 
  -- permit doublequotes to delimit data 
  SET @sqlmode=(SELECT @@sql_mode); 
  SET @@sql_mode=''; 
  SET @sql = CONCAT( "CALL ", p1 ); 
  PREPARE stmt FROM @sql; 
  EXECUTE stmt; 
  DROP PREPARE stmt; 
  SET @sql = CONCAT( "CALL ", p2 ); 
  PREPARE stmt FROM @sql; 
  EXECUTE stmt; 
  DROP PREPARE stmt; 
  SET @@sql_mode=@sqlmode; 
END; 
| 
DELIMITER ;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Variable-length argument for query IN() clause

To have an sproc accept a variable-length parameter list for an IN(...) clause in a query, code the sproc to PREPARE the query statement:


DROP PROCEDURE IF EXISTS passInParam; 
DELIMITER | 
CREATE PROCEDURE passInParam( IN qry VARCHAR(100), IN param VARCHAR(1000) ) 
BEGIN 
  SET @qry = CONCAT( qry, param, ')' ); 
  PREPARE stmt FROM @qry; 
  EXECUTE stmt; 
  DROP PREPARE stmt; 
END; 
| 
DELIMITER ;

For this example, the query string should be of the form:


SELECT ... FROM ... WHERE ... IN (

but so long as it has those elements, it can be as complex as you like. When you call the sproc:
1. Quote each argument with a pair of single quotes,
2. Separate these quoted arguments with commas,
3. Surround the whole param string with another set of single quotes:


CALL passInParam( 'SELECT * FROM tbl WHERE colval IN (', ('''abc'',''def'',''ghi''' ));

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Count delimited substrings

Here is a function to count substrings delimited by a constant delimiting string:


DROP FUNCTION IF EXISTS strcount; 
SET GLOBAL log_bin_trust_function_creators=1; 
DELIMITER | 
CREATE FUNCTION strCount( pDelim VARCHAR(32), pStr TEXT) RETURNS int(11) 
BEGIN 
  DECLARE n INT DEFAULT 0; 
  DECLARE pos INT DEFAULT 1; 
  DECLARE strRemain TEXT; 
  SET strRemain = pStr; 
  SET pos = LOCATE( pDelim, strRemain ); 
  WHILE pos != 0 DO

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Count substrings

To count instances of a search string in a target string ...

in the target string, replace the search string with a single character,
subtract the length of the modified target string from the length of the original target string,
divide that by the length of the search string:


SET @str  = "The quick brown fox jumped over the lazy dog";  
SET @find = "the"; 
SELECT ROUND(((LENGTH(@str) - LENGTH(REPLACE(LCASE(@str), @find, '')))/LENGTH(@find)),0)  
AS COUNT;  
+-------+ 
| COUNT | 
+-------+ 
|     2 | 
+-------+

Note that REPLACE() does a case-sensitive search; to get a case-insensitive result you must coerce target and search strings to one case.

To remove decimals from the result:


SELECT CAST((LENGTH(@str) - LENGTH(REPLACE(LCASE(@str)), @find, '')))/LENGTH(@find) AS SIGNED) AS COUNT;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Levenshtein distance

The Levenshtein distance between two strings is the minimum number of operations needed to transform one string into the other, where an operation may be insertion, deletion or substitution of one character. Jason Rust published this MySQL algorithm for it at http://www.codejanitor.com/wp/.


CREATE FUNCTION levenshtein( s1 VARCHAR(255), s2 VARCHAR(255) ) 
  RETURNS INT 
  DETERMINISTIC 
  BEGIN 
    DECLARE s1_len, s2_len, i, j, c, c_temp, cost INT; 
    DECLARE s1_char CHAR; 
    -- max strlen=255 
    DECLARE cv0, cv1 VARBINARY(256); 
    SET s1_len = CHAR_LENGTH(s1), s2_len = CHAR_LENGTH(s2), cv1 = 0x00, j = 1, i = 1, c = 0; 
    IF s1 = s2 THEN 
      RETURN 0;

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Proper case

The basic idea is...

lower-case the string
upper-case the first character if it is a-z, and any other a-z character that follows a punctuation character

Here is the function. To make it work with strings long than 128 characters, change its input and return declarations accordingly:


DROP FUNCTION IF EXISTS proper; 
SET GLOBAL  log_bin_trust_function_creators=TRUE; 
DELIMITER | 
CREATE FUNCTION proper( str VARCHAR(128) ) 
RETURNS VARCHAR(128) 
BEGIN 
  DECLARE c CHAR(1);

Last updated 20 Jun 2012

Back to the top

Subscribe for Us

Feedback

Retrieve octets from IP addresses

If ip is for the form N.N.N.N where N may be 1, 2 or 3 digits, how to group and count by just the first three octets, ie the ip class?


SELECT  
  LEFT(ip, CHAR_LENGTH(ip) - LOCATE('.', REVERSE(ip))) as ipclass, 
  COUNT(*) 
FROM tbl 
GROUP BY ipclass;

Hamilton Turner notes we can find the first octet with LEFT(ip,LOCATE('.',ip)-1).

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Return digits or alphas from a string

Return only the digits from a string:


SET GLOBAL log_bin_trust_function_creators=1; 
DROP FUNCTION IF EXISTS digits; 
DELIMITER | 
CREATE FUNCTION digits( str CHAR(32) ) RETURNS CHAR(32) 
BEGIN 
  DECLARE i, len SMALLINT DEFAULT 1; 
  DECLARE ret CHAR(32) DEFAULT ''; 
  DECLARE c CHAR(1); 
  SET len = CHAR_LENGTH( str ); 
  REPEAT 
    BEGIN

Last updated 22 May 2009

Back to the top

Subscribe for Us

Feedback

Split a string


CREATE FUNCTION SPLIT_STR( s VARCHAR(255), delim VARCHAR(16),  pos INT ) 
RETURNS VARCHAR(255) 
RETURN REPLACE( SUBSTRING( SUBSTRING_INDEX(s, delim, pos), 
                           LENGTH(SUBSTRING_INDEX(s, delim, pos -1)) + 1), 
                delim,  
                ''
              );

For longer strings, change the datatype.

Last updated 27 Jan 2012

Back to the top

Subscribe for Us

Feedback

Strip HTML tags

Ported from a T-SQL function by Robert Davis:


SET GLOBAL log_bin_trust_function_creators=1; 
DROP FUNCTION IF EXISTS fnStripTags; 
DELIMITER | 
CREATE FUNCTION fnStripTags( Dirty varchar(4000) ) 
RETURNS varchar(4000) 
DETERMINISTIC  
BEGIN 
  DECLARE iStart, iEnd, iLength int; 
  WHILE Locate( '<', Dirty ) > 0 And Locate( '>', Dirty, Locate( '<', Dirty )) > 0 DO 
    BEGIN 
      SET iStart = Locate( '<', Dirty ), iEnd = Locate( '>', Dirty, Locate('<', Dirty ));

Last updated 16 Jul 2012

Back to the top

Subscribe Us

Feedback

↧

Leonids - Android Particle Animation Library

November 14, 2016, 8:34 pm

≫ Next: Top 8 Features of Snapchat Glass : You should know

≪ Previous: Important Common List of MySQL Queries

Leonids is a particle system library that works with the standard Android UI.

The library is extremely lightweight, LeonidsLib.jar is just 81Kb.

You can download Leonids Demo from Google Play to check out what can be done with it.

Setup

Leonids is available in jcenter as well as a jar file to fit both Android Studio and Eclipse.

Android Studio / gradle

Add the following dependency to the build.gradle of your project₹₹₹₹

dependencies {
    compile 'com.plattysoft.leonids:LeonidsLib:1.3.1'
}

Note: If you get an error, you may need to update the jcenter repository to:

repositories {
    jcenter{
        url "http://jcenter.bintray.com/"
    }
}

Eclipse / jar file

Just put LeonidsLib.jar into the libs folder of your app.

Why this library?

Particle systems are often used in games for a wide range of purposes: Explosions, fire, smoke, etc. This effects can also be used on normal apps to add an element of "juiciness" or Playful Design.

Precisely because its main use is games, all engines have support for particle systems, but there is no such thing for standard Android UI.

This means that if you are building an Android app and you want a particle system, you have to include a graphics engine and use OpenGL -which is quite an overkill- or you have to implement it yourself.

Leonids is made to fill this gap, bringing particle sytems to developers that use the standard Android UI.

Basic usage

Creating and firing a one-shot particle system is very easy, just 3 lines of code.

newParticleSystem(this, numParticles, drawableResId, timeToLive)
.setSpeedRange(0.2f, 0.5f)
.oneShot(anchorView, numParticles);

Note that the ParticleSystem checks the position of the anchor view when oneShot (or emit) is called, so it requires the views to be measured. This means that ParticleSystem won't work properly if you call oneShot or emit during onCreate.

When you create the particle system, you tell how many particles will it use as a maximum, the resourceId of the drawable you want to use for the particles and for how long the particles will live.

Then you configure the particle system. In this case we specify that the particles will have a speed between 0.2 and 0.5 pixels per milisecond (support for dips will be included in the future). Since we did not provide an angle range, it will be considered as "any angle".

Finally, we call oneShot, passing the view from which the particles will be launched and saying how many particles we want to be shot.

Emitters

You can configure emitters, which have a constant ratio of particles being emited per second. This is the code for the Confeti example:

newParticleSystem(this, 80, R.drawable.confeti2, 10000)
.setSpeedModuleAndAngleRange(0f, 0.3f, 180, 180)
.setRotationSpeed(144)
.setAcceleration(0.00005f, 90)
.emit(findViewById(R.id.emiter_top_right), 8);

newParticleSystem(this, 80, R.drawable.confeti3, 10000)
.setSpeedModuleAndAngleRange(0f, 0.3f, 0, 0)
.setRotationSpeed(144)
.setAcceleration(0.00005f, 90)
.emit(findViewById(R.id.emiter_top_left), 8);

It uses an initializer for the Speed as module and angle ranges, a fixed speed rotaion and extenal acceleration.

Available Methods

List of the methods available on the class ParticleSystem.

Constructors

All constructors use the activity, the maximum number of particles and the time to live. The difference is in how the image for the particles is specified.

Supported drawables are: BitmapDrawable and AnimationDrawable.

ParticleSystem(Activity a, int maxParticles, int drawableRedId, long timeToLive)
ParticleSystem(Activity a, int maxParticles, Drawable drawable, long timeToLive)
ParticleSystem(Activity a, int maxParticles, Bitmap bitmap, long timeToLive)
ParticleSystem(Activity a, int maxParticles, AnimationDrawable animation, long timeToLive)

There are also constructors that recieve a view id to use as the parent so you can put the particle system on the background (or between any two views)

ParticleSystem(Activity a, int maxParticles, int drawableRedId, long timeToLive, int parentViewId)
ParticleSystem(Activity a, int maxParticles, Drawable drawable, long timeToLive, int parentViewId)
ParticleSystem(Activity a, int maxParticles, Bitmap bitmap, long timeToLive, int parentViewId)
ParticleSystem(Activity a, int maxParticles, AnimationDrawable animation, long timeToLive, int parentViewId)

Configuration

Available methods on the Particle system for configuration are:

setSpeedRange(float speedMin, float speedMax): Uses 0-360 as the angle range
setSpeedModuleAndAngleRange(float speedMin, float speedMax, int minAngle, int maxAngle)
setSpeedByComponentsRange(float speedMinX, float speedMaxX, float speedMinY, float speedMaxY)
setInitialRotationRange (int minAngle, int maxAngle)
setScaleRange(float minScale, float maxScale)
setRotationSpeed(float rotationSpeed)
setRotationSpeedRange(float minRotationSpeed, float maxRotationSpeed)
setAcceleration(float acceleration, float angle)
setFadeOut(long milisecondsBeforeEnd, Interpolator interpolator): Utility method for a simple fade out effect using an interpolator
setFadeOut(long duration):Utility method for a simple fade out

You can start the particle system "in the future" if you want to have the particles already created and moving using

setStartTime(int time)

For more complex modifiers, you can use the method addModifier(ParticleModifier modifier). Available modifiers are:

AlphaModifier (int initialValue, int finalValue, long startMilis, long endMilis)
AlphaModifier (int initialValue, int finalValue, long startMilis, long endMilis, Interpolator interpolator)
ScaleModifier (float initialValue, float finalValue, long startMilis, long endMilis)
ScaleModifier (float initialValue, float finalValue, long startMilis, long endMilis, Interpolator interpolator)

One shot

Make one shot using from the anchor view using the number of particles specified, an interpolator is optional

oneShot(View anchor, int numParticles)
oneShot(View anchor, int numParticles, Interpolator interpolator)

Emitters

Emits the number of particles per second from the emitter. If emittingTime is set, the emitter stops after that time, otherwise it is continuous.

Basic emitters

emit (View emitter, int particlesPerSecond)
emit (View emitter, int particlesPerSecond, int emittingTime)

Emit based on (x,y) coordinates

emit (int emitterX, int emitterY, int particlesPerSecond)
emit (int emitterX, int emitterY, int particlesPerSecond, int emitingTime)

Emit with Gravity

emitWithGravity (View emiter, int gravity, int particlesPerSecond)
emitWithGravity (View emiter, int gravity, int particlesPerSecond, int emitingTime)

Update, stop, and cancel

updateEmitPoint (int emitterX, int emitterY) Updates dynamically the point of emission.
stopEmitting () Stops the emission of new particles, but the active ones are updated.
cancel () Stops the emission of new particles and cancles the active ones.

Other details

Leonids requires minSDK 11 because it uses ValueAnimators. It should be very easy, however to use nineoldandroids and make it work on Gingerbread.

The library is Free Software, you can use it, extended with no requirement to open source your changes. You can also make paid apps using it.

Each Particle System only uses one image for the particles. If you want different particles to be emitted, you need to create a Particle System for each one of them.

↧

Top 8 Features of Snapchat Glass : You should know

November 14, 2016, 10:20 pm

≫ Next: PhotoScan - A new Google Photo feature

≪ Previous: Leonids - Android Particle Animation Library

Snapchat has renamed itself to Snap inc. and enlarged its service to client hardware. the first product launched by the company is that the latest Snap Spectacles that could be a pair of glasses with in-built wireless enabled camera and reminds Google Glass. However, the popular instant electronic messaging company claims that their wearable is way higher in terms of functionality and staying power.

Spectacles smart glasses can permit users to require snaps or shoot videos of the primary person and send them on to the Snapchat app. Though, it's necessary for our users to understand that Google Glasses weren't appreciated an excessive amount of by the lots as a result of people had concern that anyone carrying such overpriced glasses will record their video secretly.

Snapchat believes that it won’t happen with their Spectacles and hopes that they'll for certain become a huge success in close to future. There are many reasons behind this confidence, Let’s see what we all know regarding Snapchat Spectacles until currently and it'd assist you to come to a decision whether or not to travel for it or not.

1. Decent value

The Spectacles can value you around $129.99 that is even lesser than the try of designer shades with none camera inbuilt. it's also terribly cheaper than Google Glass that was priced $1,500 at the time of launch. Snapchat Spectacles provide similar fun and fits into everyone’s budget, which may undoubtedly create a distinction. they'll be shipped in 3 color variants Black, Coral and Teal.

2. Focus on Single Aim

Google Glasses were launched with several options and were additional sort of a wearable computer than a mobile camera. On the opposite hand Snapchat Spectacles have the one aim to capture pictures or record video. Users will record up to ten seconds of video by sound the button mounted on the frame and every tap can records a replacement clip.

3. You'll be able to Shoot while not a Phone

The Spectacles have the flexibility to attach on to the Snapchat via Bluetooth or Wi-Fi and send the footage during a circular video format instead of rectangular or sq. in order that they'll play full screen on any device. The camera glass uses 115-degree lens to suit additional into the frame that is wider than the iPhone that provides 74 degree of viewing angle.

4. Strong Battery Backup

Spectacles ar meant to be the right answer for standalone photography or videography and a strong battery is crucial to try to to therefore. The battery comes with the glasses is capable of taking snaps (that get keep on the glass) for the day when one charge.

5. Share Videos through Snapchat App Wirelessly

As we already mentioned, you'll be able to transfer pictures or video clips via Bluetooth or Wi-Fi to your Snapchat app that is put in on your android or iPhone. you'll be able to add those pictures in your recollections and share them within the Snapchat stories. you'll be able to play videos taken by the Spectacles despite of this fact that they were captured in circular video format as a result of Snapchat can crop them automatically once you rotate the device.

6. No Secret Recording

We keep in mind that people carrying Google Glasses were late prohibited to publicly places because of the very fact that they were able to capture video or pictures of others while not their information. Though, Google created tries to form it clear that the lens of the wearable glitters once the video gets recorded, however folks weren't happy and wished one thing higher. there's no such problem in Spectacles and also the little ring round the camera lights up to let others grasp that you just ar recording.

7. Spectacles Charge just in case

Yes, you scan it right, your Snapchat Spectacles wirelessly charges in their own case. That means, you don’t got to carry wires or charges beside you whereas traveling. Your case can charge the automatically. Though, we are still waiting to check however it can be done.

8. Snapchat Spectacles availability

There is no specific time, that is mentioned by the company as once it's planning to be available within the market. however “Coming Soon” tag is on the website and it clearly means you'll be able to expect it to launch within the finish of the 2016 or early 2017.

It’s been virtually one and a half-year since the Google Glass was launched and currently things has modified lots. people are more comfortable with selfies and social networks, however still you'll be able to meet those that don’t just like the plan of carrying a try of sunglass with a mobile camera at public areas. Snapchat is presenting Spectacles as a fun device and is ready to drag it back and introduce one thing less arguable if it doesn’t work.

↧

PhotoScan - A new Google Photo feature

November 16, 2016, 1:40 am

≫ Next: Google Draw AI Experiment - Kids can learn by draw

≪ Previous: Top 8 Features of Snapchat Glass : You should know

Google has launched PhotoScan, a innovative application that allows users to scan recent photos which will thenceforth be hold on and retouched in Google Photos.
PhotoScan allows users to scan recent photos in numerous formats, and also, wherever necessary, to revive them to their former glory. The app options retouching, and rotation functionalities, together with options to enhance perspective and eliminate reflections. the concept is produce high-quality digital copies that provides a new life to recent photos.

Along with this new application, Google is additionally launching an update to Photos with a recently improved automatic improvement feature that permits users to effortlessly balance exposure and saturation to optimize image detail.

How will it work ?

The app opens to the camera and prompts users to position a photograph inside the frame.
Four dots ought to appear over the picture if done properly.
The app can automatically crop, rotate and color correct.
Google has additionally created area for these memories in Google Photos, that is accessed using the search ‘scan’.
This app is offered for each iOS and android users.

‘Scanned photos is saved in one tap to Google Photos to be organized, searchable, shared, and safely secured at high quality—for free.’

To scan the print, move your phone over every of the dots and hold it till the circle seems to be filled in.
Not solely is that this less expensive than a scanner, however as shortly as you’ve uploaded one photograph you'll be able to move right to the next.
And the app can automatically crop, rotate and color correct.

The latest version of Google Photos additionally includes 12 new filters that modify pictures by adding shadows, warming up colours, and adjusting saturation. There are advanced retouching choices (light, color, sharpness and style) that permit users to optimize the looks every kind of photos.
Google PhotoScan and Google Photos is downloaded for free of charge from Google Play (Android) and App Store (iOS).

↧

Google Draw AI Experiment - Kids can learn by draw

November 17, 2016, 12:23 am

≫ Next: HungerMoji - Android Notification Game [Source Code]

≪ Previous: PhotoScan - A new Google Photo feature

Google Draw, a brand new group of online artificial-intelligence experiments and one particularly stands out as each an habit-forming game and a desirable window on machine learning.

Quick, Draw challenges you to form a series of drawings in under twenty seconds for each one. The neural network tries to work out what you are drawing as you go. It learns from its mistakes and seeks to enhance its recognition skills.

I'm no nice visual creative person, however the AI racked up a powerful 5-out-of-6 score on my 1st round, properly identifying grass, a chopper, a penguin, a remote control and scissors. The one it lost was an outline of a feather, that I though was good, however the AI thought it looked a lot of sort of a mouse, ice cream or tennis racket.

There's an almost addictive thrill to wiggling with fast, Draw. The timer counts down and gets a lot of pressing as the seconds tick by. A voice guesses at your drawings, making it want a browser-based game of charades. generally it simply provides up, saying "I don't have any clue what you are drawing. Sorry i could not guess it." and so you are feeling like you've got let the poor AI down along with your tinny art skills.

In one stunningly bad round, I drew a kangaroo that sounded like a lizard, a sun that sounded like an exploded egg, a tornado that seemed like a cream horn and a basketball that resembled a cookie. however the neural network does not decide. It simply causes you to wish to play once more.

↧

HungerMoji - Android Notification Game [Source Code]

November 17, 2016, 8:53 pm

≫ Next: 10 Commont Jquery Mistakes You should Aware

≪ Previous: Google Draw AI Experiment - Kids can learn by draw

As you know Android has such a great scope in gaming field but you just need some innovation and different ideas which can attract the user to play your game again and again.
Today i am telling you a great idea of game with their source code. You'll play it on your notification bar. Its cool isn't it ?
This game named as HungerMoji and it is a unique Android game that takes place within the system notification drawer.

Players must feed their emoji which resides in a notification by dismissing any offending notifications, and collecting any bonus point notifications.

Download Source Code

Demo

Hope you enjoy and please give some time for feedback 😀😀😀.

↧

10 Commont Jquery Mistakes You should Aware

November 22, 2016, 2:26 am

≫ Next: How to enable Offline Download In Android

≪ Previous: HungerMoji - Android Notification Game [Source Code]

Well, everybody makes mistakes and and better part would be to not repeat them again. You must "always make new mistakes" :). jQuery is amazing library. however believe me, it will cause you to crazy if it's not used properly and with efficiency. It will hit performance of your page and you do not even understand it. Raise yourself that as a developer whereas writing jQuery code does one really think about performance or it's only if your client shouts.

I have listed down ten common jQuery mistakes from my personal expertise and their fixes that you just can be additionally doing while writing jQuery code.

1. Be courageous to remove jQuery

Sometimes things can be done easily via CSS without even thinking about jQuery but we don't realize it. Plain CSS is far better than jQuery. So be open and courageous to remove jQuery whenever needed.

2. Not using latest version of jQuery

jQuery team is keep on updating the jQuery library and the newer version comes with lots of bug fixes and performance enhancement. I understand that it is not always possible for you to use the latest version for your old projects but I suggest for your new projects, you can use latest version of jQuery.

3. Not using minified version of jQuery library

The jQuery library (when you download) comes in two versions.

Production (Compressed Version)
Development (Uncompressed Version)

For development purpose, you can choose the development version of .js file as if you want to make some changes then that can be easily done. But ensure that when your software or product goes on production, always use the production version of .js file as its size is 5 times lesser than the development version. This can save some amount of bandwidth.

4. Not loading jQuery from Google CDN

Google is sea of free services. Do you know that Google is also hosting jQuery libraries on its CDN(Content delivery network) and allows any website to use it for free.

Why to use Google CDN?

Caching: The most important benefit is caching. If any previously visited site by user is using jQuery from Google CDN then the cached version will be used. It will not be downloaded again.
Reduce Load: It reduces the load on your web server as it downloads from Google server's.
Serves fast : You will be also benefitted from speed point of view. As Google has dozen's of different servers around the web and it will download the jQuery from whichever server is closer to the user. Google's CDN has a very low latency, it can serve a resource faster than your webserver can.
Parellel Downloading: As the js file is on a separate domain, modern browsers will download the script in parallel with scripts on your domain.

5. Not loading jQuery locally when CDN fails

It is a good approach to always use CDN but sometimes what if the CDN is down (rare possibility though) but you never know in this world as anything can happen. So if you have loaded your jQuery from any CDN and it went down then your jQuery code will stop working and your client will start shouting.

I always recommend that write the code, if jQuery library is not loaded properly then it should use your local copy of jQuery.

<script type="text/javascript" src="http://ajax.googleapis.com/ajax/libs/jquery/1.9.0/jquery.min.js"></script>
<script type="text/javascript">
if (typeof jQuery == 'undefined')
{
  document.write(unescape("%3Cscript src='Scripts/jquery.1.9.0.min.js' type='text/javascript'%3E%3C/script%3E"));
}
</script>

It first loads the jQuery from Google CDN and then check the jQuery object. If jQuery is not loaded successfully then it will references the jQuery.js file from hard drive location. In this example, the jQuery.js is loaded from Scripts folder.

6. Not using selectors efficiently

Be smart while using selectors. As there are many ways to select element using selectors but that doesn't mean that all are equal. Always try to use ID and Element as selector as they are very fast. Even the class selectors are slower than ID selector.

When IDs are used as selector then jQuery internally makes a call to getElementById() method of Java script which directly maps to the element.

When Classes are used as selector then jQuery has to do DOM traversal.So when DOM traversal is performed via jQuery takes more time to select elements. In terms of speed and performance, it is best practice to use IDs as selector.

7. Using jQuery selectors repeatedly

Take a look at below jQuery code. The selectors are used thrice for 3 different operation.

$("#myID").css("color", "red");
$("#myID").css("font", "Arial");
$("#myID").text("Error occurred!");

The problem with above code is, jQuery has to traverse 3 times as there are 3 different statements.But this can be combined into a single statement.

$("#myID").css({ "color": "red", "font": "Arial"}).text("Error occurred!");

This will ensure that jQuery traverse only once through DOM while selecting the element.

8. Not knowing how selectors are executed

Do you know how the selectors are executed? Your last selectors is always executed first. For example, in below jQuery code, jQuery will first find all the elements with class ".myCssClass" and after that it will reject all the other elements which are not in "p#elmID".

$("p#elmID .myCssClass");

9. By not caching the stuff

Caching is an area which can give you awesome performance, if used properly and at the right place. While using jQuery, you should also think about caching. For example, if you are using any element in jQuery more than one time, then you must cache it. See below code.

$("#myID").css("color", "red");
//Doing some other stuff......
$("#myID").text("Error occurred!");

Now in above jQuery code, the element with #myID is used twice but without caching. So both the times jQuery had to traverse through DOM and get the element. But if you have saved this in a variable then you just need to reference the variable. So the better way would be,

var $myElement = $("#myID").css("color", "red");
//Doing some other stuff......
$myElement.text("Error occurred!");

So now in this case, jQuery won't need to traverse through the whole DOM tree when it is used second time. So in jQuery, Caching is like saving the jQuery selector in a variable. And using the variable reference when required instead of searching through DOM again.

10. Not checking while using various plugins

One of the great feature of jQuery is various plugins available created using jQuery are available for free. You like a jQuery plugin and start using it in your project. But while using plugins do you really consider,

File size
Performance of plugin
Browser Compatibility

Before using any plugin, always ensure that it must not hit performance of your page and it should not conflict with other plugins that you are using already.

Feel free to contact me for any help related to jQuery, I will gladly help you.

↧

How to enable Offline Download In Android

December 8, 2016, 3:44 am

≫ Next: Android Studio 2.2 is Out with Advanced Features

≪ Previous: 10 Commont Jquery Mistakes You should Aware

If you have mobile phones and the without internet, you want to see Android web or video files then a new feature in the Chrome browser is launched. Latest Chrome web pages and videos for you to watch later offline can download. This is many times the cost of your Internet data less.

You must know : Difference between Google Chrome V/s Microsoft Edge

How to Download Offline in Android

Chrome Browser has a new update on the latest and option to download the new browser is offline. This option will show up on the right of the web page.
Clicking the Download button will appear. Clicking it available on that page as well as photos, videos and text will be downloaded offline.
You can view it later. However, this feature of Chrome will not be able to download videos from video streaming sites.
In addition to the site, upload videos and downloading allowed then you can easily save offline.

Read this : Hack Any WhatsApp in Just 4 Steps

Important thing is that such videos Bluetooth, email and messaging apps such as WhatsApp and Hike can share and download the Google Chrome has a feature is introduced. Now if you have no videos, music or photos are downloaded and then for some reason your network is down as much as it has been downloaded will be saved. Whenever the network running again will resume from where you will download.
The advantage would be that you had to use it to download the data will be saved first.
What you think about this feature of Google . Please comment below.

↧

Android Studio 2.2 is Out with Advanced Features

December 22, 2016, 12:40 am

≫ Next: Android Floating Action Button Example

≪ Previous: How to enable Offline Download In Android

There are a huge number of Android app developers out there and it is no wonder having given it a wide range across the globe, in which majority of the developers use Android Studio for their app creation. There has been an estimated figure on its number as 92% of the top 125 apps and games make use of this Android Studio.

Android Studio is a free integrated development environment (IDE) for the Android platform developments which is based on the IntelliJ Java IDE. This year has been an exciting time for the Android Studio wherein they have come up with another latest version of it- the Android Studio 2.2.

What is Android Studio 2.2?

Google has recently launched Android Studio 2.2 which is the latest version of IDE, encrypted in it the new advanced set of features throughout every major phase of the development process like design, develop, build and test.

Now let us have a closer look on some of the features:

Design

Layout Editor:

This is one of a unique user interface assisting you to visually design the layouts in your app. This layout editor gives you the ease with simply dragging the widgets into the visual design editor. The layout editor has been modified in a way resembling a blueprint mode that hides the remaining details of your layout shortly, thereby giving you the option to check the arrangements and spacing of UI.

Constraint Layout:

A flexible layout manager designed for Android Studio’s brand new layout editor. Allowing you to create complex user interfaces without nesting multiple layouts.

Fundamentally, this is same to the Relative Layout, though Constraint layout has the added advantage of being more flexible and works perfectly with Android Studio’s new layout editor.

Layout Inspector:

A quick solution to inspect the view properties of your layout without leaving the IDE. Enabling you to debug a snapshot of your app layout running on the Android device.

Develop

Improved C++ Support:

Improved C++ support along with a number of bug fixes enables you to use CMake or ndk-build for compiling C++ projects from Gradle. Shifting the projects from CMake build systems to Android Studio can be done now with ease.

Firebase Plugin:

A new pack of services meant for the developers out there. It enables you in developing high-quality apps, widen your user base and driving in your monetary level.

Samples Browser:

With this option, referencing the Android sample code is made now much easier. It assists you in receiving high-quality sample codes provided by Google based on the symbol highlighted in your project.

Build

Instant Run:

This is one of the major added options for the quick and scalable functioning in the Android developments significantly improving the edit, build and run functions more easily for the developers.

Build Cache:

It mainly focuses on generating cache leading to the time reduction of creating both full and incremental builds.

Test

Espresso Test Recorder:

With this, you can now easily create apps wherein it records all the UI interactions and converts it into a fully reusable test code.

APK Analyzer:

With the APK analyzer option you can easily analyze your APK contents understanding the size of each component. A helpful feature used at the debugging phase of the multi-dex issues.

There are in fact a lot more advanced features, check it out in the below link.

Conclusion

Android Studio 2.2 is certainly a step ahead with its advanced set of features and is sure to impact a huge publicity among the Android app development companies in meeting up all their expectations.

↧

Android Floating Action Button Example

December 22, 2016, 9:18 pm

≫ Next: Boom Menu For Android - New Era of Android Menu UI

≪ Previous: Android Studio 2.2 is Out with Advanced Features

In my previous article i am showing Android Floating label example and FAB loading bar which are come from Google Material design. And today we'll discuss about yet another implementation of Floating Action Button for Android with lots of features.

Android Floating Action Button based on Material Design specification

Requirements

The library requires Android API Level 14+.

Download

Features

Ripple effect on Android Lollipop devices
Option to set custom normal/pressed/ripple colors
Option to set custom shadow color and offsets
Option to disable shadow for buttons and (or) labels
Option to set custom animations
Option to set custom icon drawable
Support for normal 56dp and mini 40dp button sizes
Custom FloatingActionMenu icon animations
Option to expand menu up and down
Option to show labels to the left and to the right of the menu
Option to show circle progress on FloactinActionButton
Option to add button to the FloatingActionMenu programmatically
Option to dim the FloatinActionMenu's background
Option to remove all buttons from the FloatingActionMenu
Option to set a label for the FloatingActionMenu's button

Demo

Usage

Add a dependency to your build.gradle:

dependencies {
    compile 'com.github.clans:fab:1.6.4'
}

See this(Not for beginner) : Boom Menu For Android Example - New Era of Android UI

Add the com.github.clans.fab.FloatingActionButton to your layout XML file.

<FrameLayoutxmlns:android="http://schemas.android.com/apk/res/android"
xmlns:fab="http://schemas.android.com/apk/res-auto"
android:layout_width="match_parent"
android:layout_height="match_parent">

<ListView
android:id="@+id/list"
android:layout_width="match_parent"
android:layout_height="match_parent"/>

<com.github.clans.fab.FloatingActionButton
android:id="@+id/fab"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:layout_gravity="bottom|right"
android:layout_marginBottom="8dp"
android:layout_marginRight="8dp"
android:src="@drawable/ic_menu"
fab:fab_colorNormal="@color/app_primary"
fab:fab_colorPressed="@color/app_primary_pressed"
fab:fab_colorRipple="@color/app_ripple"/>

</FrameLayout>

You can set an icon for the FloatingActionButton using android:src xml attribute. Use drawables of size 24dp as specified by guidlines. Icons of desired size can be generated with Android Asset Studio.

FloatingActionButton

Here are all the FloatingActionButton's xml attributes with their default values which means that you don't have to set all of them:

<com.github.clans.fab.FloatingActionButton
android:id="@+id/fab"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:layout_gravity="bottom|right"
android:layout_marginBottom="8dp"
android:layout_marginRight="8dp"
android:src="@drawable/your_icon_drawable"
app:fab_colorNormal="#DA4336"
app:fab_colorPressed="#E75043"
app:fab_colorRipple="#99FFFFFF"
app:fab_showShadow="true"
app:fab_shadowColor="#66000000"
app:fab_shadowRadius="4dp"
app:fab_shadowXOffset="1dp"
app:fab_shadowYOffset="3dp"
app:fab_size="normal"
app:fab_showAnimation="@anim/fab_scale_up"
app:fab_hideAnimation="@anim/fab_scale_down"
app:fab_label=""
app:fab_progress_color="#FF009688"
app:fab_progress_backgroundColor="#4D000000"
app:fab_progress_indeterminate="false"
app:fab_progress_max="100"
app:fab_progress="0"
app:fab_progress_showBackground="true"/>

All of these FloatingActionButton's attributes has their corresponding getters and setters. So you can set them programmatically.

Floating action menu

Here are all the FloatingActionMenu's xml attributes with their default values which means that you don't have to set all of them:

<com.github.clans.fab.FloatingActionMenu
android:id="@+id/menu"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:layout_alignParentBottom="true"
android:layout_alignParentRight="true"
android:layout_marginRight="10dp"
android:layout_marginBottom="10dp"
android:layout_marginLeft="10dp"
fab:menu_fab_size="normal"
fab:menu_showShadow="true"
fab:menu_shadowColor="#66000000"
fab:menu_shadowRadius="4dp"
fab:menu_shadowXOffset="1dp"
fab:menu_shadowYOffset="3dp"
fab:menu_colorNormal="#DA4336"
fab:menu_colorPressed="#E75043"
fab:menu_colorRipple="#99FFFFFF"
fab:menu_animationDelayPerItem="50"
fab:menu_icon="@drawable/fab_add"
fab:menu_buttonSpacing="0dp"
fab:menu_labels_margin="0dp"
fab:menu_labels_showAnimation="@anim/fab_slide_in_from_right"
fab:menu_labels_hideAnimation="@anim/fab_slide_out_to_right"
fab:menu_labels_paddingTop="4dp"
fab:menu_labels_paddingRight="8dp"
fab:menu_labels_paddingBottom="4dp"
fab:menu_labels_paddingLeft="8dp"
fab:menu_labels_padding="8dp"
fab:menu_labels_textColor="#FFFFFF"
fab:menu_labels_textSize="14sp"
fab:menu_labels_cornerRadius="3dp"
fab:menu_labels_colorNormal="#333333"
fab:menu_labels_colorPressed="#444444"
fab:menu_labels_colorRipple="#66FFFFFF"
fab:menu_labels_showShadow="true"
fab:menu_labels_singleLine="false"
fab:menu_labels_ellipsize="none"
fab:menu_labels_maxLines="-1"
fab:menu_labels_style="@style/YourCustomLabelsStyle"
fab:menu_labels_position="left"
fab:menu_openDirection="up"
fab:menu_backgroundColor="@android:color/transparent"
fab:menu_fab_label="your_label_here"
fab:menu_fab_show_animation="@anim/my_show_animation"
fab:menu_fab_hide_animation="@anim/my_hide_animation">

<com.github.clans.fab.FloatingActionButton
android:id="@+id/menu_item"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:src="@drawable/ic_star"
fab:fab_size="mini"
fab:fab_label="Menu item 1"/>

</com.github.clans.fab.FloatingActionMenu>

If you're using custom style for labels - other labels attributes will be ignored.

Labels shadow preferences depends on their corresponding FloatingActionButtons' shadow preferences.

↧

Boom Menu For Android - New Era of Android Menu UI

December 22, 2016, 9:31 pm

≫ Next: [Resolved] Preselect and Value Stay Problem in Multiple Select in Materialize

≪ Previous: Android Floating Action Button Example

Are you bored to see typical plain menu in android app then a great news for Android developer which is about a new era of Android UI for menu purpose which can boom your mind by this.

This concept came from floating button of Android and now firstly just look at this.

See this : Android Floating Action Button Example

Tired of these menu buttons?

Old Menu Buttons

Why not try these:

Circle Ham

List Share

Yes, this library is about a menu which can ... BOOM! It is cool isn't it.

Download

Documentation Chapters

Basic Usage
How to use BMB in just several lines of code?
Simple Circle Button
Add simple circle buttons with just an image for each to BMB.
Text Inside Circle Button
Add text inside circle buttons with a text and image inside for each to BMB.
Text Outside Circle Button
Add text outside circle buttons with a text and image outside for each to BMB.
Ham Button
Add ham buttons with with a title, subtitle and image inside for each to BMB.
Share Style
Make a share-style BMB.
Button Place Alignments
Place all the buttons to anywhere on screen.
Different Ways to Boom
Different animations when the buttons boom or re-boom.
Ease Animations for Buttons
Use different and cute ease-animations for buttons.
Different Order for Buttons
Different order enum for boom-buttons.
Other Animations Attributes for Buttons
Delay, duration, rotate-degrees, frames...
Click Event and Listener
Listener for clicking each button or animation-states.
Control BMB
Boom or re-boom BMB programmatically.
Use BMB in Action Bar
How to put BMB in action bar?
Use BMB in List
Matters need attention when you need a BMB in list-view or recycler-view.
Use BMB in Fragment
Example for use BMB in fragment.
Attributes for BMB or Pieces on BMB
How to change the size or margins of dots on BMB?
Cache Optimization & Boom Area
What if I want BMB to boom in just its parent-view?
Version History
What's more for every version?
Structure for BMB
Structure for BMB when I designed it, for sharing and communicating.

↧

[Resolved] Preselect and Value Stay Problem in Multiple Select in Materialize

December 26, 2016, 10:18 pm

≫ Next: How to Add PanoramaImageView in Android

≪ Previous: Boom Menu For Android - New Era of Android Menu UI

Hello today i am here with two problem which occur with most of developer when they start with Google's Materialized theme. In my project i also used this awesome designing them but i stuck with problems in Multiple select option. And now let us look with problem and their solution.

Problem 1 :

MaterializeCSS - Multiple Select - value stays in array after unselecting

When I unselect value in multiple selection in MaterializeCSS, it's still in the value array and I can't find a way to repair it. It works if I unselect option from original select with some function but

(please ignore the error on the screenshot, it's not related)

As you can see in screenshot firstly i selected Option 2 and Option 4 and got value in console is:

["test2", "test4"]

But when i unselect Option 2 then value of that is remain as you can show in console of second image.

Solution :

I had the same problem and wrote a small workaround.

$(document).ready(function () {
    $('select').material_select();
    $('select').change(function(){
var newValuesArr = [],
            select = $(this),
            ul = select.prev();
        ul.children('li').toArray().forEach(function (li, i) {
if ($(li).hasClass('active')) {
                newValuesArr.push(select.children('option').toArray()[i].value);
            }
        });
        select.val(newValuesArr);
    });
});

You can also see demo here

Problem 2 :

Dynamically select Materialized multiple select

After the resolving first problem i stuck again when i need to select some data automatically from database. But it doesn't have any solution for this. Then i send this issue to MaterializeCSS forum and they update their JS and gave me updated JS. So now if you want to do this then you need to follow below step:

This is multiple dropdown of Materialize

<divclass="input-field col s12">
<selectmultiple>
<optionvalue=""disabledselected>Choose your option</option>
<optionvalue="1"selected>Option 1</option>
<optionvalue="2">Option 2</option>
<optionvalue="3"selected>Option 3</option>
</select>
<label>Materialize Multiple Select</label>
</div>

And you change little change in your Javascript section :

<script src="http://materializecss.com/bin/materialize.js" />
$(document).ready(function () {
        $('select').material_select();
    });

You can also see demo here

↧

How to Add PanoramaImageView in Android

December 27, 2016, 1:51 am

≫ Next: Expanding Menu View in Android

≪ Previous: [Resolved] Preselect and Value Stay Problem in Multiple Select in Materialize

Today i am going to tell you a new Custom view which can blow your mind. This is using Gyroscope sensor.
An imageView can auto scroll with device rotating.

ScreenShots

horizontal sample

vertical sample

recyclerview sample

Include PanoramaImageView to Your Project

With gradle:

dependencies {
   compile 'com.gjiazhe:PanoramaImageView:1.0'
}

Use PanoramaImageView in Layout File Just Like ImageView

<com.gjiazhe.panoramaimageview.PanoramaImageView
android:id="@+id/panorama_image_view"
android:layout_width="match_parent"
android:layout_height="match_parent"
android:src="@drawable/img"
app:piv_enablePanoramaMode="true"
app:piv_show_scrollbar="true"
app:piv_invertScrollDirection="false"/>

Description of Attributes

Attributes	Format	Default	Description
piv_enablePanoramaMode	boolean	true	Enable panorama effect or not.
piv_show_scrollbar	boolean	true	Show scrollbar or not.
piv_invertScrollDirection	boolean	false	Invert the scroll direction or not

You can also check :Add Boom Menu For Android

All the attributes can also be set in java code:

panoramaImageView.setEnablePanoramaMode(true);
panoramaImageView.setEnableScrollbar(true);
panoramaImageView.setInvertScrollDirection(false);

Register the GyroscopeObserver

In Activity or Fragment using PanoramaImageView, you should register the GyroscopeObserver in onResume() and remember to unregister it in onPause().

publicclassMyActivityextends AppCompatActivity {

private GyroscopeObserver gyroscopeObserver;

@Override
protectedvoidonCreate(Bundle savedInstanceState) {
super.onCreate(savedInstanceState);
        setContentView(R.layout.activity_main);
// Initialize GyroscopeObserver.
        gyroscopeObserver = new GyroscopeObserver();
// Set the maximum radian the device should rotate to show image's bounds.
// It should be set between 0 and π/2.
// The default value is π/9.
        gyroscopeObserver.setMaxRotateRadian(Math.PI/9);

        PanoramaImageView panoramaImageView = (PanoramaImageView) findViewById(R.id.panorama_image_view);
// Set GyroscopeObserver for PanoramaImageView.
        panoramaImageView.setGyroscopeObserver(gyroscopeObserver);
    }

@Override
protectedvoidonResume() {
super.onResume();
// Register GyroscopeObserver.
        gyroscopeObserver.register(this);
    }

@Override
protectedvoidonPause() {
super.onPause();
// Unregister GyroscopeObserver.
        gyroscopeObserver.unregister();
    }
}

Set OnPanoramaScrollListener to observe scroll state

Check this : Add Android Fingerprint Authentication in Android

If you want to get callback when the image scrolls, set an OnPanoramaScrollListener for PanoramaImageView.

panoramaImageView.setOnPanoramaScrollListener(new PanoramaImageView.OnPanoramaScrollListener() {
@Override
publicvoidonScrolled(PanoramaImageView view, float offsetProgress) {
// Do something here.
// The offsetProgress range from -1 to 1, indicating the image scrolls
// from left(top) to right(bottom).
    }
});

↧

Expanding Menu View in Android

December 30, 2016, 12:59 am

≫ Next: Google introduce ExifInterface Support Library

≪ Previous: How to Add PanoramaImageView in Android

If you want to make a shopping Android app then a great library is for you which name as Expanding Menu View Library. This gives you a great look in your simple android application.
This is a library to help creating expanding views with animation in Android.

Download Source Code

Adding the Library to gradle file

dependencies {
    compile 'com.diegodobelo.expandingview:expanding-view:0.9.3'
}

Demo

Using the Library

Layouts

First of all include the ExpandingList in your Activity (or Fragment) layout. This will be the list of items (ExpandItem):

<com.diegodobelo.expandingview.ExpandingList
android:id="@+id/expanding_list_main"
android:layout_width="match_parent"
android:layout_height="match_parent"/>

Now create a new layout (xml) file to represent the item (such as res/layout/expanding_item.xml). This will be the Item (header) that can be expanded to show sub items:

<RelativeLayoutxmlns:android="http://schemas.android.com/apk/res/android"
xmlns:app="http://schemas.android.com/apk/res-auto"
android:layout_width="match_parent"
android:layout_height="94dp">
<TextView
android:id="@+id/title"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:layout_centerVertical="true"
android:gravity="center_vertical|left"
android:textSize="22sp"/>
</RelativeLayout>

Create another layout file to represent the sub items (such as /res/expanding_sub_item.xml). This will be the sub items that will be shown when the Item is expanded:

<RelativeLayoutxmlns:android="http://schemas.android.com/apk/res/android"
xmlns:app="http://schemas.android.com/apk/res-auto"
android:layout_width="match_parent"
android:layout_height="48dp">
<TextView
android:id="@+id/sub_title"
android:layout_width="match_parent"
android:layout_height="match_parent"
android:layout_centerVertical="true"
android:layout_marginLeft="8dp"
android:textSize="18sp"
android:gravity="center_vertical|left"/>
</RelativeLayout>

Now create a layout file (such as /res/expanding_layout.xml) to represent the whole item, including both item layout and sub item layout. We will explain each custom attribute later:

<com.diegodobelo.expandingview.ExpandingItem
xmlns:android="http://schemas.android.com/apk/res/android"
xmlns:app="http://schemas.android.com/apk/res-auto"
android:layout_width="match_parent"
android:layout_height="wrap_content"
app:item_layout="@layout/expanding_item"
app:sub_item_layout="@layout/expanding_sub_item"
app:indicator_size="42dp"
app:indicator_margin_left="16dp"
app:indicator_margin_right="16dp"
app:show_indicator="true"
app:show_animation="true"
app:start_collapsed="true"
app:animation_duration="250"/>

Note that we included expanding_item and expanding_sub_item layouts created before.

Java code

Now that you have all the required layouts you are able to use them in Java code. Let's start inflating the ExpandingList:

ExpandingList expandingList = (ExpandingList) findViewById(R.id.expanding_list_main);

Create a new ExpandingItem in the ExpandingList. This method receives the expanding_layout created before. Yes! We can have different layouts for different items in the same list. The items will be created based on expanding_item layout and the sub items will be created based on expanding_sub_item layout:

ExpandingItem item = expandingList.createNewItem(R.layout.expanding_layout);

/*ExpandingItem extends from View, so you can call
findViewById to get any View inside the layout*/
(TextView) item.findViewById(R.id.title)).setText("It Works!!");

Let's create the sub items. There is a method to create items in batch:

//This will create 5 items
item.createSubItems(5);

//get a sub item View
View subItemZero = item.getSubItemView(0);
((TextView) subItemZero.findViewById(R.id.sub_title)).setText("Cool");

View subItemOne = item.getSubItemView(1);
((TextView) subItemOne.findViewById(R.id.sub_title)).setText("Awesome");

...

For each item you can set the indicator color and the indicator icon:

item.setIndicatorColorRes(R.color.blue);
item.setIndicatorIconRes(R.drawable.ic_icon);

ExpandingItem layout attributes

Attribute Name	Type	Default Value	Meaning	Mandatory
item_layout	reference		The layout for the Item (header).	Yes
sub_item_layout	reference		The layout for the sub items.	Yes
separator_layout	reference		A layout to separate items.	No
indicator_size	dimension	0dp	The indicator size in dp.	No
indicator_margin_left	dimension	0dp	The margin between the indicator and its left.	No
indicator_margin_right	dimension	0dp	The margin between the indicator and its right.	No
show_indicator	boolean	true	true if you want to show the indicator. false otherwise.	No
show_animation	boolean	true	true if you want to show animations. false otherwise.	No
start_collapsed	boolean	true	true if you want the sub views to start collapsed. false otherwise.	No
animation_duration	integer	300ms	The animations duration in milliseconds.	No

ExpandingList public methods

`public void createNewItem(int layoutId)`

Method to create and add a new item.
Parameters: `layoutId` — The item Layout

`public void removeItem(ExpandingItem item)`

Method to remove an item.
Parameters: `item` — The item to be removed, of type {@link ExpandingItem}

`public void removeAllViews()`

Method to remove all items.

ExpandingItem public methods

`public void setStateChangedListener(OnItemStateChanged listener)`

Set a listener to listen item stage changed.
Parameters: `listener` — The listener of type {@link OnItemStateChanged}

`public boolean isExpanded()`

Tells if the item is expanded.
Returns: true if expanded. false otherwise.

`public int getSubItemsCount()`

Returns the count of sub items.
Returns: The count of sub items.

`public void collapse()`

Collapses the sub items.

`public void toggleExpanded()`

Expand or collapse the sub items.

`public void setIndicatorColorRes(int colorRes)`

Set the indicator color by resource.
Parameters: `colorRes` — The color resource.

`public void setIndicatorColor(int color)`

Set the indicator color by color value.
Parameters: `color` — The color value.

`public void setIndicatorIconRes(int iconRes)`

Set the indicator icon by resource.
Parameters: `iconRes` — The icon resource.

`public void setIndicatorIcon(Drawable icon)`

Set the indicator icon.
Parameters: `icon` — Drawable of the indicator icon.

`@Nullable public View createSubItem()`

Creates a sub item based on sub_item_layout Layout, set as ExpandingItem layout attribute.
Returns: The inflated sub item view.

`@Nullable public View createSubItem(int position)`

Creates a sub item based on sub_item_layout Layout, set as ExpandingItem layout attribute. If position is -1, the item will be added in the end of the list.
Parameters: `position` — The position to add the new Item. Position should not be greater than the list size.
Returns: The inflated sub item view.

`public void createSubItems(int count)`

Creates as many sub items as requested in {@param count}.
Parameters: `count` — The quantity of sub items.

`public View getSubItemView(int position)`

Get a sub item at the given position.
Parameters: `position` — The sub item position. Should be > 0.
Returns: The sub item inflated view at the given position.

`public void removeSubItemAt(int position)`

Remove sub item at the given position.
Parameters: `position` — The position of the item to be removed.

`public void removeSubItemFromList(View view)`

Remove the given view representing the sub item. Should be an existing sub item.
Parameters: `view` — The sub item to be removed.

`public void removeSubItem(View view)`

Remove the given view representing the sub item, with animation. Should be an existing sub item.
Parameters: `view` — The sub item to be removed.

`public void removeAllSubItems()`

Remove all sub items.

↧

Google introduce ExifInterface Support Library

January 1, 2017, 8:52 pm

≫ Next: Add Notification Badge in Android

≪ Previous: Expanding Menu View in Android

This is a class for reading and writing Exif tags in a JPEG file or a RAW image file. Supported formats are: JPEG, DNG, CR2, NEF, NRW, ARW, RW2, ORF and RAF.
Attribute mutation is supported for JPEG image files.

One thing that is important to understand with Exif data is that there are no required tags: each and every tag is optional - some services even specifically strip Exif data. Therefore throughout your code, you should always handle cases where there is no Exif data, either due to no data for a specific attribute or an image format that doesn't support Exif data at all (say, the ubiquitous PNGs or WebP images).

Add the ExifInterface Support Library to your project with the following dependency:

compile "com.android.support:exifinterface:25.1.0"

But when an Exif attribute is exactly what you need to prevent a mis-rotated image in your app, the ExifInterface Support Library is just what you need to #BuildBetterApps

The basics are still the same: the ability to read and write Exif tags embedded within image files: now with 140 different attributes (almost 100 of them new to Android 7.1/this Support Library!) including information about the camera itself, the camera settings, orientation, and GPS coordinates.

Camera Apps: Writing Exif Attributes

For Camera apps, the writing is probably the most important - writing attributes is still limited to JPEG image files. Now, normally you wouldn't need to use this during the actual camera capturing itself - you'd instead be calling the Camera2 API CaptureRequest.Builder.set() with JPEG_ORIENTATION, JPEG_GPS_LOCATION or the equivalents in the Camera1 Camera.Parameters. However, using ExifInterface allows you to make changes to the file after the fact (say, removing the location information on the user's request).

Reading Exif Attributes

For the rest of us though, reading those attributes is going to be our bread-and-butter; this is where we see the biggest improvements.

Firstly, you can read Exif data from JPEG and raw images (specifically, DNG, CR2, NEF, NRW, ARW, RW2, ORF, PEF, SRW and RAF files). Under the hood, this was a major restructuring, removing all native dependencies and building an extensive test suite to ensure that everything actually works.

For apps that receive images from other apps with a content:// URI (such as those sent by apps that target API 24 or higher), ExifInterface now works directly off of an InputStream; this allows you to easily extract Exif information directly out of content:// URIs you receive without having to create a temporary file.

Uri uri; // the URI you've received from the other app
InputStream in;
try {
  in = getContentResolver().openInputStream(uri);
  ExifInterface exifInterface = new ExifInterface(in);
// Now you can extract any Exif tag you want
// Assuming the image is a JPEG or supported raw format
} catch (IOException e) {
// Handle any errors
} finally {
if (in != null) {
try {
      in.close();
    } catch (IOException ignored) {}
  }
}

Note: ExifInterface will not work with remote InputStreams, such as those returned from a HttpURLConnection. It is strongly recommended to only use them with content:// or file:// URIs.

For most attributes, you'd simply use the getAttributeInt(), getAttributeDouble(), or getAttribute() (for Strings) methods as appropriate.

One of the most important attributes when it comes to displaying images is the image orientation, stored in the aptly-named TAG_ORIENTATION, which returns one of the ORIENTATION_ constants. To convert this to a rotation angle, you can post-process the value.

int rotation = 0;
int orientation = exifInterface.getAttributeInt(
    ExifInterface.TAG_ORIENTATION,
    ExifInterface.ORIENTATION_NORMAL);
switch (orientation) {
case ExifInterface.ORIENTATION_ROTATE_90:
    rotation = 90;
break;
case ExifInterface.ORIENTATION_ROTATE_180:
    rotation = 180;
break;
case ExifInterface.ORIENTATION_ROTATE_270:
    rotation = 270;
break;
}

There are some helper methods to extract values from specific Exif tags. For location data, the getLatLong() method gives you the latitude and longitude as floats and getAltitude() will give you the altitude in meters. Some images also embed a small thumbnail. You can check for its existence with hasThumbnail() and then extract the byte[] representation of the thumbnail with getThumbnail() - perfect to pass to BitmapFactory.decodeByteArray().

↧

Add Notification Badge in Android

January 3, 2017, 1:37 am

≫ Next: Why and How to use JRebel in Java

≪ Previous: Google introduce ExifInterface Support Library

Notification Badge is a simple feature, but it’s hugely useful because it lets you know exactly which apps need your attention, and how much attention they need.

For some reason, Google has yet to create a similar badge system in Android. Instead, notifications are collected in the notification panel, so users don’t know how many notifications they have from each app until they swipe open the panel and look.

Alternative : Make Badge(Item count) in Android

Today i am telling you about a library which can help to create notification badge with animation.

Gradle

compile 'com.nex3z:notification-badge:0.1.0'

Usage

Use setNumber(int number) or setText(String text) to add number or arbitrary text on the badge. Use clear() to clear the badge.

Downlaod library

First of all you need to define that badge in

AndroidManifest.xml

<application
android:allowBackup="true"
android:icon="@mipmap/ic_launcher"
android:label="@string/app_name"
android:supportsRtl="true"
android:theme="@style/AppTheme">
<activityandroid:name=".MainActivity">
<intent-filter>
<actionandroid:name="android.intent.action.MAIN"/>

<categoryandroid:name="android.intent.category.LAUNCHER"/>
</intent-filter>
</activity>
</application>

Now create a layout where i notify the action and for this demo i am creating three button -->> Increase, Many Notification and Clear

activity_main.xml

<ImageView
android:id="@+id/icon"
android:layout_width="48dp"
android:layout_height="48dp"
android:layout_marginTop="8dp"
android:src="@drawable/ic_chat_24dp"/>

<com.nex3z.notificationbadge.NotificationBadge
android:id="@+id/badge"
android:layout_width="28dp"
android:layout_height="28dp"
android:layout_toRightOf="@id/icon"
android:layout_alignTop="@id/icon"
android:layout_marginLeft="-16dp"
android:layout_marginTop="-8dp"
app:badgeBackground="@drawable/badge_bg_with_shadow"
app:maxTextLength="2"/>

<Button
android:id="@+id/increase"
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:layout_below="@id/icon"
android:text="@string/increase"/>

<Button
android:id="@+id/many"
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:layout_below="@id/increase"
android:text="@string/many"/>

<Button
android:id="@+id/clear"
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:layout_below="@id/many"
android:text="@string/clear"/>

And finally this time to perform action by

MainActivity.java

importandroid.os.Bundle;
importandroid.support.v7.app.AppCompatActivity;
importandroid.view.View;
importandroid.widget.Button;

importcom.nex3z.notificationbadge.NotificationBadge;

publicclassMainActivityextends AppCompatActivity {

    NotificationBadge mBadge;
privateint mCount = 0;

@Override
protectedvoidonCreate(Bundle savedInstanceState) {
super.onCreate(savedInstanceState);
        setContentView(R.layout.activity_main);

        mBadge = (NotificationBadge) findViewById(R.id.badge);

        Button increase = (Button) findViewById(R.id.increase);
        increase.setOnClickListener(new View.OnClickListener() {
@Override
publicvoidonClick(View view) {
                mBadge.setNumber(++mCount);
            }
        });

        Button many = (Button) findViewById(R.id.many);
        many.setOnClickListener(new View.OnClickListener() {
@Override
publicvoidonClick(View view) {
                mCount = 98;
                mBadge.setNumber(mCount);
            }
        });

        Button clear = (Button) findViewById(R.id.clear);
        clear .setOnClickListener(new View.OnClickListener() {
@Override
publicvoidonClick(View view) {
                mCount = 0;
                mBadge.setNumber(mCount);
            }
        });
    }
}

I hope you got the point and enjoy coding.

↧

Why and How to use JRebel in Java

January 19, 2017, 10:26 pm

≫ Next: How to Use JRebel for Android

≪ Previous: Add Notification Badge in Android

First of all you should know about what is JRebel and why you should use it with your Java program. This post reviews some of the great advantages of using JRebel in Java development, and how eradicating the daily, time-consuming process of redeploying your application to see your changes has long-lasting effects on the way that you will write code in the future.

What is JRebel ?

According to their website -

"JRebel is an anything-Java plugin that speeds up JVM-based development (Java, Scala, Groovy) by reloading changes made in your workspace into a running JVM, without restarts or redeploys, maintaining the state of the application while you’re coding."

Don't worry if you're Android Dev :How to Use JRebel for Android

What would you rather be doing?

According to above image you can easily find that when you make a Java web application then whenever you change in your Java code you need to build, deploy and then run your changes which take some crucial time. But when you use JRebel then it save your time to build and deploy. Now you just change your Java code and test your program instead of build and deploy your application.

How to add JRebel in your IDE(Eclipse, MyEclipse, Netbeans etc.)

Firstly go to JRebel website.
Then follow the instruction which are given in above link.
Activate your JRebel plugin.
After activating right click on your project in IDE and find JRebel option.
Add JRebel in that project which create rebel.xml in WEB-INF/classes.
Thats it. Change your code and run the program.

Demo :

Why you use JRebel ?

Advantage #1: See all changes immediately

Does your code smell? Find out really quickly and run your own mini-tests throughout the day. Remove bad code before it goes to QA and they make fun of you.

With JRebel, you get to see the your changes immediately after compiling them. When the IDE builds classes automatically, the process is as fast as using a scripting language or testing your latest HTML changes. In fact, I urge you to try compiling the following code snippet, start it up with JRebel enabled, and change the “Hello World” part to something else while the program is running.

publicclassTestLoop {
publicstaticvoidmain(String... args) throws Exception {
while (true) {
            saySomething();
            Thread.sleep(1000);
        }
    }
privatestaticvoidsaySomething() {
        System.out.println("Hello World");
    }
}

For me, the console output following the JRebel banner was (yes, I write fast):

Hello World
Hello World
Hello World
JRebel: Reloading class 'TestLoop'.
Goodbye Redeploys
Goodbye Redeploys

This opens up a world of possibilities that have never before been available for developers. It combines the power of Java with the fast turnaround of dynamic scripting languages. You don’t have to queue up gazillion changes before redeploying. You can code by making small incremental changes and alt-tabbing to your browser to verify the results.

Advantage #2: Fine-tuning through incremental changes

This works best when you have to implement some logic that is tricky to get just right. Imagine you have to add validation to your form. With JRebel, you can add the rules field by field and play around with the way errors should be displayed until satisfied by the results. You may want to add some new fields to your controller or validator. That is supported, too. By the way, JRebel preserves all application state, which means it does not matter how you reached the page you’re modifying.

JRebel actually goes way beyond “simple” class reloading. Applications today take advantage of frameworks and JEE technologies, such as EJBs, Hibernate, Spring Framework, JSF, Seam, Struts, etc. The thing with those frameworks and technology implementations is that they initialize themselves when the application starts up and generate lots of metadata. It is usually a good thing since nobody wants Hibernate to re-initialize itself for every HTTP request (that’s what PHP programs have to do, by the way, even with opcode caching enabled).

JRebel has addressed this by having explicit support for almost 60 popular frameworks and application containers. This means when you make a change to a “special” resource such as adding a new EJB, changing JPA mappings (either via xml or annotations), or add new Spring Beans (again, either via xml or annotations), JRebel updates the metadata to reflect those changes. Your job is to press F5 in the browser.

Advantage #3: Debug remotely for fun and profit

JRebel makes remote debugging more useful than ever. Imagine you are using it to hunt down some difficult bug. Now you can hit a breakpoint, investigate what’s going on, make a change, resume execution, and see the result almost immediately!

The new code kicks in right after compiling the code, next time when the affected method is invoked. You can use this approach to fix discovered bugs immediately or to quickly add some debugging statements in the code (just remember to avoid committing them). You can even drop to a parent frame and check the effect immediately!

Advantage #4: Explore the code with ease

One of the benefits of small turnaround is that you can test the framework features interactively. You can write a small method to test some functionality of a new framework and learn hands-on how it works. Python and Scala even have interactive shells.

The problem here is the applicability in real world. While exploring the API of Java libraries is easy (even if they lack documentation), it can be ridiculously tedious in some other languages, especially when they are dynamic in nature, lack documentation, and/or miss out on good IDE support.

Even if you can indeed enjoy the fast turnaround with PHP and the like, it has little net value when you have to dig into some big and obscure framework (like Joomla CMS, which I actually tried that a few years ago, and it wasn’t pleasant) to see how it works. The ability to ctrl-click and dive into the code, as well as enjoy the auto-completion would be great, if only they were there. Granted, there is some support for it in modern IDEs, but it is nowhere nearly as useful. The difference is in the order of magnitudes.

Java, on the other hand, suffers from the fact that while the 3rd-party code is easily navigable, writing the stubs and running them can be tedious or time consuming. Writing stand-alone tests requires effort to set up and are not exactly a substitute for a real-world use case.

Embedding a test code into an existing application is simple and easy, but restarting the server or redeploying the app to see the changes takes time...unless you’re using JRebel to eliminate the need to restart.

Advantage #5: Deploying code locally or remotely

It’s great when a developer can check out code from a repository, compile it, and deploy the thing in his own machine. Sometimes this is not possible due to performance reasons, or the whole system is simply too complex. Sometimes the system has to run in a different network than the developer’s, which cannot be easily simulated. Some developers may be using low-powered laptops. Whatever the reason, sometimes running the app in the same machine is not an option.

I can almost hear developers cry (myself included) when that is the case. JRebel can help even in this case and the capability to make a system deployable on a developer’s machine is not crucial anymore.

JRebel Remoting (beta) was introduced from version 4.6, which lets you push changes from a developer’s machine to a remote server. It works over HTTP on the same port as the server is serving the application itself, therefore needs close to zero configuration to set up and no changes to firewall settings. In the screencast, we show JRebel Remoting in action when deploying to Amazon EC2, so if it can work well in the cloud far away, it will definitely work in your LAN or on the other side of the city.

You can even let multiple developers loose on a single application instance on a remote machine (but not simultaneously). By default, JRebel Remoting reverts back to the initial codebase when the app server is restarted. You can set up an application on the remote machine, have a developer have his way with it, restart it (ok, there’s one redeploy, sorry), and allow another developer do his magic on the app. This is the case when both developers are making changes to the same application. When there are several apps deployed to the remote server, developers can bombard them without stepping on each other’s toes.

Advantage #6: No restarts = More time for better stuff

No restarts means no forced interruptions, which means devs have more time for doing better things. Sometimes it’s hard to maintain concentration on a single task for more than a few minutes, and that’s when the level of focus is totally up to me. If there was a circus clown jumping out from behind a door every 6 minutes to frighten me, I couldn’t imagine working properly throughout the day.

With JRebel, you get more time to meet your deadlines, research & learn and *gasp* communicate with people. According to a recent report on Developer Stress, it came out that developers really do care about making deadlines and their level of expertise, wishing they had more time to spend on education. We also discovered that for every 1 hour developers spend writing code, they spend 30 minutes dealing with communication overhead (i.e., meetings, chats, reporting/timesheets, etc.). JRebel’s instantly class reloading magic simply frees up more time for doing the things that developers are already spending time on. The 2012 Developer Productivity Report is FREE and packed with Java stats, analysis and interviews. Inside you’ll find expanded coverage on technologies and tools used by Java development teams along with analysis and interviews with Java experts on “What makes developers tick?”

Final words

JRebel will give you the best of both worlds—the performance, scalability and robustness of Java, as well as the FTL-turnaround and the feeling of lightness, which are typically associated with 100-line guestbook applications, not with big mature systems. JRebel behaves as if it’s in Rome—you don’t have to make changes to your code, architecture or sysops. It adapts to your environment and requirements. All you have to do is start using it. You get to keep all the effort and investments already made to your infrastructure.

↧

How to Use JRebel for Android

January 19, 2017, 10:54 pm

≫ Next: Form Submit Animation Android Example

≪ Previous: Why and How to use JRebel in Java

Android development is great so long as your project stays relatively small. As the functionality of your project grows you’ll find that your build times follow suit. This puts you in the position where you spend most of your time figuring how to make your build run faster rather than adding more value to your actual app.

You can also check : Why and How to use JRebel in Java

The interwebs are packed full with suggestions of how to squeeze the most out of your Gradle builds. There are some great posts on this, including, “Making Gradle builds faster”. Although you can win back seconds and maybe even minutes, yet some bottlenecks will still remain in your build. For example having annotation based dependency injections are nice to have for a cleaner architecture, but it has an impact on your build time.

One thing you can try is JRebel for Android. It takes a different approach by not introducing a new apk after each change. Instead apk gets installed once and delta packages are shipped over to the device or emulator and are applied during runtime. This logic is nothing new and has been present in the Java EE/SE with JRebel for more than 8 years.

Let’s take the Google IO 2015 app and see how the JRebel for Android setup works as well as how it can save you valuable time.

Installing JRebel for Android

JRebel for Android is available as a plugin for Android Studio. You can download it directly from the IDE by navigating to Plugins > Browse Repositories and searching for “JRebel for Android”.

If for some reason you can’t access the public maven repositories you can download it directly from the JetBrains homepage. After which you need to install via the Plugins > Install plugin from disk… route.

Once the plugin is installed, you’ll need to restart Android Studio, as usual after a plugin installation. After restart, you need to provide your name and email to get your free 21 day trial of JRebel for Android.

Running my application with JRebel for Android

Now the plugin is installed, you just need to click the Run with JRebel for Android button, which will always build a new apk if changes are detected between the previous installation. Run with JRebel for Android is the same as the Run action in Android Studio. So you’ll be faced with the same run flow, where you first need to pick a device and then apk is built and installed on that device etc.

To update your code and resources, JRebel for Android needs to process the project’s classes and embed an agent to the application. JRebel for Android will only run with a debuggable flavor, so your release apk is never affected. In addition no changes are required to your project. For a detailed overview of how JRebel for Android works, read this under the hood post.
So pressing Run with JRebel for Android on the Google IO 2015 application would result in the following:

Applying changes with JRebel for Android

The Apply changes button is the key when using JRebel for Android, it will do the least amount of work possible to make your changes and updates visible on your device. If you didn’t use Run with JRebel for Android to install the application yet, Apply changes will take care of the installation on your behalf.

Now let’s make a simple functional change to the application. For each session taking present in GoogleIO you can send feedback. We’ll add one additional element to the question, an input for your name and we’ll use the value from that input in a Toast thanking you for providing feedback.

Step 1: Add an EditText component to the session_feedback_fragment.xml

<FrameLayout
android:layout_width="match_parent"
android:layout_height="wrap_content">
<EditText
android:id="@+id/name_input"
android:layout_width="match_parent"
android:layout_height="wrap_content"/>
</FrameLayout>

Step 2: Fix the paddings

<FrameLayout
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:paddingLeft="@dimen/padding_normal"
android:paddingStart="@dimen/padding_normal"
android:paddingRight="@dimen/padding_normal"
android:paddingEnd="@dimen/padding_normal"
android:paddingTop="@dimen/spacing_micro"
android:paddingBottom="@dimen/padding_normal">

Step 3: Add a hint

<EditText
android:id="@+id/name_input"
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:hint="@string/name_hint"/>

During all of these changes we have remained on the same screen on our device throughout. After each Apply change Activity.recreate() is invoked by JRebel for Android. So your top most activity will go through the same callbacks as you would if you rotated your device from portrait to landscape, for example.
So far we’ve only performed resource changes, so let’s change some Java code as well.

Step 4: Show a toast in SessionFeedbackFragment.sumbitFeedback()

EditText nameInput = (EditText) getView().findViewById(R.id.name_input);
Toast.makeText(getActivity(), "Thanks for the feedback " + nameInput.getEditableText().toString(), Toast.LENGTH_SHORT).show();

Application restart vs Activity restart

Not all changes will trigger an Activity.recreate() invocation. Should you change something in the AndroidManifest, a new apk has to be built and an incremental install will be performed. In this case the application will be restarted. Application restart will also be done if you replace a superclass or change interfaces that the class is implementing. Here is a full breakdown of activity vs application restart:

Activity restart	Application restart
Adding, removing, modifying methods	Adding, removing implemented interfaces
Adding, removing, modifying constructors	Replacing superclasses
Adding, removing fields
Adding, removing classes
Adding, removing annotations
Adding, removing, modifying static fields value
Adding, removing enum values
Modifying interface
Adding, removing, modifying XML resources

Why should you try JRebel for Android?

There are loads of reasons! Here are some of the most persuasive reasons as to why you should give it a go:

Reduce the time it takes to see your changes on the device
Polishing UI and getting that pixel perfect result no longer takes hours because of long build times
No need to change anything in your project to make JRebel for Android work
Use debugger and update code and resources at the same time! That’s right JRebel for Android comes with full debugger support!

↧

Form Submit Animation Android Example

February 10, 2017, 9:40 pm

≫ Next: Kick Out the Wrong Password Android Animation

≪ Previous: How to Use JRebel for Android

Hello friends, I hope you doing well in development and my blog helped when you face any problem. Today i am sharing a cool animation tutorial for Android which you can use in any application and they boost the look of your application.

You must read this : Android Developers Get Wrong in Material Design

In old way when you make a form in Android with submit button it stores the data with boring loading bar. But after this tutorial you can make cool animation for this. Just take a look in below image :

Cool isn't it. If you liked then just dive in the code now.

See the difference :Google ProgressBarV/sFAB ProgressBar

Download Source Code

In this project we use ButterKnife for Field and method binding with Android views.

Now start with simple layout with three EditText with floating action button

<LinearLayout
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:orientation="vertical">

<EditText
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:hint="First edit text"/>

<EditText
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:hint="Second edit text"/>

<EditText
android:layout_width="match_parent"
android:layout_height="match_parent"
android:hint="Third edit text"/>
</LinearLayout>

<android.support.design.widget.FloatingActionButton
android:id="@+id/fab"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:layout_gravity="bottom|end"
android:layout_margin="@dimen/fab_margin"
android:src="@drawable/ic_send_white_24dp"/>

And now we want to add some animation in our simple layout so firstly we add all layout in FrameLayout and now our layout look likes

<FrameLayout
android:id="@+id/inputs_layout"
android:layout_width="match_parent"
android:layout_height="match_parent"
android:background="@android:color/background_light"
android:paddingBottom="@dimen/activity_vertical_margin"
android:paddingLeft="@dimen/activity_horizontal_margin"
android:paddingRight="@dimen/activity_horizontal_margin"
android:paddingTop="@dimen/activity_vertical_margin"
app:layout_behavior="@string/appbar_scrolling_view_behavior">

<LinearLayout
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:orientation="vertical">

<EditText
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:hint="First edit text"/>

<EditText
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:hint="Second edit text"/>

<EditText
android:layout_width="match_parent"
android:layout_height="match_parent"
android:hint="Third edit text"/>
</LinearLayout>

<android.support.design.widget.FloatingActionButton
android:id="@+id/fab"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:layout_gravity="bottom|end"
android:layout_margin="@dimen/fab_margin"
android:src="@drawable/ic_send_white_24dp"/>

</FrameLayout>

Now at last we use some more layouts to achieve our aim like io.codetail.widget.RevealFrameLayout and final layout xml would be

activity_main.xml

<?xml version="1.0" encoding="utf-8"?>
<android.support.design.widget.CoordinatorLayoutxmlns:android="http://schemas.android.com/apk/res/android"
xmlns:app="http://schemas.android.com/apk/res-auto"
xmlns:tools="http://schemas.android.com/tools"
android:layout_width="match_parent"
android:layout_height="match_parent"
android:fitsSystemWindows="true"
tools:context="com.jaouan.sendinganimationexample.MainActivity">

<android.support.design.widget.AppBarLayout
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:theme="@style/AppTheme.AppBarOverlay">

<android.support.v7.widget.Toolbar
android:id="@+id/toolbar"
android:layout_width="match_parent"
android:layout_height="?attr/actionBarSize"
android:background="?attr/colorPrimary"
app:popupTheme="@style/AppTheme.PopupOverlay"/>

</android.support.design.widget.AppBarLayout>

<RelativeLayout
android:id="@+id/sky_layout"
android:layout_width="match_parent"
android:layout_height="match_parent"
android:background="@color/sky">

<FrameLayout
android:id="@+id/plane_layout"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:layout_alignParentBottom="true"
android:layout_alignParentEnd="true">

<ImageView
android:id="@+id/plane"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:layout_margin="@dimen/plane_margin"
android:rotation="180"
android:src="@drawable/ic_send_white_24dp"/>
</FrameLayout>

<io.codetail.widget.RevealFrameLayout
android:layout_width="150dp"
android:layout_height="150dp"
android:layout_centerInParent="true">

<RelativeLayout
android:id="@+id/sent_layout"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:background="@drawable/rounded">

<ImageView
android:id="@+id/check"
android:layout_width="match_parent"
android:layout_height="match_parent"
android:layout_centerInParent="true"
android:padding="20dp"
android:src="@drawable/ic_check_white_24dp"/>
</RelativeLayout>
</io.codetail.widget.RevealFrameLayout>

</RelativeLayout>

<FrameLayout
android:id="@+id/inputs_layout"
android:layout_width="match_parent"
android:layout_height="match_parent"
android:background="@android:color/background_light"
android:paddingBottom="@dimen/activity_vertical_margin"
android:paddingLeft="@dimen/activity_horizontal_margin"
android:paddingRight="@dimen/activity_horizontal_margin"
android:paddingTop="@dimen/activity_vertical_margin"
app:layout_behavior="@string/appbar_scrolling_view_behavior">

<LinearLayout
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:orientation="vertical">

<EditText
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:hint="First edit text"/>

<EditText
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:hint="Second edit text"/>

<EditText
android:layout_width="match_parent"
android:layout_height="match_parent"
android:hint="Third edit text"/>
</LinearLayout>

<android.support.design.widget.FloatingActionButton
android:id="@+id/fab"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:layout_gravity="bottom|end"
android:layout_margin="@dimen/fab_margin"
android:src="@drawable/ic_send_white_24dp"/>

</FrameLayout>


</android.support.design.widget.CoordinatorLayout>

Now we add animation in our layout by Activity file. Here we'll firstly click on floating button by which all the input layout hide and rotate the fab and then flying fab button animation will start with full screen cover and at last restore the input layout.

We just start with binding the ButterKnife library in onCreate function.

protectedvoidonCreate(Bundle savedInstanceState) {
super.onCreate(savedInstanceState);
        setContentView(R.layout.activity_main);
        setSupportActionBar((Toolbar) findViewById(R.id.toolbar));
        ButterKnife.bind(this);
    }

Now make function for onClick on fab button

@OnClick(R.id.fab)
voidsend() {
// - Prepare views visibility.
        mCheckImageView.setVisibility(View.INVISIBLE);
        mSentLayout.setVisibility(View.INVISIBLE);

// - Rotate fab.
final RotateAnimation rotateAnimation = new RotateAnimation(0, 180, Animation.RELATIVE_TO_SELF, .5f, Animation.RELATIVE_TO_SELF, .5f);
        rotateAnimation.setInterpolator(new AccelerateDecelerateInterpolator());
        rotateAnimation.setDuration(280);
        mFab.startAnimation(rotateAnimation);

// - Hide inputs layout.
final Animator circularReveal = ViewAnimationUtils.createCircularReveal(mInputsLayout, (int) (mFab.getX() + mFab.getWidth() / 2), (int) (mFab.getY() + mFab.getHeight() / 2), mInputsLayout.getHeight(), 0);
        circularReveal.setDuration(250);
        circularReveal.addListener(new AnimatorListenerAdapter() {
@Override
publicvoidonAnimationEnd(Animator animation) {
// - Update views visibility.
                mInputsLayout.setVisibility(View.INVISIBLE);

// - Fly away.
                flyAway();
            }
        });
        circularReveal.start();

    }

Implement fly button animation

privatevoidflyAway() {
// - Combine rotation and translation animations.
final RotateAnimation rotateAnimation = new RotateAnimation(0, 180);
        rotateAnimation.setDuration(1000);
        mPlaneImageView.startAnimation(rotateAnimation);
        Revealator.reveal(mSentLayout)
                .from(mPlaneLayout)
                .withTranslateDuration(1000)
                .withCurvedTranslation(new PointF(-1200, 0))
                .withRevealDuration(200)
                .withHideFromViewAtTranslateInterpolatedTime(.5f)
                .withEndAction(new Runnable() {
@Override
publicvoidrun() {

// - Display checked icon.
final ScaleAnimation scaleAnimation = new ScaleAnimation(0, 1, 0, 1, Animation.RELATIVE_TO_SELF, .5f, Animation.RELATIVE_TO_SELF, .5f);
                        scaleAnimation.setInterpolator(new BounceInterpolator());
                        scaleAnimation.setDuration(500);
                        scaleAnimation.setAnimationListener(new AnimationListenerAdapter() {
@Override
publicvoidonAnimationEnd(Animation animation) {
                                mInputsLayout.postDelayed(new Runnable() {
@Override
publicvoidrun() {

// - Restore inputs layout.
                                        retoreInputsLayout();

                                    }
                                }, 1000);
                            }
                        });
                        mCheckImageView.startAnimation(scaleAnimation);
                        mCheckImageView.setVisibility(View.VISIBLE);

                    }
                }).start();
    }

After animation done we just restore our input fields

privatevoidretoreInputsLayout() {
        mInputsLayout.postDelayed(new Runnable() {
@Override
publicvoidrun() {

final Animator circularReveal = ViewAnimationUtils.createCircularReveal(mInputsLayout, (int) (mFab.getX() + mFab.getWidth() / 2), (int) (mFab.getY() + mFab.getHeight() / 2), 0, mInputsLayout.getHeight());
                circularReveal.setDuration(250);
                circularReveal.start();

                mInputsLayout.setVisibility(View.VISIBLE);
            }
        }, 1000);
    }

Final Main_Activity.java

importandroid.animation.Animator;
importandroid.animation.AnimatorListenerAdapter;
importandroid.graphics.PointF;
importandroid.os.Bundle;
importandroid.support.v7.app.AppCompatActivity;
importandroid.support.v7.widget.Toolbar;
importandroid.view.Menu;
importandroid.view.View;
importandroid.view.ViewAnimationUtils;
importandroid.view.animation.AccelerateDecelerateInterpolator;
importandroid.view.animation.Animation;
importandroid.view.animation.BounceInterpolator;
importandroid.view.animation.RotateAnimation;
importandroid.view.animation.ScaleAnimation;

importcom.jaouan.revealator.Revealator;
importcom.jaouan.revealator.animations.AnimationListenerAdapter;

importbutterknife.BindView;
importbutterknife.ButterKnife;
importbutterknife.OnClick;

publicclassMainActivityextends AppCompatActivity {

@BindView(R.id.fab)
    View mFab;

@BindView(R.id.plane)
    View mPlaneImageView;

@BindView(R.id.plane_layout)
    View mPlaneLayout;

@BindView(R.id.inputs_layout)
    View mInputsLayout;

@BindView(R.id.sky_layout)
    View mSkyLayout;

@BindView(R.id.sent_layout)
    View mSentLayout;

@BindView(R.id.check)
    View mCheckImageView;

@Override
protectedvoidonCreate(Bundle savedInstanceState) {
super.onCreate(savedInstanceState);
        setContentView(R.layout.activity_main);
        setSupportActionBar((Toolbar) findViewById(R.id.toolbar));
        ButterKnife.bind(this);
    }

/**
     * Send something.
     */
@OnClick(R.id.fab)
voidsend() {
// - Prepare views visibility.
        mCheckImageView.setVisibility(View.INVISIBLE);
        mSentLayout.setVisibility(View.INVISIBLE);

// - Rotate fab.
final RotateAnimation rotateAnimation = new RotateAnimation(0, 180, Animation.RELATIVE_TO_SELF, .5f, Animation.RELATIVE_TO_SELF, .5f);
        rotateAnimation.setInterpolator(new AccelerateDecelerateInterpolator());
        rotateAnimation.setDuration(280);
        mFab.startAnimation(rotateAnimation);

// - Hide inputs layout.
final Animator circularReveal = ViewAnimationUtils.createCircularReveal(mInputsLayout, (int) (mFab.getX() + mFab.getWidth() / 2), (int) (mFab.getY() + mFab.getHeight() / 2), mInputsLayout.getHeight(), 0);
        circularReveal.setDuration(250);
        circularReveal.addListener(new AnimatorListenerAdapter() {
@Override
publicvoidonAnimationEnd(Animator animation) {
// - Update views visibility.
                mInputsLayout.setVisibility(View.INVISIBLE);

// - Fly away.
                flyAway();
            }
        });
        circularReveal.start();

    }

/**
     * Starts fly animation.
     */
privatevoidflyAway() {
// - Combine rotation and translation animations.
final RotateAnimation rotateAnimation = new RotateAnimation(0, 180);
        rotateAnimation.setDuration(1000);
        mPlaneImageView.startAnimation(rotateAnimation);
        Revealator.reveal(mSentLayout)
                .from(mPlaneLayout)
                .withTranslateDuration(1000)
                .withCurvedTranslation(new PointF(-1200, 0))
                .withRevealDuration(200)
                .withHideFromViewAtTranslateInterpolatedTime(.5f)
                .withEndAction(new Runnable() {
@Override
publicvoidrun() {

// - Display checked icon.
final ScaleAnimation scaleAnimation = new ScaleAnimation(0, 1, 0, 1, Animation.RELATIVE_TO_SELF, .5f, Animation.RELATIVE_TO_SELF, .5f);
                        scaleAnimation.setInterpolator(new BounceInterpolator());
                        scaleAnimation.setDuration(500);
                        scaleAnimation.setAnimationListener(new AnimationListenerAdapter() {
@Override
publicvoidonAnimationEnd(Animation animation) {
                                mInputsLayout.postDelayed(new Runnable() {
@Override
publicvoidrun() {

// - Restore inputs layout.
                                        retoreInputsLayout();

                                    }
                                }, 1000);
                            }
                        });
                        mCheckImageView.startAnimation(scaleAnimation);
                        mCheckImageView.setVisibility(View.VISIBLE);

                    }
                }).start();
    }

/**
     * Restores inputs layout.
     */
privatevoidretoreInputsLayout() {
        mInputsLayout.postDelayed(new Runnable() {
@Override
publicvoidrun() {

final Animator circularReveal = ViewAnimationUtils.createCircularReveal(mInputsLayout, (int) (mFab.getX() + mFab.getWidth() / 2), (int) (mFab.getY() + mFab.getHeight() / 2), 0, mInputsLayout.getHeight());
                circularReveal.setDuration(250);
                circularReveal.start();

                mInputsLayout.setVisibility(View.VISIBLE);
            }
        }, 1000);
    }

@Override
publicbooleanonCreateOptionsMenu(Menu menu) {
        getMenuInflater().inflate(R.menu.menu_main, menu);
returntrue;
    }
}

Thats it. Hope you like this tutorial and if you've any issues or problem or suggestion then please comment in below and like my Facebook page.

↧

Kick Out the Wrong Password Android Animation

February 12, 2017, 10:24 pm

≫ Next: Top 25 Android Libraries you can try in 2017 - Part 3

≪ Previous: Form Submit Animation Android Example

Hello friends now a day many of android applications are making and the successful measure of popular application is their UI, Clean up and animation effect too. In my previous article i show you how to use beautiful form submit animation. Just look at that tutorial.

In this tutorial you are gonna see kick out the wrong password animation for Android application. Most of android application need authentication and find wrong password if any intruder want to access.

For fast android development : How to Use JRebel for Android

We use that form but not in traditional way just look at below demo and tell me about it.

Cool isn't it...
If you think it is cool and want to implement this then dive in the code now.

What we do in our animation ?

Firstly it showing normal form with CircularImageView and single EditText.
When you entered wrong password and click on submit button then it lock image move and kick the wrong password and remain blank the password and fadeout a red message "Wrong password" and after some time it gone and EditText is blank again.

Play and edit with Full source code : HungerMoji - Android Notification Game

In layout here we create password layout(layout_password.xml) separately from normal form for avoid confusion and then include that layout in our main layout(activity_main.xml).

layout_password.xml

<?xml version="1.0" encoding="utf-8"?>
<LinearLayoutxmlns:android="http://schemas.android.com/apk/res/android"
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:layout_marginTop="20dp"
android:background="@drawable/bg_edit_text"
android:orientation="horizontal">


<ImageView
android:id="@+id/ivLock"
android:layout_width="wrap_content"
android:layout_height="match_parent"
android:padding="10dp"
android:src="@drawable/ic_locked_padlock"/>

<FrameLayout
android:layout_width="match_parent"
android:layout_height="wrap_content">

<EditText
android:id="@+id/etPassword"
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:background="@drawable/bg_edit_text"
android:inputType="textPassword"
android:paddingBottom="5dp"
android:paddingLeft="10dp"
android:paddingTop="5dp"
android:text="Hello"
android:textSize="30sp"/>

<LinearLayout
android:id="@+id/llParent"
android:layout_width="match_parent"
android:layout_height="wrap_content"
android:layout_marginTop="15dp"
android:gravity="center"
android:orientation="horizontal"
android:paddingBottom="5dp"
android:paddingLeft="10dp"
android:paddingTop="5dp">

<ImageView
android:id="@+id/ivDummy"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:src="@drawable/ic_dot"
android:visibility="gone"/>

<EditText
android:id="@+id/etDummy"
android:layout_width="0dp"
android:layout_height="0dp"
android:focusable="true"
android:focusableInTouchMode="true"></EditText>
</LinearLayout>
</FrameLayout>
</LinearLayout>

What we done above in layout ?

Line 10-15 : It is a padlock ImageView which will move and take a role to kick out the wrong password from EditText.
Line 33-57 : This linear layout will holds all dots generated during animation. Padding, margin of this linear layout must be in match with Edit text.
Line 44-49 : This is just to simulate position of dots. Visibility of this image view must be in GONE state.
Line 51-56 : This is used to remove focus of password edittext.

Now we implement animation in our layout from activity.
Main_Activity.java

importandroid.content.Context;
importandroid.os.Bundle;
importandroid.support.v7.app.AppCompatActivity;
importandroid.view.View;
importandroid.view.animation.Animation;
importandroid.view.animation.AnticipateInterpolator;
importandroid.view.animation.CycleInterpolator;
importandroid.view.animation.RotateAnimation;
importandroid.view.animation.TranslateAnimation;
importandroid.view.inputmethod.InputMethodManager;
importandroid.widget.Button;
importandroid.widget.EditText;
importandroid.widget.ImageView;
importandroid.widget.LinearLayout;



publicclassMainActivityextends AppCompatActivity {

    EditText etPassword, etDummy;
    Button btnLogin;
private LinearLayout llParent;
private ImageView ivLock;

@Override
protectedvoidonCreate(Bundle savedInstanceState) {
super.onCreate(savedInstanceState);

        setContentView(R.layout.activity_main);
        etPassword = (EditText) findViewById(R.id.etPassword);
        etDummy = (EditText) findViewById(R.id.etDummy);
        ivLock = (ImageView) findViewById(R.id.ivLock);
        llParent = (LinearLayout) findViewById(R.id.llParent);
        btnLogin = (Button) findViewById(R.id.btnLogin);

        btnLogin.setOnClickListener(new View.OnClickListener() {
@Override
publicvoidonClick(View view) {
//  Hide cursor/focus from edit text after click
                removeFocus();

                String password = etPassword.getText().toString();
                etPassword.setText("");

                validatePassword(password);
            }
        });
    }

privatevoidvalidatePassword(String password) {
//  TODO Perform password validation here and if failed then proceed with below

int passLength = password.length();
        playHangingAnimation(ivLock);

//  Generate number of dots (imageview) similar to length of input password
for (int i = 0; i < passLength; i++) {
            ImageView imageView = new ImageView(this);
            imageView.setImageResource(R.drawable.ic_dot);
            LinearLayout.LayoutParams params = new LinearLayout.LayoutParams(LinearLayout.LayoutParams.WRAP_CONTENT,
                    LinearLayout.LayoutParams.WRAP_CONTENT);
            params.setMargins(3, 0, 3, 0);
            imageView.setLayoutParams(params);
            llParent.addView(imageView);

//  Play animation for each and every dots separately
            playKickOutAnimation(imageView, i);
        }
    }

privatevoidplayKickOutAnimation(final View view, int i) {
//  Parameters are startPositionX,endPositionX, startPositionY, endPositionY
//  Intentionally changed position of X at runtime based on i(index of dot) to give elastic effect
//  Play around these numbers and let community know if any combination is giving better result :)
        Animation animation = new TranslateAnimation(-20 + i * 5, 400, 0, 0);

//  Intentionally changed duration at runtime based on i(index of dot) to give elastic effect
        animation.setDuration(700 + i * 20);

//  To give kick out effect. Read about all Inter polator here - http://cogitolearning.co.uk/?p=1078
        animation.setInterpolator(new AnticipateInterpolator());

        animation.setAnimationListener(new Animation.AnimationListener() {
@Override
publicvoidonAnimationStart(Animation animation) {

            }

@Override
publicvoidonAnimationEnd(Animation animation) {
//  Remove dots from screen once animation has been stopped
                llParent.removeView(view);
            }

@Override
publicvoidonAnimationRepeat(Animation animation) {

            }
        });
        view.startAnimation(animation);
    }

privatevoidplayHangingAnimation(View v) {
int pivot = Animation.RELATIVE_TO_SELF;
//  Parameter defines how many times the cycle should happen
        CycleInterpolator interpolator = new CycleInterpolator(2.5f);
//  Parameters are fromDegree,toDegree,positionXType(from self in this case),positionX,positionYType,positionY
//  Play around these values to get to know the behaviour more closely
        RotateAnimation animation = new RotateAnimation(0, 20, pivot, 0.47f, pivot, 0.05f);
        animation.setStartOffset(100);
        animation.setDuration(900);
        animation.setRepeatCount(0);// Animation.INFINITE
        animation.setInterpolator(interpolator);
        v.startAnimation(animation);
    }

privatevoidremoveFocus() {
        etPassword.clearFocus();
        etDummy.requestFocus();
        InputMethodManager imm = (InputMethodManager) getSystemService(Context.INPUT_METHOD_SERVICE);
        imm.hideSoftInputFromWindow(etDummy.getApplicationWindowToken(), 0);
    }
}

What we do in activity above ?

Firstly we got value of EditText and call validatePassword(password) function in onCreate from line 36-48 and definition of that is start from line 50-69 in which we perform password validation here and if failed then proceed with below and call playHangingAnimation(ivLock) whose play a role to animate padlock from line 103-115.
Generate number of dots (imageview) similar to length of input password and then play animation for each and every dots separately by calling playKickOutAnimation(imageView, i) from line 71-101.
And finally we removeFocus from Edit text from line 117-122.

Download Source code

Hope you like this tutorial please comment below and like my Facebook page.

↧