reading a large number of files

[TheKKK]

I have to read and extract data from a large number of files:

file1.dat
file2.dat
file3.dat
.........
file100.dat

Is there a fast way of reading all these files one at a time?

My simple idea is to write all the filenames in a text file from which i will read a filename at a time. If anyone knows a faster solution im glad to hear it.

Thanx

 

[IRstuff]

Speed of reading files is dependent on the IO method, usually, not how you get your filenames. Nonetheless, since your filenames are sequential, you could just store the common root and append the number for each new file.

As for reading the file itself, the preference would be binary format, since that goes much faster than text.

TTFN

FAQ731-376

 

[TheKKK]

What if my filenames have not a common root?

If i have:

GR01.dat
GR07.dat
FR01.dat
FR02.dat
IT01.dat
IT06.dat
........

Is there a way fortan could read one by one all these...
if i dont want to write all the filenames in a .txt file?

Thanx!

 

[Denial]

Presumably your data files has some unique defining characteristic, such as all having a .dat filename extension and being the only .dat files in the directory.

If so, have your program initiate a directory operation to list the files into some new file (of a different extension). Once this file has been created, your program can open it, then read the .dat file names one by one and open them for you to extract the stuff you want.

In good ol' DOS, the directory operation would have been
DIR *.DAT /B > FILES.LST

 

[TheKKK]

What about in linux?

 

[IRstuff]

The file output redirection is similar in Unix

TTFN

FAQ731-376

 

[SomptingGuy]

ls -1 > files.dat

(One, not l or L)

In most dialects.

- Steve

 

[johnhors]

to get just the dat files in linux

ls -1 *.dat > files.txt