Define regularExpression for Strsplit?

Hi, i have a text file consisting of multiple headers with Data in between. Now, i want to use Strsplit to find the sequence of the headers and save them under a Variable.

The Problem is the all Headers are different from each other, instead of the first Word and letter.

ex.

1. abc 0 AAA BB CC DDD 111
2. abc 0 EEE FF GGG HH II 120

1. and 2. are not the Content of the Header. so the abc and 0 are tab delimited. after 0 there is no common sequence for the line, some of the words are tab delimited some with space, some have numbers some not.

The ending of the File can be 'just' one of These, because it has more thousands headers.

110, 005, 006, 010, 133/1A, 230, 400, NWD

The Expression i started Looks like this:

xpr = '(?m-s)^abc\s+';

but how to define the ending?

Anyone to suggest something?

Answer 1

[per isakson]

I've understood the question differently

str = fileread( ... );
[ data_blocks, headers ] = strsplit( str, '(?m-s)^abc\t0.+$', 'DelimiterType','RegularExpression' )

[FishermanJack]

Thanks a lot, this works for me.. i dont understand why this works because as i read the Help the '\w*m$' matches words ending with m at the end of the text.

So how does the 0.+$ know i have NWD at the end.???

[Walter Roberson]

It does not. That expression will match from ^abc to the end of that one line. Your ?-s flag is equivalent to 'dotexceptnewline', so the .+ will not cross line boundaries.

[FishermanJack]

thanks for the Explanation :D

[per isakson]

I didn't fully understood the questions and thus I made some assumptions:

1. The entire lines shown in the question are headers. The blocks of data between the headers are not shown in the question.
2. Every line starting with abc\t0 is a header. ( \t for tab )
3. I still don't understand the role of the string NWD. Does it indicate the last header of the file? I ignored it.

[FishermanJack]

1. Yes 2. Yes 3. No These are some random letters that are at the end of some Headers, like 110 or 005

Answer 2

[Walter Roberson]

S = fileread('YourFile.txt');
regexp(S, '^abc\s.*110,\s+005,\s+006,\s+010,\s+133/1A,\s+230,\s+400,\s+NWD', 'match', 'lineanchors')

If all of the headers are the same and there is nothing between the 110* line and the next header, then consider

regexp(S, '^abc\s', 'split', 'lineanchors')

The 'abc' will be removed from each block during the splitting; it would be possible to get around that but doing so is somewhat more obscure.