Yoshinobu Kano wrote: > Hi, > > I am trying to embed our text mining system into Taverna workflow. > Since I myself is not a biologist in any meaning, > I am not sure how the text/document is handled in the Taverna (or > biological) community. > A couple of questions regarding the text handling issue. > > > In the Results tab, when Result Type is "Text" newlines are not shown. > Is there any way to display newlines as really newlines? > How about wrapping lines?
What version of Taverna are you running? In Taverna 1.7.1, 2.0 and 2.1 beta 1, I get newlines in output text. > Iteration Strategy. > It seems like that the iteration strategy is handled using Java > objects like ArrayList, > the workflow itself is not iterated. > Is this a correct understanding, or is there any way to iterate the > same workflow as a "batch" like way? In Taverna 1.7.1 you can feed a list into a workflow port that expects a single value. The workflow will iterate using the same rules as if you were passing lists into a service. So if the workflow has two ports that each expect a single value, and you pass [a,b,c] and [1,2], the workflow will iterate 6 times, once for a&1, a&2, b&1, b&2, c&1, c&2. In Taverna 2.0 (and 2.1 beta 1) you cannot feed a list into a workflow port if the port is expecting a single value. The input dialog will not allow you to specify multiple values. However, you can alter the depth of the workflow port to accept lists. That will often achieve what you want. You can still, of course, run it with a single value just by specifying only one element in the list. Katy Wolstencroft has raised workflow iteration as an issue that needs to be fixed. > How do you handle a large scale document set, e.g. the whole Pubmed > papers, to avoid the large memory consumption? The best way is to design the services so that they take URLs rather than the actual documents. For Taverna 1.7.1, you can use the data proxy mechanism http://www.taverna.org.uk/associated-tools/webservice-data-proxy/ In Taverna 2, the data handling is better and much improved in 2.1 beta. > If there is such a batch mode exists, how do I notice the end of the batch? In Taverna 1.7.1, the output from iterations will not appear until all the iterations have finished. In Taverna 2, the values are output as they are produced. In Taverna 2.1 beta 1, there is an explicit indication of when the workflow has finished. > Further, what is the most popular unit to handle text in this > community - sentence, document, word... ? I do not know. > Any help appreciated! > > Thanks, > > -Yoshinobu Alan ------------------------------------------------------------------------------ Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensing option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects _______________________________________________ taverna-hackers mailing list [email protected] Web site: http://www.taverna.org.uk Mailing lists: http://www.taverna.org.uk/taverna-mailing-lists/ Developers Guide: http://www.mygrid.org.uk/tools/developer-information
