How to Run the Hacks
The programmatic hacks in this book run either on the command line
(that's Terminal for Mac OS X folk, DOS command
window for Windows users) or as CGI scripts—dynamic pages
living on your web site, accessed through your web browser.
Command-Line Scripts
Running a hack on the command line invariably involves the following
steps:
Type the program into a garden-variety text editor: Notepad on
Windows, TextEdit on Mac OS X, vi or
Emacs on Unix/Linux, or anything else of the sort.
Save the file as directed—usually as
scriptname.pl (the pl bit
stands for Perl, the predominant programming language used in
Google Hacks). Alternatively, you can download the code for all of the hacks online
at http://www.oreilly.com/catalog/googlehks2, a
ZIP archive filled with individual scripts already saved as text
files. Get to the command line on your computer or remote server. In Mac OS
X, launch the Terminal
(ApplicationsUtilitiesTerminal). In Windows,
click the Start button, select Run..., type
command, and hit the Enter/Return key on your
keyboard. In Unix ... well, we'll just assume you
know how to get to the command line. Navigate to where you saved the script at hand. This varies from
operating system to operating system, but usually involves something
like cd ~/Desktop
(that's your Desktop on the Mac). Invoke the script by running the programming
language's interpreter (e.g., Perl) and feeding it
the script (e.g., scriptname.pl) like so: $ perl scriptname.pl
Most often, you'll also need to pass along some
parameters—your search query, the number of results
you'd like, and so forth. Simply drop them in after
the script name, enclosing them in quotes if they're
more than one word or if they include an odd character or three: $ perl scriptname.pl '"much ado about nothing" script' 10
The results of your script are almost always sent straight back to the command-line window in which you're working, like so: $ perl scriptname.pl '"much ado about nothing" script' 10
1. "Amazon.com: Books: Much Ado About Nothing: Screenplay ..."
[http://www.amazon.com/exec/obidos/tg/detail/-/0393311112?v=glance]
2. "Much Ado About Nothing Script"
[http://www.signal42.com/much_ado_about_nothing_script.asp]
...
|
The elllpsis (...) bit signifies that
we've cut off the output for
brevity's sake.
|
|
To stop output scrolling off your screen faster than you can read it,
on most systems you can "pipe"
(read: redirect) the output to a little program called more: $ perl scriptname.pl | more
Hit the Enter/Return key on your keyboard to scroll through line by
line, the space bar to leap through page by page. You'll also sometimes want to direct output to a
file for safekeeping, importing into your spreadsheet application, or
displaying on your web site. This is as easy; refer to the code shown
next. $ perl scriptname.pl > output_filename.txt
And to pour some input into your script from a file, simply do the
opposite: $ perl scriptname.pl < input_filename.txt
Don't worry if you can't remember
all of this; each hack has a "Running the
Hack" section, and some even have a
"The Results" section that shows
you just how it's done.
CGI Scripts
CGI scripts—programs that run on your web site and produce
pages dynamically—are a little more complicated if
you're not used to them. While fundamentally
they're the same sort of scripts as those run on the
command line, they are more troublesome because setups vary so
widely. You may be running your own server, your web site may be
hosted on an Internet service provider's (ISP)
server, your content may live on a corporate intranet server—or
anything in between.
Since going through every possibility is beyond the scope of this (or
any) book, you should check your ISP's knowledge
base or call their technical support department, or ask your local
system administrator for help.
Generally, though, the methodology is the same:
Type the program in to a garden-variety text editor: Notepad on
Windows, TextEdit on Mac OS X, vi or
Emacs on Unix/Linux, or anything else of the sort.
Save the file as directed—usually as
scriptname.cgi (the cgi bit
reveals that you're dealing with a
CGI—that's common gateway
interface—script). Move the script over to wherever your web site lives. You should have
some directory on a server somewhere in which all of your web pages
(all those .html files) and images (ending in
.jpg, .gif, etc.) live.
Within this directory, you'll probably see something
called a cgi-bin directory: this is where CGI
scripts must usually live in order to be run rather than just
displayed in your web browser when you visit them. You usually need to bless CGI scripts as executable—to be run
rather than displayed. Just how you do this depends on the operating
system of your server. If you're on a Unix/Linux or
Mac OS X system, this usually entails typing the following on the
command line: $ chmod 755
scriptname.cgi
Now you should be able to point your web browser at the script and
have it run as expected, behaving in a manner similar to that
described in the "Running the Hack"
section of the hack at hand. Just what URL you use once again varies wildly. It should, however,
look something like this: http://www.your_domain.com/cgi-bin/scriptname.cgi, where
your_domain.com is your web site domain,
cgi-bin refers to the directory in which your
CGI scripts live, and scriptname.cgi is the
script itself. If you don't have a domain and are hosted at an ISP,
the URL is more likely to look like this:
http://www.your_isp.com/~your_username/cgi-bin/scriptname.cgi, where
your_isp.com is your
ISP's domain,
~your_username is
your username at the ISP, cgi-bin refers to the
directory in which your CGI scripts live, and
scriptname.cgi is the
script itself.
If you come up with something called an "Internal
Server Error" or see the error code 500,
something's gone wrong somewhere in the process. At
this point you can take a crack at debugging (read: shaking the bugs
out) yourself or ask your ISP or system administrator for help.
Debugging—especially CGI debugging—can be a little more
than the average newbie can bear, but there is help in the form of a
famous Frequently Asked Question (FAQ): "The
Idiot's Guide to Solving Perl CGI
Problems." Google for it and step through as
directed.
Using the Google API
Be sure to consult Chapter 9 for an
introduction to the Google API, how to sign up for a
developer's key—you'll need
one for many of the hacks in this book—and the basics of
programming Google in a selection of languages to get you going.
Learning to Code
Fancy trying your hand at a spot of programming?
O'Reilly's best-selling
Learning Perl (http://www.oreilly.com/catalog/lperl3) by
Randal L. Schwartz and Tom Phoenix provides a good start. Apply what
you learn to understanding and using the hacks in this book, perhaps
even taking on the "Hacking the
Hack" sections to tweak and fiddle with the scripts.
This is a useful way to get a little programming under your belt if
you're a searching nut, since it's
always a little easier to learn how to program when you have a task
to accomplish and existing code to leaf through.
|