Programming ColdFusion

CFHTTP Considerations

While CFHTTP is extremely powerful, there are a few limitations in the current version that are worth mentioning. These issues include problems with CFHTTP parsing text files that don’t contain column headers and methods for interacting with pages that require authentication. If you plan to make heavy use of CFHTTP, you should carefully consider the issues discussed here.

Column Headers and Delimited Text Files

There is a problem with the current implementation of CFHTTP regarding the parsing of text files (that don’t contain column headers) into query objects. Using CFHTTP to get a delimited text file (that doesn’t contain column headers in the first row of data) causes CFHTTP to ignore the first row of the file. This results in the first row of data in the text file being left out of any subsequent operations. This occurs even if alternate column headers are specified in the COLUMNS attribute. There is currently no workaround for this problem other than ensuring that the first row of the delimited file contains column headers or a dummy record. Example 14-8 illustrates this point, by showing how CFHTTP retrieves a text file (that contains stock quote information) from Yahoo’s web site. The text file doesn’t contain any column headers. In this case, it is possible to get a dummy record by having the CGI program that generates the file include a duplicate record for the first record. In this example, passing the ticker symbol “YHOO” twice in the URL does ...