Is it possible to use two textscan statements in a row ?

Question

Gwendal Marrec 2022-3-28

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1682529-is-it-possible-to-use-two-textscan-statements-in-a-row

评论： Gwendal Marrec 2022-3-30

采纳的回答： Stephen23

Hello,

I am trying to extract some data from a text file.

My text file contains several lines written this way:

I want to extract the parameters after the colons.

Here's my code :

It is supposed to return me C1 = "Mars 2022" and C2 = "16:17:50".

I get the following error message :

It seems like I cannot use two textscan statements in a row. In the workspace, C1 = "Mars 2022" but C2 is an empty cell.

Two additional questions :

how to handle the empty string after 'nom' ? I would like that C3 = "".
'affaire' has multiple occurences within my textfile, but I'm only interested in the first one. How to do so ?

Any help would be greatly appreciated,

Gwendal

2 个评论
显示无隐藏无

Stephen23 2022-3-28

@Gwendal Marrec: please upload a sample file by clicking the paperclip button.

Gwendal Marrec 2022-3-28

I've uploaded a copy of the text file I'm working on as well as the lines of code I'm trying to use.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Stephen23 2022-3-28

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1682529-is-it-possible-to-use-two-textscan-statements-in-a-row#answer_929079

编辑：Stephen23 2022-3-28

在 MATLAB Online 中打开

textscan_error.txt

"Is it possible to use two textscan statements in a row ?"

Yes, you can call TEXTSCAN as many times as you want on one file, I have often used this to read blocks of data separated by intermediate "header" lines. Quite handy, but probably not the best approach for your data file.

str = fileread('textscan_error.txt');
tmp = regexpi(str,'(?<=INFORMATION\s*\{\s*)[^\}]+','once','match');
tkn = regexp(tmp,'(\w+)\s*:\s*"([^"]*)','tokens');
tkn = vertcat(tkn{:}).';
out = struct(tkn{:})
out = struct with fields:
          date: 'Mars 2022'
         heure: '16:17:50'
           nom: ''
       affaire: 'stage'
    protection: '_'
out.heure
ans = '16:17:50'
out.affaire
ans = 'stage'

If you really want to use TEXTSCAN:

opt = {};
fmt = '%s:%q';
[fid,msg] = fopen('textscan_error.txt','rt');
assert(fid>0,msg)
str = 'X';
while ~startsWith(str,'INFORMATION')
    str = fgetl(fid);
end
tmp = textscan(fid,fmt,opt{:});
fclose(fid);
tmp{:}
ans = 6×1 cell array
    {'date'      }
    {'heure'     }
    {'nom'       }
    {'affaire'   }
    {'protection'}
    {'}'         }
ans = 5×1 cell array
    {'Mars 2022'}
    {'16:17:50' }
    {0×0 char   }
    {'stage'    }
    {'_'        }