glimpse and webglimpse


Glimpse is a marvelous indexing facility. The following fragmentary notes were left over from some of my first adventures in installing software on a Linux system.

using glimpse

My first experiment with glimpse correctly indexed all of Swan's cpp files -- and here was my hugely successful second experiment: from my home directory,

glimpseindex ~

indexed all of my files (takes less than a minute for 6000 files) -- then

glimpse -F .cpp friend

listed the path to every occurrence of the word "friend" in all of my .cpp files, together with the line in which that word occurs -- in a few seconds!! One such entry might be

/home/parrish/swan/c11/mutual.cpp: friend ostream &

installing glimpse

Download and unpack glimpse into a local directory.

cp files in /bin subdirectory of the glimpse local directory to /usr/bin (or some other directory on your search path)

cp files in /man subdirectory of glimpse local directory to /usr/local/man/man1

installing webglimpse

Download and unpack into a local directory. Read installation instructions.

Run ./wginstall

It will ask some questions. I chose the defaults in every case.

It answers

You may need to edit webserver conf files and specify /usr/local/webglimpse/cgi-bin as a cgi-bin directory. For Apache use ScriptAlias directive.

Then it says

You may configure an archive at any time with the command


A directory should be a full path starting with /
A URL should be an absolute path starting with http://

modifying Apache server to be webglimpse compatible

As root, I entered the following Apache directive in /etc/httpd/conf/srm.conf >ScriptAlias /cgi-bin/ /usr/local/webglimpse/cgi-bin/

but I have not yet tested webglimpse because I only have one page at my website!