Reading a string to get required data

Im trying to write a program where I can read HTML code for the purposes of extracting the data, for some analyse im conducting. Ive managed to remove the HTML jargon and am now left with a string which contains the data i require. However im trying to convert the data into a readable cell array;

Mar25,2011>4.88>4.88>4.83>4.88>51,000>Mar24,2011>4.72>4.72>4.72>4.72>13,300>Mar22,2011>4.88>4.88>4.88>4.88>0>Mar18,2011>5.00>5.00>5.00>5.00>0>Mar17,2011>4.81>4.89>4.81>4.89>1,001>

I know this may seem rather simple to most of you but im new to Matlab. Basically im trying to convert this string into a column array, firstly with the date followed by the sucessive five numbers for the whole data set. Any help on this would be greatly appreciated.

Answer 1

[Walter Roberson]

textscan('%s%f%f%f%f%f', 'Delimiter', '>', 'CollectOutput', 1)

You might need to change the shapes around afterwards. I am not clear on what you are envisioning for a "column array".

[Tom Pearce]

Basically I just want the data in a list (6 Columns wide) from which i can write to file and produce a graph. Ive tried textscan but keeps returning {0x1} [0x1] [0x1] [0x1] [0x1] [0x1]. Now i realise im along the right lines i will persivere. Thanks Very Much for your help.

[Walter Roberson]

Ah, you have commas in your fifth numeric field; that throws off parsing them as a number. Also I forgot to show the string field.

Let T be the string you have the line stored in. Then,

Q = textscan(T,'%s%f%f%f%f%s', 'Delimiter', '>');

Q{6} = str2double(regexprep(Q{6},',',''));

Answer 2

[Clemens]

Personally I don't remove the "html jargon" in such cases. I use regexps like:

table_lines = regexp(table,'<tr [^>]*>(.*?)</tr>','tokens');

table_line_entry = regexp(table_line,'<td [^>]*>(.*?)</td>','tokens');

This has the advantage that it keeps the structure information of original table.

Also you might run into problems if in a table cell is html code, or just a ">" sign.