Finding specfic string within another string

I am having trouble finding a specific part of a file path. The path is something along: \\blabla\blabla\blabla\blabla\blabla\blabla\04 - Black\05 - TEST

I need to find "04 - Black" in the path name. The name is changing and so is its position in the path - Therefor I need to find the first instance of 'XX - '. Afterwards I need to find the "05 - TEST" folder, which is also changing name and position.

Can anyone help?

Answer 1

[KSSV]

str1 = '\\blabla\blabla\blabla\blabla\blabla\blabla\04 - Black\05 - TEST';

str2 = '04 - Black' ;

idx = strfind(str1,str2)
str1(idx:(idx+length(str2)-1))

[Jan]

Safer:

idx = strfind(str1, ['\', str2, '\'])

Answer 2

[José-Luis]

yourStr = '\blabla\blabla\blabla\blabla\blabla\blabla\04 - Black\05 - TEST'
re = '\\[0-9]{2}\s-\s[a-zA-Z]+';
result = regexp(yourStr,re,'match');

[Guillaume]

Other possible regex, which is more permissive:

re = '\\[0-9]{2}\s-\s[^\\]+';

Allows anything that is not \ after the two digits, whitespaces, and dash.

Answer 3

[Guillaume]

if the aim is to obtain the name of the folder after the 'XX - ', then:

folder = regexp(path, '(?<=\\[0-9]{2}\s-\s[^\\]+\\)[^\\]+', 'match', 'once')

Answer 4

[Cedric]

Why not simply

 >> str =  '\\blabla\blabla\blabla\blabla\blabla\blabla\04 - Black\05 - TEST' ;

 >> folders = regexp( str, '\d\d - [^\\]+', 'match' )

 folders =
  1×2 cell array
    '04 - Black'    '05 - TEST'

[Cedric]

Note that to make it portable, you should build the pattern based on filesep.

[Stephen23]

Or just use a forward-slash, which works on Windows and Linux.

[Cedric]

Not in this direction ;-)

If he gets the path from PWD, DIR, FULLFILE, etc on a Windows machine, it will have backslashes.

[Image Analyst]

Those functions may return backslashes, but you can use forward slashes in your strings that you make up and send into functions like that with no problems. Like Stephen says, Windows accepts either forward or backward slashes equally well.

[Cedric]

Yes but we need to match a sub-string in this context and for this he needs to use the correct separator in the pattern, for REGEXP to work. He could use both for ending the match by defining the pattern as '\d\d - [^\\/]+'.

But again, if there is no problem in the "other direction" (because MATLAB can deal with either filesep), here both possibilities have to be managed, either by building the pattern using filesep or by making the pattern flexible enough.

Answer 5

[Lucas Silva]

Eu tenho que fazer um processo semelhante, mas o método de separar as seqüências de caracteres é o ';' Eu tentei identificar o; como se fosse parte do comando, mas o matlab não aceita como.

Existe alguma maneira de obter essa informação?