OReilly web database applications with PHP and MySQL ISBN mar 2002 0596000413

A join query is a querying technique that matches rows from two or more tables based on a join condition in a WHERE clause and outputs onlythose rows that meet the condition.. So, a real

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

only for RuBoard - do not distribute or recompile

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

authenticating

managing

validation script for only for RuBoard - do not distribute or recompile

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

only for RuBoard - do not distribute or recompile

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

system( )

only for RuBoard - do not distribute or recompile

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

X

XTemplate class library

only for RuBoard - do not distribute or recompile

[Symbol][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Z]

Z

Zend engine , 2nd

only for RuBoard - do not distribute or recompile

A join query is a querying technique that matches rows from two or more

tables based on a join condition in a WHERE clause and outputs onlythose rows that meet the condition As part of the process of convertingthe winestore entity-relationship model to SQL statements, we have

SELECT region_id FROM winery

WHERE winery_name='Ryan Ridge Winery';

Recall that the ORDER BY clause sorts the results after the query hasbeen evaluated; it has no effect on which rows are returned from thequery Here is the first part of the result of the query with the ORDER BY

3.7.2 Elementary Natural Joins

A cartesian product isn't the join we want Instead, we want to limit theresults to only the sensible rows, where the winery is actually located inthe region From a database perspective, we want only rows in which the

region_id in the winery table matches the corresponding region_id

in the region table This is a natural join [1]

[1] It isn't quite true to say that the joins described here are naturaljoins A true natural join doesn't require you to specify the join

condition, because "natural" implies that the system figures this outitself So, a real natural join doesn't need the WHERE clause; one isautomatically included "behind the scenes." The joins described

example, SELECT retrieves rows from the tables winery and region.

Attributes in the WHERE clause are specified using both the tablename and attribute name, separated by a period This usually

disambiguates uses of the same attribute name in different tables

So, for example, region_id in the region table and region_id in

the winery table are disambiguated as region.region_id and

winery.region_id This procedure can also be used for clarity inqueries, even if it isn't required It can be used in all parts of the

query, not just the WHERE clause

The WHERE clause includes a join clause that matches rows betweenthe multiple tables In this example, the output is reduced to thoserows where wineries and regions have matching region_id

attributes, resulting in a list of all wineries and which region they arelocated in This is the key to joining two or more tables to producesensible results

3.7.2.1 Examples

The natural join can be used in many other examples in the winestore.Consider another example that finds all the wines made by all the

You can extend this query to produce a list of wines made by a specificwinery or group of wineries To find all wines made by wineries with aname beginning with Borg, use:

To find the name of the region Ryan Ridge Winery is situated in:

SELECT region.region_name FROM region,winery

WHERE winery.region_id=region.region_id AND winery.winery_name='Ryan Ridge Winery';

with cust_id=1 in table c2

3.7.3 Joins with More than Two Tables

Queries can join more than two tables In the next example, the queryfinds all details of each item from each order by a particular customer,customer #2 The example also illustrates how frequently the Booleanoperators AND and OR are used:

same in the orders and items tables.

If you remove the cust_id=2 clause, the query outputs all items in allorders by all customers This is a large result set, but still a sensible onethat is much smaller than the cartesian product!

SELECT region_name FROM wine,winery,region

WHERE wine.winery_id=winery.winery_id AND

winery.region_id=region.region_id AND

1 The WHERE clause restricts the winery rows to those that bear the

Buonopane Wines DISTINCT is used to show each customer only

once ORDER BY sorts the customer rows into telephone directory

order

Here is the example query that joins all three tables To find what grapevarieties are in wine #1004, use:

SELECT variety FROM grape_variety, wine_variety, wine WHERE wine.wine_id=wine_variety.wine_id AND

wine_variety.variety_id=grape_variety.variety_id AND wine.wine_id=1004

We've now covered as much complex querying in SQL as we need to inthis chapter If you'd like to learn more, see the pointers to resourcesincluded in Appendix E SQL examples in web database applicationscan be found throughout Chapter 4 to Chapter 13.

In this section, we show how to engineer a front-page panelwe call this

the panel The completed panel was shown in Figure 4-1 We use the

techniques discussed so far in this chapter to present more attractiveHTML <table> formatted results, to process multiple query results, and

to customize the output based on the data retrieved No significant newconcepts are introduced in the case study

The panel case study is a progressive development of a script to displaythe details of new wines We show the following details in the panel:

Information about the three wines most recently added to the

database, including the vintage year, the winery, the wine name, andthe varieties

The review written by a wine writer

How much a bottle costs, how much a case of a dozen bottles costs,and any per-bottle discount users receive if they purchase a case

To achieve the outcome of a functional and attractive panel, you need to

query the wine, winery, inventory, grape_variety, and wine_variety tables.

You also need to use the structure of the HTML <table> environment toachieve distinct presentation of the three componentsthe details, thereview, and the priceof each newly added wine Last, you need somemathematics to calculate any savings for buying a case and present

these savings to the user

The panel component developed in this chapter is the basis of the frontpage of our online winestore However, shopping cart features that arenot discussed in detail here have been added to the production versionshown in Figure 4-1 The finalized code that includes the shopping-cartfunctionality is discussed further in Chapter 5, and the completed code

is presented in Chapter 11

In engineering the panel, we use the following techniques:

components of the system Moreover, PHP is good at reporting errors tothe browser, and the HTML output can easily be viewed In most

browsers, right-clicking on the HTML document in the web browser

window offers the option to view the HTML source

4.3.1 Step 1: Producing Visually Appealing Tables

Example 4-9 shows a script that is the first step in producing the panel.Not surprisingly, the script combines the same querying process

described earlier with an HTML <table> environment to wrap the

output The output is more attractive than in previous examples and theoutput in a Netscape browser is shown in Figure 4-3

Figure 4-3 The first step in producing a front-page panel that shows more attractive

presentation

The query returns one row for each inventory of a wine If a wine hasmultiple inventories, the wine appears multiple times The query alsooutputs the wine's winery_name, the vintage attribute year, the

wine_name, and a descriptive review, description The WHERE clauseensures that only reviewed winesthose with a description that isn'temptyare returned The WHERE clause also implements a natural join with

the wine table using the primary keys of the winery and inventory tables.

date_added added inventory, and the LIMIT 3 ensures only the three latest-addedinventories are retrieved

if (!($result = @ mysql_query ($query, $connection))) showerror( );

Different background colors for the single <td> element are set ineach table row <tr>; the colors are maroon, silver, and gray

The color attribute of the <font> tag is set to white for the heading

of each wine

The bold tag <b> is used for pricing information

A blank row between wines is used for spacing in the presentation

mysql_fetch_array( ) is used to retrieve rows This has theadvantage that the elements of the $row array can be referenced byattribute name The resultant code is more readable and more

query-independent than if mysql_fetch_row( ) is used

Manipulating presentation by using structure is, unfortunately, part ofworking with HTML

4.3.1.1 Limitations of Step 1

introduction to the case study Three particular limitations are:

The varieties of the wines are not shown For example, you can't tellthat the first-listed Binns Hill Vineyard Morfooney is a Cabernet

Sauvignon variety

The user expects that the dozen price represents a per-bottle savingover purchasing bottles in smaller quantities However, the frontpanel doesn't show the saving, and the user needs a calculator todecide whether a dozen bottles is worth the discount

The first-listed wine appears twice There are two inventory entries for the same wine, and the query has returned two rows for that same wine, with the only difference being the prices.

Another explanation for a double appearance could be that there are

two wines with the same review and year, but with different

grape_varieties This is very unlikely and isn't the case here.

We improve the panel progressively in the next section to address theselimitations, while also adding new features

4.3.2 Step 2: Adding Varieties to the Panel

To add varieties to the panel, you need two SQL queries in a single

script This next step adds an additional query to find the varieties of awine, and the consolidated varieties are presented together with the

vintage, winery, and wine name

The second addition to the panel in this step is the calculation and

conditional display of results We introduce a new feature to the panelthat calculates the savings in buying a dozen bottles and shows the userthe per-bottle saving of buying a case of wine, but only when there issuch a saving We don't deal with the situation where a case costs morethan 12 single purchases

The script showing these two new concepts is in Example 4-10 Thescript improves on Example 4-9 by removing the first two limitations

Figure 4-4 Adding wine varieties and discounts to the panel

Example 4-10 An improved display with varieties and the dozen-bottle discount

<!DOCTYPE HTML PUBLIC

"-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">

if (!mysql_select_db($databaseName, $connection))

showerror( );

// Run the query created above on the database through // the connection

if (!($result = @ mysql_query ($query, $connection))) showerror( );

winery_name, the inventory data (cost and case_cost), and the

the information (all the data from wine, winery, and inventory) The

second query is nested inside the first; that is, for each row retrieved fromthe result set of the first query, you run the new query to get the varieties.The result is that the script runs four queries: one to retrieve the threewines, and three queries to get their varieties

Let's return to Example 4-10 The first query has not changed and stillreturns one row per inventory of each of the most recently added winesthat has a written review For each wine, the script produces a headingshowing the year, winery_name, and wine_name

It is after this query is run and the year, winery_name, and wine_name

output that the new functionality of an additional query begins In thisexample, a function, showVarieties( ), is called This function runs aquery to find the varieties of a particular wine with a wine_id value thatmatches the parameter $wineID:

winestore; the approach is used in the panel to produce order receipts forpresentation and email confirmation, and in many of the stock and

Why isn't such data stored? The answer is usually that it is redundant:storing it adds no more information to the database The down side isthat you need calculations to recreate output when it is needed In thissection, this is illustrated with a simple example that shows the per-bottlesaving when a user purchases more than a dozen bottles

Returning to the script in Example 4-10, having produced a completeheading that now includes the wine variety, we produce the wine review

The element $row["cost"] is the cost of a single bottle, and

$row["case_cost"] is the cost of a case Since a case contains 12bottles, it follows that the cost of 1 bottle in the case is

$row["case_cost"]/12 The difference between the price of a singlebottle and the price of bottle that comes in a case is then:

$row["cost"]-($row["case_cost"]/12)

A saving is printed out only if there is one; that is, when $dozen_saving

is greater than zero In the case where buying a dozen bottles at oncecosts the same as 12 separate purchases (or maybe more!), nothing isshown printf is used in preference to echo, so that you can includethe formatting string %-.2f to show exactly two decimal places (that is,the cents of the $dozen_saving)

There are many examples of calculations that are performed on the rawdata from the database to present information to the user in our

winestore These include calculating order totals, discounts, receipt

information, delivery charges, and so on Elementary mathematics is acommon component of most web database applications; it's used

throughout later examples

4.3.3 Step 3: Finishing the Panel

We have built a satisfactory component However, one problem identifiedearlier still remains The first-listed wine appears twice In this case it isbecause there are two inventory entries for the same wine, with the onlydifference being the prices Of course, our user will pick the cheapest

4.3.3.1 Fixing the queries

To address the inventory problemwhere a wine appears multiple times inthe front panel if there are multiple inventories of that wineyou need tomodify the initial query