man llsearch (Commandes) - Search a GNIS file for place names within a given block of latitude/longitude
NAME
llsearch - Search a GNIS file for place names within a given block of latitude/longitude
SYNOPSIS
llsearch [-L] | [latitude_low longitude_low latitude_high longitude_high]
DESCRIPTION
The U.S. Geological Survey supports sites on the Internet with Geographic Names Information System (GNIS) files. These files contain lists of place names, complete with their latitude/longitude and other information. There are separate files for each of the U.S. states, and each file contains many, many, many place names. If you want to use this data with drawmap, it is useful to reduce the data to only the items that you need. Llsearch lets you filter a GNIS file and winnow out only those place names that fall within the latitude/longitude boundaries that you specify. (You may want to specify boundaries that are a tiny bit larger than what you are interested in, so that numerical quantization doesn't eliminate locales that fall exactly on your boundaries.)
Latitudes and longitudes are positive for north latitude and east longitude, and negative for south latitude and west longitude. Llsearch expects you to enter them in decimal degrees. (The latitudes and longitudes in the GNIS file are in degrees-minutes-seconds format, followed by 'N', 'S', 'E', or 'W'. However, there are two available file formats, and one of the formats also contains the latitudes/longitudes in decimal degrees.) Typical usage is as follows:
- gunzip -c california.gz | llsearch 33 -118 34 -117 > gnis_santa_ana_west
If you enter the "-L" option, the program will print some license information and exit.
Once you have reduced the data to some subset of interest, you can search for particular items via the grep or perl commands, or other search commands, or you can simply edit the results with your favorite text editor. Search commands are useful in reducing the sheer volume of data to a more manageable size (by extracting, say, all mountain summits or all streams), but you will probably ultimately end up looking through the remaining data manually. The individual records contain codes, such as "ppl" for populated places, and "summit" for mountain tops, that can help you pick and choose.
There is considerable redundancy in place names, and human intelligence is useful in sorting things out. While I was writing drawmap and llsearch, I frequently gazed out my office window, where I could spot at least two, and possibly three Baldy Mountains. There are also quite a few Beaver Creeks, Bear Canyons, Saddle Buttes, and Springfields out there. By taking a close look at the information associated with each place name, you can find the particular locations that interest you.