Append date to time column entries where time is sporadically placed throughout the table

My data looks something like this:

time: service: user: otherdata:

09:00:00 @SERVICE first.last@machineName somedata

09:00:03 @SERVICE TIMESTAMP 12/31/2019

09:05:00 @SERVICE first.last@machineName somedata

09:08:00 @SERVICE first.last@machineName somedata

What I would like to do is append that 12/31/2019 to the time column, but the problem is that it's sporadically placed throughout the file. The main consistancy is the TIMESTAMP data that's in the user column, but I'm not exactly sure how to format the parsing to find it.

Any help would be greatly apprecaited.

[Mohammad Sami]

If the format of the row with date is alway the same, you can use that to find the rows which contain date.

index = contains(stringdata,'@SERVICE TIMESTAMP');
rows_with_timestamp = stringdata(index);
% you can then parse using regexp textscan e.t.c

[Tyler Rohren]

I'm more concerned with appending that data to the time entry and then changing to the next one when the timestamp updates to a newer date, and idea on how to do that? The parsing it's that big of a deal.

Answer 1

[Cris LaPierre]

Identifying a "best" solution will require some creativity and some knowledge of your data. Here's one simple approach that may get you started.

data = readtable('TylerRohren_data.txt')

ind = find(string(data.user_)=="TIMESTAMP",1,"first");
TS = datetime(data.otherdata_(ind));
for r = 1:height(data)
    if string(data.user_(r))=="TIMESTAMP"
        TS = datetime(data.otherdata_(r),'InputFormat',"MM/dd/uuuu");
    end
    time(r,1) = TS + data.time_(r);
end
data.time_ = time;

[Tyler Rohren]

Thanks! That was immensely helpful.