Learn how to create request header fields and parse response header fields when making HTTP calls.
In the first tip, I introduced the new suite of HTTP functions in DB2 for i 7.1 and demonstrated the HTTPGETCLOB and HTTPGETBLOB scalar functions. I'll continue the topic by considering how to include and retrieve HTTP header information with each request. This tip assumes you have some familiarity with the new HTTP functions and with web development concepts. IBM i 7.1, DB2 PTF group SF99701 Level 23, and Java 6 are required.
Anyone who has done web development knows that HTTP header fields are part of the standard HTTP protocol and are used to pass additional information about an HTTP request or response between the client (usually a browser) and the HTTP server. A "request" is made to the web server each time a web page, web service, or other resource is requested. The "response" consists of the information that the web server returns.
The table below shows a list of common request HTTP header fields (from Wikipedia). Your HTTP-enabled application can set any of these fields when making a request to a web server:
HTTP Header Fields |
||
Field Name |
Description |
Example |
Accept |
Content types that are acceptable for the response |
Accept: text/plain |
Accept-Encoding |
Accept-Encoding: gzip, deflate |
|
Accept-Language |
List of acceptable human languages for response |
Accept-Language: en-US |
Authorization |
Authentication credentials for HTTP authentication |
Authorization: Basic QWxhZGRpbjpvcGVuIHNlc2FtZQ== |
Cache-Control |
Used to specify directives that must be obeyed by all caching mechanisms along the request/response chain |
Cache-Control: no-cache |
An HTTP cookie previously sent by the server with Set-Cookie (in the next table) |
Cookie: $Version=1; Skin=new; |
|
Content-Length |
Content-Length: 348 |
|
Content-Type |
The MIME type of the body of the request (used with POST and PUT requests) |
Content-Type: application/x-www-form-urlencoded |
Host |
The domain name of the server (for virtual hosting) and the TCP port number on which the server is listening |
Host: en.wikipedia.org:80 |
Pragma |
Implementation-specific headers that may have various effects anywhere along the request-response chain |
Pragma: no-cache |
Proxy-Authorization |
Authorization credentials for connecting to a proxy |
Proxy-Authorization: Basic QWxhZGRpbjpvcGVuIHNlc2FtZQ== |
Range |
Request only part of an entity. Bytes are numbered from 0. |
Range: bytes=500-999 |
This is the address of the previous web page from which a link to the currently requested page was followed. (The word "referrer" is misspelled in the RFC as well as in most implementations.) |
Referer: http://en.wikipedia.org/ wiki/Main_Page |
|
User-Agent |
The user agent string of the user agent |
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:12.0) Gecko/20100101 Firefox/21.0 |
As you can see, there are things of interest to a DB2 developer, such as "Cookie" (in case you have to set cookie information when exchanging data with a traditional web app) and "Authorization" (for when you have to "login" before accessing a service). When sending information using POST or PUT, Content-Type can be used to specify whether the content is plain text, XML, a jpg image, etc.
Likewise, when an HTTP server sends a response back to a client (such as HTML, XML, or a file), it returns extra information in the form of response header fields. The table below shows a few of them (from Wikipedia):
Response Header Fields |
||
Field Name |
Description |
Example |
Allow |
Valid actions for a specified resource. To be used for a 405 Method not allowed. |
Allow: GET, HEAD |
Content-Disposition |
An opportunity to raise a "File Download" dialogue box for a known MIME type with binary format or suggest a filename for dynamic content. Quotes are necessary with special characters. |
Content-Disposition: attachment; filename="fname.ext" |
Content-Encoding |
The type of encoding used on the data. See HTTP compression. |
Content-Encoding: gzip |
Content-Language |
The language the content is in |
Content-Language: da |
Content-Length |
Content-Length: 348 |
|
Content-Location |
An alternate location for the returned data |
Content-Location: /index.htm |
Content-MD5 |
A Base64-encoded binary MD5 sum of the content of the response |
Content-MD5: Q2hlY2sgSW50ZWdyaXR5IQ== |
Content-Range |
Where in a full body message this partial message belongs |
Content-Range: bytes 21010-47021/47022 |
Content-Type |
Content-Type: text/html; charset=utf-8 |
|
Date |
The date and time that the message was sent |
Date: Tue, 15 Nov 1994 08:12:31 GMT |
Expires |
Gives the date/time after which the response is considered stale |
Expires: Thu, 01 Dec 1994 16:00:00 GMT |
Last-Modified |
The last modified date for the requested object, in RFC 2822 format |
Last-Modified: Tue, 15 Nov 1994 12:45:26 GMT |
Pragma |
Implementation-specific headers that may have various effects anywhere along the request-response chain |
Pragma: no-cache |
Retry-After |
If an entity is temporarily unavailable, this instructs the client to try again after a specified period of time (seconds). |
Retry-After: 120 |
Server |
A name for the server |
Server: Apache/2.4.1 (Unix) |
Set-Cookie |
Set-Cookie: UserID=JohnDoe; Max-Age=3600; Version=1 |
|
Status |
Status: 200 OK |
|
Transfer-Encoding |
The form of encoding used to safely transfer the entity to the user. Currently defined methods are chunked, compress, deflate, gzip, identity. |
Transfer-Encoding: chunked |
WWW-Authenticate |
Indicates the authentication scheme that should be used to access the requested entity |
WWW-Authenticate: Basic |
When you surf the web with your favorite web browser, you don't need to pay attention to what's being exchanged in these fields because it's done behind the scenes. However, when programming an HTTP request to, say, communicate with a web service using these DB2 for i tools (or similar tools), the web server may require that portions of the header be populated in a special way.
Since you may be making an HTTP call (of any kind) that requires specifying or inspecting a header field, I'll discuss how to do this with the new DB2 HTTP functions. For example, I wrote a program to communicate with a web service in which the service provider, as a security measure, required a specific value to be placed in the User-Agent header request field. The User-Agent normally holds a description of the browser and host OS, but it can be overridden to anything and this vendor required an override. Without setting this header field appropriately, the application wouldn't work.
Parsing the DB2 for i HTTP Response Headers
The HTTP protocol requires header fields to be specified in plain text and follow a certain format (name value pairs, using carriage return and line feed, etc.). Wikipedia covers this. Fortunately, we don't have to worry about that stuff. The DB2 functions use an easy-to-understand XML template to exchange this information. I'll explain how to read the server's response header fields and how to send custom values in the request header fields.
To examine the HTTP response header fields returned from a web call, use table function HTTPGETCLOBVERBOSE. This function is similar to HTTPGETCLOB with the exception that it's a table function that returns one row with two columns:
- RESPONSEMSG contains the data returned from the web server (same info as returned by HTTPGETCLOB).
- RESPONSEHTTPHEADER contains information about the headers returned from the server. These values can be interrogated for error codes and the header field name/value pairs.
Most of the other scalar functions have a VERBOSE table function companion (e.g., HTTPGETBLOB (scalar) and HTTPGETBLOBVERBOSE (table)).
Recall the simple Google search example from the first tip that simply returns the HTML results of a Google search for A&W Root Beer:
Select SYSTOOLS.HTTPGETCLOB
('https://www.google.com/search?q=A%26W+Root+beer&;client=AS400','')
From SYSIBM.SYSDUMMY1;
By invoking the verbose table function, it's possible to review the response header fields that Google.com's web server returns as follows:
SELECT * FROM TABLE(SYSTOOLS.HTTPGETCLOBVERBOSE
('https://www.google.com/search?q=A%26W+Root+beer&;client=AS400','')) AS HTTPDATA
The RESPONSEMSG column returns the same information as the scalar function HTTPGETCLOB (in this case, HTML search results). The RESPONSEHTTPHEADER column returned the following XML (formatted and abridged):
<?xml version="1.0" encoding="UTF-8" ?>
<httpHeader responseCode="200">
<responseMessage>OK</responseMessage>
<header name="HTTP_RESPONSE_CODE" value="HTTP/1.1 200 OK"/>
<header name="Cache-Control" value="private, max-age=0"/>
<header name="Server" value="gws"/>
<header name="X-XSS-Protection" value="1; mode=block"/>
<header name="Expires" value="-1"/>
<header name="Transfer-Encoding" value="chunked"/>
<header name="X-Frame-Options" value="SAMEORIGIN"/>
<header name="Date" value="Sat, 08 Jun 2013 03:15:43 GMT"/>
<header name="Content-Type" value="text/html; charset=ISO-8859-1"/><header name="Set-Cookie" value="NID=67=PO7G2a7WLH9QQsCaw1Kn-w80EDGbwuzGiV1s7dLXbTZ-AKzX01Pk-JuOXapUOL_RpUEYoPrCdYarz3Ce5u5SeX4L18FO73ve-Yrp70cBCTVGfdKSirT5RVOHMclc0km7; expires=Sun, 08-Dec-2013 03:15:43 GMT; path=/; domain=.google.com; HttpOnly"/>
</httpHeader>
Besides the response fields, this XML contains additional information from the HTTP server (shown in bold). The responseCode attribute and the responseMessage contain status messages the server returns. The responseCode attribute should always be a three-character response code. The standard HTTP status code classifications are listed below, many of which your application should be prepared to handle when encountered (from Wikipedia):
- 1xx Informational—The request has been received and is continuing the process.
- 2xx Success—This class of status codes indicates the action requested by the client was received, understood, accepted, and processed successfully.
- 3xx Redirection—The client must take additional action to complete the request.
- 4xx Client Error—This status code is intended for cases in which the client seems to have erred.
- 5xx Server Error—The server failed to fulfill an apparently valid request.
If a 2xx value (Success) isn't returned, prepare to have the app do something about it. If an application receives a 3xx redirection error, the HTTP functions will by default follow the redirection (except when the redirect specifies a different protocol, such as from HTTP to HTTPS).
After the HTTP status is listed, the XML presents a series of <header> elements that contain the header field name/value pairs. The following query will parse the response's status information and the header fields using two instances of the XMLTABLE table function combined with a UNION ALL:
WITH ResponseHeader AS (
SELECT XMLPARSE(DOCUMENT RESPONSEHTTPHEADER) as "doc"
FROM TABLE(SYSTOOLS.HTTPGETCLOBVERBOSE
('https://www.google.com/search?q=A%26W+Root+beer&;client=AS400',''))
AS HTTPDATA
)
SELECT 'Response' As Type,HeaderResponse.*
FROM XMLTABLE(
'$doc/httpHeader' PASSING (SELECT "doc" FROM ResponseHeader) as "doc"
COLUMNS
Name VARCHAR(10) PATH '@responseCode',
Value VARCHAR(128) PATH 'responseMessage'
) HeaderResponse
UNION ALL
SELECT 'HeaderField' As Type,HeaderResponse.*
FROM XMLTABLE(
'$doc/httpHeader/header' PASSING (SELECT "doc" FROM ResponseHeader) as "doc"
COLUMNS
Name VARCHAR(128) PATH '@name',
Value VARCHAR(128) PATH '@value'>
) HeaderResponse;
The results are shown in this table:
Query Results |
||
Type |
Name |
Value |
Response |
200 |
OK |
HeaderField |
HTTP_RESPONSE_CODE |
HTTP/1.1 200 OK |
HeaderField |
Cache-Control |
private, max-age=0 |
HeaderField |
Server |
gws |
HeaderField |
X-XSS-Protection |
1; mode=block |
HeaderField |
Expires |
-1 |
HeaderField |
Transfer-Encoding |
Chunked |
HeaderField |
X-Frame-Options |
SAMEORIGIN |
HeaderField |
P3P |
CP="This is not a P3P policy! See http://www.google.com/support/accounts/bin/answer.py?hl=en&;answer=151657 for more info." |
HeaderField |
Date |
Thu, 14 Mar 2013 19:00:44 GMT |
HeaderField |
Content-Type |
text/html; charset=ISO-8859-1 |
HeaderField |
Set-Cookie |
NID=67=PO7G2a7WLH9QQsCaw1Kn-w80EDGbwuzGiV1s7dLXbTZ-AKzX01Pk-JuOXapUOL_RpUEYoPrCdYarz3Ce5u5SeX4L18FO73ve-Yrp70cBCTVGfdKSirT5RVOHMclc0km7 |
This query demonstrates how to get the web server's header response (status code, summary, and header fields) simultaneously. In practice, one would normally fetch the response into a variable, check the status info and, if the request is OK, move on to processing the header fields and then the data itself.
Passing Request Headers
As previously discussed, sometimes it's a requirement to pass specific values in a request header to a web server in order for the request to be successful. I mentioned an application I wrote where I had to override the User-Agent field.
So using that simple example, how can an application using the new DB2 for i functions set the User-Agent request field to the arbitrary value of "AS/400" before making the request? The answer is that the DB2 HTTP functions have a second parameter called HTTPHEADER. Although the parameter is defined as a CLOB, it is intended to receive XML data formatted as follows:
<httpHeader>
<header name="User-Agent" value="AS/400"/>
</httpHeader>
This is basically the same format as the response header XML parsed in the prior section. So if HTTPGETBLOB is being used to make an HTTP call, setting the User-Agent field is as simple as slapping in the properly formatted XML in the second parameter:
SELECT SYSTOOLS.HTTPGETBLOB('http://mydomain.com/appext/',
'<httpHeader>
<header name="User-Agent" value="AS/400" />
</httpHeader>')
FROM SYSIBM.SYSDUMMY1;
Note that header field names are supposed to be case-insensitive, so User-Agent and USER-AGENT should work the same.
One special thing to note about passing header fields is that IBM has allowed a few additional options to be specified that can control the behavior of the HTTP functions. The table below contains the attribute names that can be added to the httpHeader element (from "HTTP request header fields and connection properties" in Accessing web services using IBM DB2 for i HTTP UDFs and UDTFs):
Attribute Names That Can Be Added to the httpHeader |
|||
Name |
Value |
Default |
Comment |
connectTimeout
|
Integer |
System default |
Maximum amount of time the JVM will wait for the connection in milliseconds |
readTimeout |
Integer |
System default |
Maximum amount of time the JVM will wait for reading data in milliseconds |
followRedirects |
True/False |
True |
Indicates whether redirects should be implicitly followed when a 3xx response code is received from the server |
useCaches |
True/False |
True |
Instructs the JVM that caches are allowed to be used if available. DB2 for i does not implement a cache in the HTTP UDFs and UDTFs; however, a default cache might be used if the default cache is registered with the JVM. |
To override the connection timeout to 1/2 second (500ms) and specify that redirects should not be automatically followed, specify the attributes as follows:
<httpHeader connectTimeout="500" followRedirects="false">
<header name="User-Agent" value="AS/400"/>
</httpHeader>
Note that the IBM manual has a typo error in that connectTimeout is incorrectly shown in the table as connectionTimeout.
For one more example, let's look at populating the Authorization field for when the web server requires basic authentication (i.e., a plain-text user name and password). With most modern web servers, when basic authentication is required, the user name and password can be passed in plain text on the URL as follows:
https://user:
As you might imagine, using secure HTTP (SSL/TSL) is desirable; otherwise, the plain text information can be easily captured.
If credentials can be passed via URL, why bother populating the authorization field? Because I found it doesn't always work in the URL.
When researching the prior tip on using the new HTTP functions, I found an example of how to access Gmail with the DB2 HTTP functions on IBM's developerWorks site. Search the page for "Listing 5," where an example shows how to query a Gmail account and return all unread messages. The result is an Atom (XML) document that can easily be shredded by the XMLTABLE table function.
The example the developerWorks article shows how to access your Gmail info by placing the user name and password in the URL. However, with the DB2 for i functions, I found at least one case when the credentials wouldn't pass via the URL; this happens when you have a password containing the "at" (@) symbol. In this case, I found the only way I could log on is by passing credentials via the Authorization request field instead of the URL.
When passing the authorization value, the credentials should be sent as user:password. This result should then be encoded as Base64.
If my Gmail account is MikeSanso and password is F@t@l, the authorization value (before Base64 encoding) would look like this:
MikeSanso:F@t@l
Further, before encoding this credential as Base64, make sure to cast it to UTF-8 (aka CCSID 1208); otherwise, you'll Base64 encode an EBCDIC string, which won't work!
Here's the revised developerWorks query with my modifications for passing the Authorization info:
WITH Authorization AS (
SELECT SYSTOOLS.BASE64ENCODE(CAST('MikeSanso:F@t@l' AS VARCHAR(64) CCSID
1208)) AS Authorization_Identity
FROM SYSIBM.SYSDUMMY1
)
SELECT RESULT.*
FROM Authorization, XMLTABLE('$result/*[local-name()=''feed'']/*[local-name()=''entry'']'
PASSING
XMLPARSE(DOCUMENT
SYSTOOLS.HTTPGETBLOB('https://mail.google.com/mail/feed/atom/',
'<httpHeader>
<header name="Authorization" value="Basic '||Authorization_Identity||'" />
</httpHeader>'))
AS "result"
COLUMNS
title VARCHAR(128) PATH '*[local-name()=''title'']',
summary VARCHAR(1024) PATH '*[local-name()=''summary'']',
author_name VARCHAR(255) PATH '*[local-name()=''author'']/*[local-name()=''name'']',
author_email VARCHAR(255) PATH '*[local-name()=''author'']/*[local-name()=''email'']'
) AS RESULT
Using the authentication header to pass credentials instead of the URL, although more complex, is recommended by IBM because there are cases when an SQL error or job log entry may reveal the URL and thereby disclose a user ID and password. If you have a Gmail account, you can try this out yourself.
In the end, the ability to pass header request values and inspect response values gives your application the versatility needed to exchange data with HTTP servers. Using headers allows your HTTP application interfaces to manage cookies, specify a content type when sending files, and send credentials, to name just a few possibilities.
References
TechTip: Check Out the New DB2 for i HTTP Functions
Accessing web services using IBM DB2 for i HTTP UDFs and UDTFs
LATEST COMMENTS
MC Press Online