MacLochlainns Weblog

This is probably defensive because I’ve had to answer the question about two dozen times since the iPad product announcement. The question is naturally, what do yo think about the iPad?

My perspective is biased by the fact that I’ve been using both DOS/Windows PCs and Macs since the 1980s. They each have merits but in short, unlike the media, I have a bias toward Apple products. In fact, I’m an old NeXT system administrator (software gone from the scene because as rumors have it, Steve wouldn’t think of letting the company become ONLY a software company).

I think the idea of the iPad for eBooks is awesome, the features are terrific. It clearly is a better opportunity for my digital movies but a bit awkward because of its size.

I can’t travel with an iPad by itself because it doesn’t support Microsoft Excel, Word, or Visio. That means I’d have to have my MacBook Pro and iPad. Ouch, the security folks will go nuts at the airport, and my bags are now heavier by about 2 pounds. The iPad is 1.5 pounds but the charger has weight too.

I understand all the logic for the device but there’s an underlying assumption in placing everything on the web. Some data can’t be on the web because of legal limits. This goes to my sticking point. Apple’s Office Suite isn’t as robust as Microsoft’s Office Suite. Keynote is easier to use and easily preferred over PowerPoint, but Numbers isn’t even close to Excel (here Apple fails). The problem with Pages is that many companies have templates built around Word and there’s no easy migration back and forth.

Perhaps Apple will reach out to Power Excel users and invest in Numbers to bring it into this millennium; and maybe they’ll also fix the portability between Word and Pages too. For example, one company I work with insists that I use Word 2003 because they’ve never updated their templates to Word 2007 (easy to do through VMWare Fusion). Then, all that’s needed is a rock solid replacement for Visio on Mac OS X.

I think that I might buy one to experiment with, just so I’m current with the product and new features. I’ve also got some product ideas that I’d like to explore but I don’t think this is a home-run like the iPod and iPhone without vitualization software to enable Windows. As an afterthought, maybe the announcement this summer will be “you can have it all now” when they port most features to the core OS X operating system. That would induce me to upgrade my MacBook Pro, wouldn’t it get you to do so too?

A few of my students wanted me to post sample cursor loops, so here are examples of simple, FOR, and WHILE loops. There are a couple variations on approaches that demonstrate %TYPE and %ROWTYPE anchoring to data dictionary table definitions and local cursors.

Part of the idea behind these examples is to show the basic structure while mimicking the \G option of MySQL. The \G (Go) displays results as a list of column names and values by row. Ever since I discovered that in MySQL, I’ve hoped Oracle would incorporate something similar in their product. While discussing my wish list, I’d also like Oracle to make the FROM dual optional (like MySQL does) when selecting a string or numeric literal. You can find an implementation here, that leverages an example from Tom Kyte.

You can click any of the titles to view the code, which isn’t needed when you don’t have JavaScript enabled or the RSS expands them for you.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

SET SERVEROUTPUT ON SIZE 1000000
 
DECLARE
 
  -- Declare local variables that are anchored to column data types.
  lv_title     item.item_title%TYPE;
  lv_subtitle  item.item_subtitle%TYPE;
  lv_rating    item.item_rating%TYPE;
 
  -- Declare a static cursor.
  CURSOR c IS
    SELECT   i.item_title AS title
    ,        i.item_subtitle AS subtitle
    ,        i.item_rating AS rating
    FROM     item i;
 
BEGIN
  -- Open the cursor.
  OPEN c;
 
  -- Print a starting line.
  dbms_output.put_line('----------------------------------------------------------------------');
 
  -- Start the simple loop block.
  LOOP
 
    -- Fetch a row of the cursor and assign it to the three local variables.
    FETCH c
    INTO  lv_title
    ,     lv_subtitle
    ,     lv_rating;
 
    -- Exit when there aren't any more records in the cursor, without this you loop infinitely.
    EXIT WHEN c%NOTFOUND;
 
    -- Print the local variables on a single line each to mimic MySQL \G equivalent.
    dbms_output.put_line('ITEM.ITEM_TITLE    ['||lv_title||']');
    dbms_output.put_line('ITEM.ITEM_SUBTITLE ['||lv_subtitle||']');
    dbms_output.put_line('ITEM.ITEM_RATING   ['||lv_rating||']');
 
    -- Print an ending line.
   dbms_output.put_line('----------------------------------------------------------------------');
 
  END LOOP;
 
  -- Close the cursor and release the resources.
  CLOSE c;
END;
/

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

SET SERVEROUTPUT ON SIZE 1000000
 
DECLARE
 
  -- Declare a local record data type, with explicit data types (you could use %TYPE here too).
  TYPE title_type IS RECORD
  ( title     VARCHAR2(60)
  , subtitle  VARCHAR2(60)
  , rating    VARCHAR2(8));
 
  -- Declare a local variable of the local record structure data type.
  item_record TITLE_TYPE;
 
  -- Declare a static cursor.
  CURSOR c IS
    SELECT   i.item_title AS title
    ,        i.item_subtitle AS subtitle
    ,        i.item_rating AS rating
    FROM     item i;
 
BEGIN
 
  -- Open the cursor.
  OPEN c;
 
  -- Print a starting line.
  dbms_output.put_line('----------------------------------------------------------------------');
 
  -- Start the simple loop block.
  LOOP
 
    -- Fetch a row of the cursor and assign it to the local record structure variable.
    FETCH c
    INTO  item_record;
 
    -- Exit when there aren't any more records in the cursor, without this you loop infinitely.
    EXIT WHEN c%NOTFOUND;
 
    -- Print the local variable elements on a single line each to mimic MySQL \G equivalent.
    dbms_output.put_line('ITEM.ITEM_TITLE    ['||item_record.title||']');
    dbms_output.put_line('ITEM.ITEM_SUBTITLE ['||item_record.subtitle||']');
    dbms_output.put_line('ITEM.ITEM_RATING   ['||item_record.rating||']');
 
    -- Print an ending line.
   dbms_output.put_line('----------------------------------------------------------------------');
 
  END LOOP;
 
  -- Close the cursor and release the resources.
  CLOSE c;
END;
/

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46

SET SERVEROUTPUT ON SIZE 1000000
 
DECLARE
 
  -- Declare a static cursor.
  CURSOR c IS
    SELECT   i.item_title AS title
    ,        i.item_subtitle AS subtitle
    ,        i.item_rating AS rating
    FROM     item i;
 
  -- Declare a local variable of that inherits its structure from a local cursor.
  item_record c%ROWTYPE;
 
BEGIN
 
  -- Open the cursor.
  OPEN c;
 
  -- Print a starting line.
  dbms_output.put_line('----------------------------------------------------------------------');
 
  -- Start the simple loop block.
  LOOP
 
    -- Fetch a row of the cursor and assign it to the local record structure variable.
    FETCH c
    INTO  item_record;
 
    -- Exit when there aren't any more records in the cursor, without this you loop infinitely.
    EXIT WHEN c%NOTFOUND;
 
    -- Print the local variable elements on a single line each to mimic MySQL \G equivalent.
    dbms_output.put_line('ITEM.ITEM_TITLE    ['||item_record.title||']');
    dbms_output.put_line('ITEM.ITEM_SUBTITLE ['||item_record.subtitle||']');
    dbms_output.put_line('ITEM.ITEM_RATING   ['||item_record.rating||']');
 
    -- Print an ending line.
   dbms_output.put_line('----------------------------------------------------------------------');
 
  END LOOP;
 
  -- Close the cursor and release the resources.
  CLOSE c;
END;
/

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

SET SERVEROUTPUT ON SIZE 1000000
 
DECLARE
 
  -- Declare a static cursor.
  CURSOR c IS
    SELECT   i.item_title AS title
    ,        i.item_subtitle AS subtitle
    ,        i.item_rating AS rating
    FROM     item i;
 
BEGIN
 
  -- Print a starting line.
  dbms_output.put_line('----------------------------------------------------------------------');
 
  -- Start a cursor FOR loop block.
  FOR i IN c LOOP
 
    -- Print the local variable elements on a single line each to mimic MySQL \G equivalent.
    dbms_output.put_line('ITEM.ITEM_TITLE    ['||i.title||']');
    dbms_output.put_line('ITEM.ITEM_SUBTITLE ['||i.subtitle||']');
    dbms_output.put_line('ITEM.ITEM_RATING   ['||i.rating||']');
 
    -- Print an ending line.
   dbms_output.put_line('----------------------------------------------------------------------');
 
  END LOOP;
END;
/

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

SET SERVEROUTPUT ON SIZE 1000000
 
BEGIN
 
  -- Print a starting line.
  dbms_output.put_line('----------------------------------------------------------------------');
 
  -- Start a cursor FOR loop block with the static cursor in the definition.
  FOR i IN (SELECT   i.item_title AS title
	        ,        i.item_subtitle AS subtitle
	        ,        i.item_rating AS rating
	        FROM     item i) LOOP
 
    -- Print the local variable elements on a single line each to mimic MySQL \G equivalent.
    dbms_output.put_line('ITEM.ITEM_TITLE    ['||i.title||']');
    dbms_output.put_line('ITEM.ITEM_SUBTITLE ['||i.subtitle||']');
    dbms_output.put_line('ITEM.ITEM_RATING   ['||i.rating||']');
 
    -- Print an ending line.
   dbms_output.put_line('----------------------------------------------------------------------');
 
  END LOOP;
END;
/

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44

DECLARE
 
  -- Declare a static cursor.
  CURSOR c IS
    SELECT   i.item_title AS title
    ,        i.item_subtitle AS subtitle
    ,        i.item_rating AS rating
    FROM     item i;
 
  -- Declare a local variable of that inherits its structure from a local cursor.
  item_record c%ROWTYPE;
 
BEGIN
 
  -- Open the cursor.
  OPEN c;
 
  -- Print a starting line.
  dbms_output.put_line('----------------------------------------------------------------------');
 
  -- Fetch the first record to put into context the cursor attributes, like %FOUND.
  FETCH c INTO item_record;
 
  -- Start the simple loop block with a guard on entry condition.
  WHILE (c%FOUND) LOOP
 
    -- Print the local variable elements on a single line each to mimic MySQL \G equivalent.
    dbms_output.put_line('ITEM.ITEM_TITLE    ['||item_record.title||']');
    dbms_output.put_line('ITEM.ITEM_SUBTITLE ['||item_record.subtitle||']');
    dbms_output.put_line('ITEM.ITEM_RATING   ['||item_record.rating||']');
 
    -- Print an ending line.
    dbms_output.put_line('----------------------------------------------------------------------');
 
    -- Fetch the second and subsequent rows of the cursor and assign it to a local variables.
    FETCH c
    INTO  item_record;
 
  END LOOP;
 
  -- Close the cursor and release the resources.
  CLOSE c;
END;
/

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43

DECLARE
 
  -- Declare a static cursor.
  CURSOR c IS
    SELECT   i.item_title AS title
    ,        i.item_subtitle AS subtitle
    ,        i.item_rating AS rating
    FROM     item i;
 
  -- Declare a local variable of that inherits its structure from a local cursor.
  item_record c%ROWTYPE;
 
BEGIN
 
  -- Open the cursor.
  OPEN c;
 
  -- Print a starting line.
  dbms_output.put_line('----------------------------------------------------------------------');
 
  -- Start the simple loop block with a guard on entry condition.
  WHILE (TRUE) LOOP
 
    -- Fetch the record set into a user-defined variable.
    FETCH c INTO item_record;
 
    -- Exit when there aren't any more records in the cursor, without this you loop infinitely.
    EXIT WHEN c%NOTFOUND;
 
    -- Print the local variable elements on a single line each to mimic MySQL \G equivalent.
    dbms_output.put_line('ITEM.ITEM_TITLE    ['||item_record.title||']');
    dbms_output.put_line('ITEM.ITEM_SUBTITLE ['||item_record.subtitle||']');
    dbms_output.put_line('ITEM.ITEM_RATING   ['||item_record.rating||']');
 
    -- Print an ending line.
    dbms_output.put_line('----------------------------------------------------------------------');
 
  END LOOP;
 
  -- Close the cursor and release the resources.
  CLOSE c;
END;
/

I’m sure this will help my students and hope it helps somebody else.

I was working on a code example for my database class, got in a hurry, and changed a table name without dropping the original table. Oops!

Naturally, I got this error message.

  LOB (administrator_photo) STORE AS admin_photo
                                     *
ERROR at line 5:
ORA-00955: name IS already used BY an existing object

The LOB segment name existed but why and where. It was in the table that I forgot to drop. This query find the latent table and column while illustrating the relationship (for my students and others) between an OBJECT_NAME and SEGMENT_NAME:

SELECT TABLE_NAME, column_name
FROM   user_lobs
WHERE  segment_name = (SELECT object_name
                       FROM   user_objects
                       WHERE  object_name = UPPER('&object_name')
                       AND    object_type = 'LOB');

Now, I can grab it later because it’ll happen again. 😉 At Samy mentioned in his comment you also have the option of using ALL_ or DBA_ views when you’re a DBA.

Alas, poor ereg() I abused you well. PHP 5.3 has deprecated ereg() and now we must move forward with preg_match(). Along with that change, somebody asked me to show how to upload images to the file system as opposed to the database. Personally, I think they should be stored in the database.

With my bias toward databases, I threw in a virtual directory mapping in a MySQL database because it doesn’t natively support an Oracle equivalent BFILE data type. You can see this older post how to use the DBA_DIRECTORIES view in Oracle to mimic this behavior.

Naturally, MySQL is the preferred database of the person asking the question. You could also implement this exactly the same in Oracle but you really don’t want to do so. Using Oracle’s virtual directories has it’s own pre-built set of security features. They provide a more robust solution.

The code is presented as follows (setup for MySQL instructions):

1. Create and seed the DIRECTORY table in MySQL:

-- Create a directory table.
CREATE TABLE directory
( directory_id   INT PRIMARY KEY AUTO_INCREMENT
, virtual_name   VARCHAR(30)
, directory_name VARCHAR(60));
 
-- Seed the table with a virtual directory mapping.
INSERT INTO directory VALUES ( NULL,'CMS_IMAGES','C:\\Data' );

2. Create a MySQLCredentials.inc credentails file for inclusion in the PHP program:

1
2
3
4
5
6
7

<?php
  // Connection variables.
  define('HOSTNAME',"localhost");
  define('USERNAME',"student");
  define('PASSWORD',"student");
  define('DATABASE',"sampledb");
?>

3. Create the PHP uploading program, named MySQLFileUpload.php:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98

<?php
  // Set database credentials.
  include_once("MySQLCredentials.inc");
 
  // Declare input variables.
  $id = (isset($_POST['id'])) ? $_POST['id'] : 1021;
 
  // Upload a file to server in a mapped physical drive location.
  if (process_uploaded_file(map_virtual_directory($id)))
    print "Successfully Uploaded the file.<br />";
 
  // Map a virtual directory to a physical directory.
  function map_virtual_directory($virtual) {
 
    // Return successful attempt to connect to the database.
    if (!$c = @mysqli_connect(HOSTNAME,USERNAME,PASSWORD,DATABASE)) {
 
      // Print user message.
      print "Sorry! The connection to the database failed. Please try again later.";
 
      // Return error message.
      print mysqli_error();
 
      // Kill the resource.
      die();
    }
    else {
 
      // Initialize a statement in the scope of the connection.
      $stmt = mysqli_stmt_init($c);
 
      // Declare a case insensitive dynamic SQL statement.
      $sql = "SELECT directory_name FROM directory WHERE virtual_name = UCASE(?)";
 
      // Prepare the statement.
      if (mysqli_stmt_prepare($stmt,$sql)) {
 
        // Bind the input parameter to the prepared statement.
        mysqli_stmt_bind_param($stmt,'s',$virtual);
 
        // Execute the prepared statement.
        if (mysqli_stmt_execute($stmt)) {
 
          // Bind the result to a local variable.
          mysqli_stmt_bind_result($stmt,$directory);
 
          // FetchPrepare statement and link it to a connection.
          while (mysqli_stmt_fetch($stmt))
            return $directory;
        }
        else
          // Return error message.
          print mysqli_error();
      }
      else
        // Return error message.
        print mysqli_error();
 
          // Disconnect from database.
      mysqli_close($c);
    }
  }
 
  // Manage file upload.
  function process_uploaded_file($directory) {
 
    /* Assume the application may allow a virtual directory with a trailing backslash or forward
       slash to be stored in the database, and manage both scenarios across Windows and Linux. */
    if (preg_match(".Win32.",$_SERVER["SERVER_SOFTWARE"]))
      if (preg_match("/\b\\\/",$directory));
      else if (preg_match("/\b\//",$directory)) {
        $directory = substr($directory,0,strlen($directory)-1);
        $directory = $directory."\\";
      }
      else $directory = $directory."\\";
    else
      if (preg_match("/\b\//",$directory))
        $directory = substr($directory,0,strlen($directory)-1);
      else
        $directory = $directory."/";  
 
    // Check for, move uploaded file, and confirm processing.
    if (is_uploaded_file($_FILES['userfile']['tmp_name'])) {
 
      // Move temporary cache into a file directory with the uploaded file name.
      move_uploaded_file($_FILES['userfile']['tmp_name'],$directory.$_FILES['userfile']['name']);
 
      // Remove this from real code, it's here for example only. ;-)
      print "Uploaded [".$_FILES['userfile']['name']."] to".$directory."<br />";
 
      // Return true to encapsulate the functional logic on success.
	    return true;
    }
    else
      // Return false to encapsulate the functional logic on failure.
	    return false;
  }
?>

4. Create a web page to test it:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

<html>
<head>
<title>
  UploadFileFormMySQL.htm
</title>
</head>
<body>
<form id="uploadForm"
      action="MySQLFileUpload.php"
      enctype="multipart/form-data"
      method="post">
  <table border=0 cellpadding=0 cellspacing=0>
    <tr>
      <td width=125>Item Number</td>
      <td>
        <input id="id" name="id" type="text">
      </td>
    </tr>
    <tr>
      <td width=125>Select File</td>
      <td>
        <input id="uploadfilename" name="userfile" type="file">
      </td>
    </tr>
    <tr>
      <td width=125>Click Button to</td>
      <td><input type="submit" value="Upload File"></td>
    </tr>
  </table>
</form>
</body>
</html>

Hope this helps a few folks. I imagine that the prepared statement with bound variables may help a few folks because it’s not found (at writing) on the php.net web site.

While reconfiguring my iMac, it was interesting to note the bumpy road to implementing Code Weaver’s CrossOver Plus 8.0.1 for the whole Microsoft Office Suite. Previously, it had only been used to support Microsoft Visio 2007.

I discovered that a Shutdown and Start operation was required after installing CrossOver Plus because a Restart hadn’t worked. Attempting to create the WINE bottle with a reboot gave me the following error:

The Microsoft Office 2007 package requires CrossOver HTML engine to install and
run properly. Continuing with the install may produce unpredictable results.

I also found that patching the Office product failed. It’s possible that could be the network here, but there’s also no manual way to download *.msi files and apply them against the WINE bottle. It’s possible that a trick was missed and there is a way to do it. Downloading and right clicking on the *.msi file on the Mac OS X side allows you to look for an application to help but I couldn’t find a helper application in the CrossOver Plus folder tree. It may be there, however, where it is eluded me in the time my patience would allow (please comment if you know its location)?

Rather than use Google Quick Search Box I thought symbolic links would be helpful in the Applications folder since I actually also run Microsoft Office 2008. Turning to Chapter 6 in the product documentation didn’t solve the problem. It led me to an incorrect folder location.

They’re not here, as stated in the product documentation:

You can find them in the WINE bottle folders here, the default is for Windows XP:

Within the Bottle folder, you can find the Microsoft Office 2007 links here:

The CrossOver Plus installation suppresses the options dialog to choose a typical, full, or custom installation of Microsoft Office 2007. That means you get typical, which means you can’t connect to a remote database using the JDBC et cetera. This doesn’t meet my needs and means I’ll have to put another VMWare instance together to work with Microsoft Excel 2007 reasonably. Drat!

The rumor is that my database is hard, but I’ve always hoped it was fun and laid the foundation of success for my students. As I walked into class to give them their final exam, they had their jackets on, which isn’t uncommon for Rexburg, Idaho in December. However, that’s normal outside but unusual in the heated classroom. Then, they all took their jackets off to show their new t-shirts.

They thought it would be fun to post on the blog, so here it is. The shirts says:

SELECT   i.survived
FROM     michael_mclaughlin i
WHERE    class= CIT 320;

It’s unfortunate that they missed the enclosing quote marks around the string literal. 😉 It should be like this:

SELECT   i.survived
FROM     michael_mclaughlin i
WHERE    i.class= 'CIT 320';

Here’s to a great group of students who know how to read, write, and think SQL. Any openings out there for internships, please drop me a note.

Happy holidays!

When I run into failures on VMWare Fusion, they’re always a bit tedious. This one happened on my iMac (OS X Leopard) running VMWare 2 (both constrained to old releases by university governance policies). The VM is Microsoft Vista in an IDE partition, it hung after running too long. I had to force quit the application. On reboot the socket file was still there, and it gave the following error message when trying to start it.

Here’s the error in plain text, so search engines can find it for others.

Virtual device serial0: File "/var/folders/Sf/SfvoJITAHMq1Vp8bNI7QZU+++TM/-Tmp-//vmware-mmclaugh/thnuclnt-641/socket" exists, but no server is listening to it.
 
There are three possible causes for this:
 - The server is alive but not ready yet, and you can retry later.
 - The server is busy communicating with another client, so you cannot run this client at the same time.
 - A previous server exited abruptly, and you can remove the file and try again.
 
The device will be disconnected.

How to fix it?

Delete the file, right? Yes, but there’s a trick. Navigating through the -Tmp- directory required a Unix shell trick because the - (dash) is a switch and backquoting it with a \ (backslash) didn’t work. Jeff Yoder, told me the trick to change directory into a dash leading directory name. It was this:

cd -- -Tmp-

The -- is how most shells mark the end of options to a command. After a -- all - (dashes) are treated as ordinary characters.

Mark Olaveson reminded me that using the present working directory before the directory name also worked. It demotes the dash to an ordinary character too.

cd ./-Tmp-

When I got to the directory, there was the socket file. I deleted it and everything worked like a charm.

srwxrwxrwx  1 mmclaugh  staff     0 Dec  8 13:04 socket

Coming from Solaris Unix and Red Hat Linux to Ubuntu, Mac OS X and other distributions of Linux was interesting a couple years ago. The fact that you couldn’t su to the root account was tedious. It didn’t take too long (a couple momentes) to recall that I could assume those privileges in a shell, like:

admin_user@machine_name:~$ sudo sh
#

Naturally, this avoids entering sudo before a series of commands and makes administering any box simpler. It should work on all distributions but I’ve not checked ALL of them because they’re too numerous anymore. I know it works on the Mac OS X, Ubuntu, and now Fedora distributions.

Today, I got a kick from the message provided by Fedora 10 when you assume root permissions. It’s been over 20 years since I got that lecture on an AT&T box at First Interstate Bank. I imagine that any equivalent box to that is in a museum, while that bank was acquired by Wells Fargo in the early 1990s. The message from Fedora is just too funny to pass on making a comment. Here’s the screen shot:

Hope it brought a smile to some faces …

We got away with a bit of lazy PHP, at least through PHP 5.29. Especially when we worked with LOB columns. All that is at an end with PHP 5.3. You must now call the mysqli_stmt_store_results($stmt); function (line number 76 in the program snippet below from source) between the execute and bind result phrases.

The function transfers a result set from a prepared statement. You need to use it when performing a SELECT, SHOW, DESCRIBE, or EXPLAIN statement that you will call by the mysqli_stmt_fetch($stmt); function.

Without using it, you may only see part of a MEDIUMTEXT column displayed. Likewise, an image from a MEDIUMBLOB column won’t be displayed at all unless you store their results too, by using the same function call.

68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87

// Prepare statement.
if (mysqli_stmt_prepare($stmt,$sql)) {
  mysqli_stmt_bind_param($stmt,"i",$id);
 
  // Execute it and print success or failure message.
  if (mysqli_stmt_execute($stmt)) {
 
    // Store result, which eliminates memory cutting off lob streams.
    mysqli_stmt_store_result($stmt);
 
    // Bind result to local variable.
    mysqli_stmt_bind_result($stmt, $desc);
 
    // Read result.
    mysqli_stmt_fetch($stmt);
 
    // Print result.
    print $desc;
  }
}

Also, that oldie ereg() function is deprecated for preg_match() in PHP 5.3 and removed in PHP 6. Don’t forget to make those changes if you were lazy like me before updating to PHP 5.3.

It’s a bit awkward when a post generates a new question, but here’s a quick explanation and example of using XDB (XML Database Server) outside of the realm of APEX. More or less, XDB is an Apache Server equivalent configured inside the database. It’s really a protocol server tied into the Shared Server Oracle*Net Architecture (a correction provided by Marco Gralike). As a note, testing was done by using a NAT static IP addressing for the virtual Windows XP, Vista, and 7 environments.

This blog post will show you how to experiment with the PL/SQL Web Toolkit and build both password protected and unprotected database content. It assumes you have access to the SYS privileged account.

Setting Up a Secure DAD

There’s secure and then there’s secure. This falls in the less than secure category but it does provide a password and uses basic HTTP authentication. The USER is the schema name, and the PASSWORD is the same as that for the SQL*Plus access to the schema.

1. Connect as the privileged SYS user and run the following script. It creates a generic STUDENT user and grants minimalist privileges, then it creates a DAD (Data Access Descriptor), and authorizes the DAD. Don’t run the command if you’re actively using Oracle APEX on the default configuration of port 8080. It’s there for those folks you are running Tomcat on 8080.

-- This resets the default port so that it doesn't conflict with other environment.
EXECUTE dbms_xdb.SETHTTPPORT(8181);
 
-- This creates the STUDENT Data Access Descriptor.
EXECUTE dbms_epg.create_dad('STUDENT_DAD','/sampledb/*');
 
-- This authorizes the STUDENT_DAD
EXECUTE dbms_epg.authorize_dad('STUDENT_DAD','STUDENT');

2. Connect as the STUDENT user and run the following script to create a PL/SQL Web Toolkit procedure.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

CREATE OR REPLACE PROCEDURE HelloWorld AS
BEGIN
  -- Set an HTML meta tag and render page.
  owa_util.mime_header('text/html');  -- <META Content-type:text/html>
  htp.htmlopen;                       -- <HTML>
  htp.headopen;                       -- <HEAD>
  htp.htitle('Hello World!');         -- <TITLE>HelloWorld!</TITLE>
  htp.headclose;                      -- </HEAD>
  htp.bodyopen;                       -- <BODY>
  htp.line;                           -- <HR>
  htp.print('Hello ['||USER||']!');   -- Hello [dynamic user_name]!
  htp.line;                           -- <HR>
  htp.bodyclose;                      -- </BODY>
  htp.htmlclose;                      -- </HTML>
END HelloWorld;
/

3. Open a browser of your choice, and enter the following URL.

http://localhost:8181/sampledb/helloworld

Setting Up an Unsecured DAD

The trick here is building on what you did by eliminating the authentication. You do this by using the ANONYMOUS account, like Oracle’s APEX does. Well, not quite like it does because APEX provides a very good user authentication model. It allows you to connect to the ANONYMOUS user where you present and validate your credentials.

Since you have to do all the prior steps, these steps are numbered after those above. You start with step #4.

4. Generally, the XML configuration is missing one key node that allows repository anonymous access. The missing node disallows anonymous login. You can run the code below as the SYS privileged user to open that up if its missing. You can check whether or not it’s missing by running this as the SYS user:

SQL> @?/rdbms/admin/epgstat.sql

+-------------------------------------------------------------------+
| ANONYMOUS access to XDB repository:                               |
|  To allow public access to XDB repository without authentication, |
|  ANONYMOUS access to the repository must be allowed.              |
+-------------------------------------------------------------------+
 
Allow repository anonymous access?
----------------------------------
false
 
1 row selected.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40

SET SERVEROUTPUT ON
DECLARE
  lv_configxml XMLTYPE;
  lv_value     VARCHAR2(5) := 'true'; -- (true/false)
BEGIN
  lv_configxml := DBMS_XDB.cfg_get();
 
  -- Check for the element.
  IF lv_configxml.existsNode('/xdbconfig/sysconfig/protocolconfig/httpconfig/allow-repository-anonymous-access') = 0 THEN
    -- Add missing element.
    SELECT insertChildXML
           ( lv_configxml
           , '/xdbconfig/sysconfig/protocolconfig/httpconfig'
           , 'allow-repository-anonymous-access'
           , XMLType('<allow-repository-anonymous-access xmlns="http://xmlns.oracle.com/xdb/xdbconfig.xsd">'
       	   || lv_value
       	   || '</allow-repository-anonymous-access>')
       	   , 'xmlns="http://xmlns.oracle.com/xdb/xdbconfig.xsd"')
    INTO   lv_configxml
    FROM   dual;
 
    DBMS_OUTPUT.put_line('Element inserted.');
  ELSE
    -- Update existing element.
    SELECT updateXML
           ( DBMS_XDB.cfg_get()
           , '/xdbconfig/sysconfig/protocolconfig/httpconfig/allow-repository-anonymous-access/text()'
           , lv_value
           , 'xmlns="http://xmlns.oracle.com/xdb/xdbconfig.xsd"')
    INTO   lv_configxml
    FROM   dual;
 
    DBMS_OUTPUT.put_line('Element updated.');
  END IF;
 
  -- Configure the element.
  DBMS_XDB.cfg_update(lv_configxml);
  DBMS_XDB.cfg_refresh;
END;
/

5. These tasks also require the privileged SYSTEM user account, and you should sign on to that account to run these commands. The first thing you may need to do is unlock the ANONYMOUS account. It is locked by default. After you unlock it, you’ll need to verify no default password was assigned by unassigning a password. The following two commands accomplish those tasks.

-- Unlock the user account.
ALTER USER anonymous ACCOUNT UNLOCK;
-- Ensure a password is assigned to the account so you can create a synonym later.
ALTER USER anonymous IDENTIFIED BY ANONYMOUS;

6. These tasks require the privileged SYS user account because you’re going to create and authorize another DAD.

-- This creates the STUDENT_DB_DAD Data Access Descriptor.
EXECUTE dbms_epg.create_dad('STUDENT_DB_DAD','/db/*');
 
-- This authorizes the STUDENT_DB_DAD
EXECUTE dbms_epg.authorize_dad('STUDENT_DB_DAD','ANONYMOUS');
 
-- Open the anonymous account by setting the database-username parameter and value.
EXECUTE dbms_epg.set_dad_attribute('STUDENT_DB_DAD','database-username','ANONYMOUS');

7. Connect as the STUDENT user and grant EXECUTE permissions on the HELLOWORLD procedure to the ANONYMOUS user account. The GRANT allows you to give unrestricted access to the ANONYMOUS account, which in turn provides it to your web audience.

SQL> GRANT EXECUTE ON helloworld TO anonymous;

8. Connect as the ANONYMOUS user and create a local synonym that point to the STUDENT.HELLOWORLD procedure. The SYNONYM provides a program name for the URL statement. It’s hides the ownership of the actual procedure by supressing the schema name. (You may need to grant CREATE ANY SYNONYM as the SYSTEM user to the ANONYMOUS user.)

SQL> CREATE SYNONYM helloworld FOR student.helloworld;

SQL> ALTER USER anonymous IDENTIFIED BY NULL;

9. Open a browser of your choice, and enter the following URL, which won’t require a User Name or Password.

http://localhost:8181/db/helloworld