You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@taverna.apache.org by Gale Naylor <Ga...@noventussolutions.com> on 2016/05/26 23:47:02 UTC
Edited Command-line Tool Web Pages
Hello, all.
I edited the Command-line Tool web pages (to fix the tool name, etc.) and
they are staged in the CMS.
Please review:
(A) Download page: http://taverna.staging.apache.org/download/commandline/
(B) Documentation page:
http://taverna.staging.apache.org/documentation/command-line/
I added sections to the documentation page and did some significant editing
for grammar and style. I also have a few questions/issues on the
documentation page. In particular:
(1) "Important: The database name in the jdbcuri property must be
T2Provenance, as this database name is hard-coded into the MySQL provenance
SQL queries." (I presume this is different now?)
(2) Part of the section on creating lists is confusing: "If a list of
greater depth is required (i.e., a list or lists or deeper)" I presume this
is referring to the indentation level of the list, but I don't understand
"a list or lists or deeper." Can someone give me an example of what a lists
might look like, and what a wrapped list would look like?
(3) Is "extended version" relevant for 3.x? - this link goes to
http://dev.mygrid.org.uk/wiki/display/tav250/Extended+version
(4) Is everything on this page current for Taverna 3? There are a few
references to taverna-cmdline-2.x.x or T2.
Thanks!
Gale
Re: Edited Command-line Tool Web Pages
Posted by Stian Soiland-Reyes <st...@apache.org>.
eOn 27 May 2016 at 00:47, Gale Naylor <Ga...@noventussolutions.com> wrote:
> Hello, all.
>
> I edited the Command-line Tool web pages (to fix the tool name, etc.) and
> they are staged in the CMS.
Thanks!
> Please review:
> (A) Download page: http://taverna.staging.apache.org/download/commandline/
> (B) Documentation page:
> http://taverna.staging.apache.org/documentation/command-line/
>
> I added sections to the documentation page and did some significant editing
> for grammar and style. I also have a few questions/issues on the
> documentation page. In particular:
>
> (1) "Important: The database name in the jdbcuri property must be
> T2Provenance, as this database name is hard-coded into the MySQL provenance
> SQL queries." (I presume this is different now?)
You are right - I think we can remove or comment out the sections
about the database as it is effectively not used by this particular
release. (That is, the Hibernate-based reference implementation was
removed due to licensing issues - and the provenance layer is captured
differently in Taverna 3 without a database - so setting up a database
probably has no effect on the running of the workflow) .
> (2) Part of the section on creating lists is confusing: "If a list of
> greater depth is required (i.e., a list or lists or deeper)" I presume this
> is referring to the indentation level of the list, but I don't understand
> "a list or lists or deeper." Can someone give me an example of what a lists
> might look like, and what a wrapped list would look like?
It IS confusing..
http://dev.mygrid.org.uk/wiki/display/tav250/List+depth
explains it a bit. Basically - if we pretend some JSON:
depth 0: "value"
depth 1: ["value1", "value2", "value3"]
depth 2: [ ["value1", "value2",], ["value3", "value4"], ["value5", "value6"] ]
depth 3: [ [ [ "value1" ], ... ], .. ]
Lists in Taverna must have consistent depth - so all children of depth
2 must be of depth 1 (or be empty), etc.
Obviously providing such nested lists on the command line is a bit
tricky - the -inputdelimiter parameter lets you split a particular
input by a specified character (typically , or newline) - supporting
depth 2 or above can't be done that way.
Also I realize now we expose the -inputdoc option - but it does not
seem to be handled by the InputsHandler (only -inputfile and
-inputvalue ) -- we have not conversion from the -inputdoc baclava
documents to the Data Bundle. It should however be possible to do a
-inputbundle and then provide a list of lists as a zipped folder
structure. It should be in Jira, if not I'll add it.
Support for a JSON input document could also be useful. New feature we
can plan for 3.2.0?
Perhaps I need to update the -help text before the release so it
doesn't include the database and -inputdoc thing - I'll raise that in
Jira as well.
> (3) Is "extended version" relevant for 3.x? - this link goes to
> http://dev.mygrid.org.uk/wiki/display/tav250/Extended+version
No longer relevant - as the Interaction activity and support for those
properties are included - however while component is not yet included
it's a separate plugin
https://github.com/apache/incubator-taverna-plugin-component
So perhaps promote up the taverna.interaction.* properties in the
documentation and remove references to "Extended Version".
> (4) Is everything on this page current for Taverna 3? There are a few
> references to taverna-cmdline-2.x.x or T2.
No, for instance the mentions of "the bundled tool" is no longer
relevant, as I don't think we'll include the command line in the
Taverna 3 workbench distribution (for the same reasons as listed on
that page :)
--
Stian Soiland-Reyes
Apache Taverna (incubating), Apache Commons
http://orcid.org/0000-0001-9842-9718