AIX

hello everybody

i am looking for a shell to cut a flat file (with a long unique line) according to a certain number of characters and redirect every result to an output file.

here is an example

MyFile :

12 3 456 12 3 456 12 3 456 .....

and i took every 9-characters including BLANKS in MyFile
so the ouput file OutFile will be like that

12 3 456
12 3 456
12 3 456
.......

i have flat files with an average of 2 Millions of characters that's why i really want to do it with a shell

Tried with awk, cut and sed but i am not really a specialist so if there is a clue just answer me cause i am in a hurry to find a solution

Used dd before?

2 megabytes really isn't that big.

Try 'man split'.

hello and thx for answering me so fast

sorry i was wrong, actually it's a text file of 200 MegaBytes !
i need to change the format of the content so it will be possible to insert the datas in a database.

i googled for "split" but i found it works with bytes and i am not sure that 1 character = 1 byte

i've heard about awk and i keep looking, if there's something that works i'll tell you

---------- Post updated at 09:36 PM ---------- Previous update was at 09:28 PM ----------

no not yet but i sounds really good, i am gonna give it a try.

anything with awk (i heard that it's a powerful tool)?

Still doable. As long as you're not doing a sort on it or any other sort of thing that requires chunking it all into memory at once, the only difference is it'll take longer. ... Try reading its man page. No googling, it just comes up. 'man split'. The version I have works with bytes OR lines, which makes it pretty clear to me that 1 byte = 1 char. But you want it in one file, not separate files, so that suggestion of mine was wrong... I don't think awk, or any other line-based tool, can operate on a line 200 megabytes long. I'll try and find something in bash...

[edit] bash took minutes and minutes to process a 5 meg file. dd works better but you have to read the man page... the 'unblock' mode converts trailing spaces to newlines, so, for the data you gave:

Code:

dd cbs=9 ibs=9 obs=9 if=longline of=longfile conv=unblock

converted a 10 meg file in 5 seconds.

Thank you very much !
you guys are great, dd works perfectly
49 seconds to convert a 200Mb-File it's really awesome

Merci beaucoup Monsieur vbe, vous êtes le meilleur!

THANK'S Corona, you're also the best, and you're right , i found this

Code:

 cat File | awk ' { for ( i = 1; i <= length($0); i++ ) a[i]=substr($0,((i-1)*9)+1,9) } END { for ( i in a ) print a[i] } ' >> output

which is couldn't work if the lengh is more than 3000 (awk limitation)

i must confess i was lost with the configuration of "dd" being a newbie in Unix i was really in need of help

dd's configuration has always been a little... a lot odd compared to other UNIX tools. When browsing the jargon file I found, much to my surprise, that its interface was perpetrated on UNIX as a practical joke poking fun at a computing platform, architecture, and interface they all hated! but they wrote it too well and it became indispensable, its basic form kept unaltered(and/or faithfully duplicated!) for decades now. The revenge of the System/360...