Re: [h5md-user] graphical notation

[Pierre de Buyl]

Le 30 août 11 à 15:57, Felix Höfling a écrit :
> > Looking into "more usual", I can think of keeping the present notation  
> > for groups and datasets. Altough it does not make a distinction between  
> > the two, it has never posed a problem to me.
> OK, a dataset can be identified as a leaf of the tree (once the  
> attributes are distinguished).
Fine :-)

> > Regarding attributes, I agree that they should be denoted in a  
> > different way.
> > I propose
> >      trajectory
> >       \-- group1
> >            \-- position
> >            |    \-- sample
> >            |    |    +-- minimum
> >            |    |    +-- maximum
> >            |    \-- step
> >            |    \-- time
> > Or
> >      trajectory
> >       \-- group1
> >            \-- position
> >            |    \-- sample
> >            |    |    --- minimum
> >            |    |    --- maximum
> >            |    \-- step
> >            |    \-- time
> > that is, using "+--" or "---" for the attributes.
> I prefer "+--", it provides still some connection to the line above. An  
> alternative I was thinking about was to put the attributes at the same  
> line as the dataset:
>
>       trajectory
>        \-- group1
>             \-- position
>             |    \-- sample --> minimum, maximum
>             |    \-- step
>             |    \-- time
The attributes on the same line makes it less readable, the "+--" is
adopted.

> Shall optional attributes be enclosed in brackets? E.g.,
>
>       trajectory
>        \-- group1
>             \-- position
>             |    \-- sample
>             |    |    +-- [minimum]
>             |    |    +-- [maximum]
>             |    \-- step
>             |    \-- time
Excellent idea, it is a quite common notation.

Pierre