Re: Filter by Selection (since v9.5)

On 21-Sep-2015 09:52 -0600, Buck Calabro wrote:

On 9/21/2015 10:19 AM, Edmund Reinhardt wrote:

But I wanted to tell you that there another way of accomplishing
search for all occurrences of a variable. On the toolbox above the
editors there is a highlighter looking button. The hover says
"Toggles marking occurrences of the given entity" <ed: TMOOTGE>.
If you toggle this on, and then double click on the name of
interest to select it, then all references and declarations will
be highlighted throughout the source.

This is pretty cool but I haven't found it to be 100% reliable, in
that there are instances of a procedure which are not identified. One
such example is
dcl-s detailProc_p pointer(*proc) inz(%paddr(detailproc));
This doesn't show up in the Outline View either, so I'm not surprised
that TMOOTGE doesn't identify it. [1]

The right hand side of the editor will have rectangles
corresponding to each occurrence, allowing you to quickly navigate
to the them. The advantage is that it understands RPG and will not
be fooled by similar spelling. The color if the highlight can be
customized under Preferences>General>Editors>Text
Editor>Annotations Occurrences
(org.eclipse.imp.occurrenceAnnotation)

Thanks for the colour tip - I found the default colour practically
invisible on the right side.

--
--buck
[1] The compiler finds this line, but RDi 9.5 and predecessors does
not.

001600 dcl-s detailProc_p pointer(*proc) inz(%paddr(detailproc));
...
002500 dcl-proc detailProc;
002600 dcl-pi detailProc;

DETAILPROC PROTOTYPE 001600 002500 002600

Interestingly, TMOOTGE finds the PR of 'detailProc' inside a /copy,
but it flags the wrong /copy... Outline View doesn't seem to notice
what's inside the /copy. Not yet sure what I think of that.

Bottom line: I still trust the compiler listing more than Outline
View

FWiW, per the explanation of the note [1] having been placed in the message following the signature delimiter, the text was cropped in the archive; the text has since been included in the above message... so if the drop was intentional, the text will now be there in the archives, despite that intent.