convert a html table to csv format

I need to convert the table in the following url to csv format. Since I have to convert many tables, I can't use cope paste. http://climate.weatheroffice.gc.ca/climate_normals/results_e.html?stnID=2046&lang=e&dCode=0&province=ALTA&provBut=Search&month1=0&month2=12

[Matt Kindig]

Do you have to use Matlab for this purpose? The reason I ask is because other languages that are more commonly used for website development have good HTML parsing capabilities, whereas such features are more limited in Matlab--in Matlab you'd basically have to resort to complex regexp statements.

I would recommend Python and the BeautifulSoup package to do this, actually.

Answer 1

[Jan]

You can import the HTML table to Matlab at first by FEX: htmltableToCell or FEX: get-html-table-data-into-matlab. Then an export to CSV depends on the contents of the data.

Answer 2

[Cedric]

As Matt mentions, Python + package would be perfect for this part. Here is one way to do it using REGEXP in MATLAB.. not the full stuff though, but enough to illustrate.

 % - Get HTML page.
 url = 'http://climate.weatheroffice.gc.ca/climate_normals/results_e.html?stnID=2046&lang=e&dCode=0&province=ALTA&provBut=Search&month1=0&month2=12' ;
 buffer = urlread(url) ;

 % - Extract horizontal header.
 p = '(?<=<td class="dataTableColHeader">).*?(?=</td>)' ;
 hheader = regexp(buffer, p, 'match') ;

 % - Extract vertical header.
 p = '(?<=<td class="dataTableRowHeader">).*?(?=</td>)' ;
 vheader = regexp(buffer, p, 'match') ;

 % - Extract/reshape data.
 p = '(?<=<td class="dataTableRowData">).*?(?=</td>)' ;
 data = regexp(buffer, p, 'match') ;
 data = reshape(data, 12+2, []).' ;

 % - Build and export the whole.
 content = [vheader.',[hheader; data]] ;
 xlswrite('example.xlsx', content) ;

Let me know if you want to go this way and I can improve a little this code. There would be still quite a bit of work to do on your side, e.g. to manage some inconsistency in the way they build the HTML table, to detect/manage failures in the processing, to export to CSV instead of XLSX, etc.