Hello all, I've got a rather large txt here at work that we received from another company that I'm trying to get into a database. The file shows that the furthest character is 1025 spaces out in each line. The last 700 spaces or so are blank. And the area that does have data in it (up around 340 or so characters or spaces in) is a bit longer than any of my data base utilities will handle. I'm using ssed ver. 3.48 on a Win2K workstation. I'm thinking I can use: ssed "s/[ \t]*$//" filename > newfile name to delete the trailing white spaces. But ssed didn't like that command. So I'm now trying ssed "s/\s\'//" filename > newfile name. But I'm getting a file with the same white spaces at the end. Any assistance would be much appreciated! Working guy

Hi: First, of all, by "ssed", i think you mean "sed"? Second, what was it that sed didn't like? This command works on my Solaris 7 system: ssed "s/[ \t]*$//" filename provided the \t is an actual tab character. In vi you create this by being in insert/add mode, pressing keys control-v, and then hitting the tab key. With lines as long as you're talk about, you might have overrun the sed buffer? If that's the case, you might true the GNU version of sed available for download at gnu.org. Regards, Nails

Nails, Thanks for the reply! I'm using a windows 2000 OS, no unix. The ssed stands for Super Sed. It's a windows/updated version of sed. The command ssed "s/[ \t]*$//" filename probably does run correctly in a unix environment, but I don't have one here to work on. A few years ago I did some unix admin on a HPUX box, but have no access to it or linux at this time. I'll see if I can't find a straight version of sed around that will work in Windows... The closest command I could find that would work with ssed was: ssed "s/\s\'//" filename > newfilename The features I found were: \s - any whitespace character [space, TAB, VT, FF, \n] \' - matches the end of the pattern space: same as "$" So I was hoping it would work. It seems to create the second file but all the data comes over including the 700 or so Whitespaces. Thanks again for your input. Working guy

ssed appears to use perl regexps but anyway it looks like you only searched for one whitespace character. Did you try this?: ssed "s/\s+\'//" filename > newfilename

Woldbone, I just tried: ssed "s/\s+\'//" filename > newfilename And got the same thing. The last part of the data on every line in the file is 2002. The first set of data on each line is the same as is the last set of data. Maybe I could try a command to just grab everything from the first set of data and the last set on each line and export it to a new file? An example would be: ssed "s/\{start,end\}/ err.. I'm not sure what the rest would be.. any help would be great here.. Thanks Working guy

I'm not sure, I thought from what you wrote earlier that the rightmost part of each line is just whitespace but if it is "2002" then you'd need to use that instead of "\'" but if not then Nails would know better about perl style regexps than me anyway.

The right 700 charcters are all whitespaces. The "2002" starts at 321 and ends at 325. I was suggesting maybe I could grab everything between the beginning of the line "start" and the end of the data in the line "2002". Does that make sense? Working guy