FWD » Core Development » Conversion Tools

05/31/2017

07:50 PM Feature #3277: implement call-graph v3

Greg, see revision 11228 - there was another problem, between first 2 contains levels no edges were being added.
B... Constantin Asofiei

06:58 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> These should not be using the "contains" edges. We have already traversed the "contains" up on l... Constantin Asofiei

06:44 PM Feature #3277: implement call-graph v3

The changes in 11227 are not quite correct.... Greg Shah

06:32 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> One other thing I need help with: I need a simple and generic way to get a descriptive text strin... Constantin Asofiei

06:30 PM Feature #3277: implement call-graph v3

One other thing I need help with: I need a simple and generic way to get a descriptive text string for each node. If... Greg Shah

06:29 PM Feature #3277: implement call-graph v3

Greg, please see revision 11227 - this shows data into the D3 web graph, but it doesn't look OK; not sure what is wro... Constantin Asofiei

06:26 PM Feature #3277: implement call-graph v3

> you are assuming that node-id is unique for the entire callgraph? This is true only in the same resource-domain: i.... Greg Shah

05:58 PM Feature #3277: implement call-graph v3

Greg, beside some other JS errors, I can get data into the web graph. But, I don't understand something: what kind o... Constantin Asofiei

05:39 PM Feature #3277: implement call-graph v3

And something else; in this code:... Constantin Asofiei

05:24 PM Feature #3277: implement call-graph v3

Greg, and the root cause why the graph isn't loading: in @createSnippet@, you are using @node-id@. Instead, you can ... Constantin Asofiei

05:18 PM Feature #3277: implement call-graph v3

Greg, I see two issues:
# when processing BOTH directions, you need to make sure a vertex is processed only once - o... Constantin Asofiei

04:57 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> The problem is that the @CGH.loadFileNodes()@ seems to return the @EXTERNAL_PROCEDURE@ node, not ... Constantin Asofiei

04:31 PM Feature #3277: implement call-graph v3

I'm testing a first pass at the @CallGraphApi.readGraph(String[] files, int direction, int levels)@ API call. It suc... Greg Shah

04:24 PM Feature #3277: implement call-graph v3

The approach makes sense. Greg Shah

04:03 PM Feature #3277: implement call-graph v3

For any kind of callgraph report, I'm planning to send these details about it, to the function which will build it:
... Constantin Asofiei

07:04 AM Feature #3277: implement call-graph v3

When you have the reports fixed and available in the API, you can work on the client side up call and display of the ... Greg Shah

06:56 AM Feature #3277: implement call-graph v3

1514a rev 11220 has support for these items:
* address the initial layout, positioning bias and collision avoidanc... Greg Shah

06:52 AM Feature #3277: implement call-graph v3

>>> Constantin wrote:
>>>
>>> Currently I'm almost finished optimizing the ambiguous sites for Hotel GUI, so that A... Greg Shah

06:50 AM Feature #3277: implement call-graph v3

>> Greg wrote:
>> I was thinking that your next step would be to implement an API call for each of the reports and w... Greg Shah

05/30/2017

12:31 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> It does and we will be using it. Anything that would be moving off the displayed graph will be p... Constantin Asofiei

11:56 AM Feature #3277: implement call-graph v3

> Can't we make this configurable in the UI?
Yes, that is the idea.
> Doesn't D3 allow a way to 'insert' dynami... Greg Shah

11:43 AM Feature #3277: implement call-graph v3

Greg Shah wrote:
> * see relationships, patterns and connections of the application that would be much harder to und... Constantin Asofiei

05/29/2017

01:13 PM Feature #3277: implement call-graph v3

I guess we need a way to get the list of the file set names.
Then for a given file set name (or ID of the file set... Greg Shah

12:26 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> the user should be able to select a different file set to load (Eric has an implementation of thi... Eric Faulhaber

11:31 AM Feature #3277: implement call-graph v3

>> rework the current reports, but as dynamically calculated results that we will send down to the browser for displa... Greg Shah

12:58 PM Feature #1514: implement fully dynamic reports (backed by the database)

Greg Shah wrote:
> The following is a description of the transport and API approach that we are using. Eric, please... Eric Faulhaber

11:13 AM Feature #1514: implement fully dynamic reports (backed by the database)

The following is a description of the transport and API approach that we are using. Eric, please correct or add to t... Greg Shah

05/28/2017

02:07 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> > if @ON ... PERSISTENT RUN@ is a call-site (as in #3277-41), then why is not a @SUBSCRIBE@ a cal... Constantin Asofiei

05/26/2017

07:10 PM Feature #3277: implement call-graph v3

> if @ON ... PERSISTENT RUN@ is a call-site (as in #3277-41), then why is not a @SUBSCRIBE@ a call-site, too? Because... Greg Shah

06:41 PM Feature #3277: implement call-graph v3

Current status (rev 11204):
!call_graph_visualization_branch_1514a_rev_11204_20170526.png!
* link z-order is hi... Greg Shah

06:00 PM Feature #3277: implement call-graph v3

Greg, regarding events/triggers, can you help me with something: if @ON ... PERSISTENT RUN@ is a call-site (as in #32... Constantin Asofiei

05/24/2017

12:30 PM Feature #3277: implement call-graph v3

The following is a really great resource for understanding zooming/panning and transforms in d3:
https://stackover... Greg Shah

10:47 AM Feature #3277: implement call-graph v3

Greg, I think everything is there for the OO call-sites (except events/triggers at class-level, which I haven't check... Constantin Asofiei

07:14 AM Feature #3277: implement call-graph v3

Greg Shah wrote:
> > create an edge to CLASS_DEF
>
> This one.
OK
> > the s:ToString() will point to the Sta... Constantin Asofiei

07:05 AM Feature #3277: implement call-graph v3

> create an edge to CLASS_DEF
This one.
> the s:ToString() will point to the Stateful implementation, and not A... Greg Shah

03:35 AM Feature #3277: implement call-graph v3

Greg Shah wrote:
> > At least the implicit c'tor should be referenced by the call-site. Or I can just link to CLASS... Constantin Asofiei

05/23/2017

06:14 PM Feature #3277: implement call-graph v3

> At least the implicit c'tor should be referenced by the call-site. Or I can just link to CLASS_DEF vertex?
I th... Greg Shah

02:11 PM Feature #3277: implement call-graph v3

Greg, some questions about OO:
* when an implicit constructor or destructor is called (as there is no actual impleme... Constantin Asofiei

07:40 AM Feature #3277: implement call-graph v3

Greg Shah wrote:
> In @SymbolResolver@, you can set breakpoints on access to @ROOT_OBJ_NAME@ to see how the @Progres... Constantin Asofiei

05/22/2017

07:15 PM Feature #3277: implement call-graph v3

Something is definitely wrong there. All of these parsed the last time I tried it. The implicit parent class was wo... Greg Shah

05:27 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> Check out the latest @cfg/class_map.xml@. See if that makes a difference.
No, it doesn't work. ... Constantin Asofiei

05:20 PM Feature #3277: implement call-graph v3

Check out the latest @cfg/class_map.xml@. See if that makes a difference. Greg Shah

05:11 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> From the project root, you should have a @p2j/skeleton/@ directory.
Tried your idea, also I've s... Constantin Asofiei

04:02 PM Feature #3277: implement call-graph v3

From the project root, you should have a @p2j/skeleton/@ directory. Greg Shah

03:12 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> > But aren't these supposed to be loaded into SymbolResolver$WorkArea.classDict
>
> The actual... Constantin Asofiei

03:06 PM Feature #3277: implement call-graph v3

> But aren't these supposed to be loaded into SymbolResolver$WorkArea.classDict
The actual skeleton classes need t... Greg Shah

03:00 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> See "OO Skeleton Classes":https://proj.goldencode.com/projects/p2j/wiki/Downloads#Object-Oriented... Constantin Asofiei

02:11 PM Feature #3277: implement call-graph v3

> Why there are two RAISES edges? One is from the subscriber (which expects an EVENT to be published by that external... Greg Shah

02:02 PM Feature #3277: implement call-graph v3

See "OO Skeleton Classes":https://proj.goldencode.com/projects/p2j/wiki/Downloads#Object-Oriented-Skeleton-Classes Greg Shah

01:10 PM Feature #3277: implement call-graph v3

Greg, any idea if the @testcases/uast/oo@ tests will parse in FWD? Because I'm getting some weird parsing errors... Constantin Asofiei

12:48 PM Feature #3277: implement call-graph v3

Greg, about PUBLISH/SUBSCRIBE. The events are raised by the "external procedure", and not by internal procedure or f... Constantin Asofiei

10:13 AM Feature #3277: implement call-graph v3

This is a summary/grouping of #2251-7
What is marked with an @X@ means is implemented in current 1514a branch.
h2... Constantin Asofiei

08:17 AM Feature #3277: implement call-graph v3

As of branch 1514a rev 11188, there is a good starting point for the call graph visualization. I have implemented th... Greg Shah

05/17/2017

11:07 AM Feature #3277: implement call-graph v3

> I assume we don't want to add to the graph the entire schema, correct?
Yes.
> Also, should I fix the existing... Greg Shah

09:42 AM Feature #3277: implement call-graph v3

Greg, a quick status:
# func/proc calls, super and dynamic calls are added (including 'references' via hints for 'vi... Constantin Asofiei

05/16/2017

02:31 PM Feature #3277: implement call-graph v3

> There is some redundant information in create_edge:
> ...
> which is used by current reports, but we can derive t... Greg Shah

02:14 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> > If there is no local definition, the target will be either a super definition or an external pr... Constantin Asofiei

02:04 PM Feature #3277: implement call-graph v3

> A secondary issue: for defined procedures or functions, which are 'not local' (i.e. IN SUPER, IN handle or EXTERNAL... Greg Shah

05/15/2017

04:55 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> Constantin, please rework and extend the run_statements.rules, for each call-site:
> * establis... Constantin Asofiei

11:09 AM Feature #3277: implement call-graph v3

Constantin Asofiei wrote:
> - create @calls@ links to all internal procs/funcs with the same name
This is @referenc... Constantin Asofiei

11:08 AM Feature #3277: implement call-graph v3

Eric Faulhaber wrote:
> Constantin Asofiei wrote:
> > Greg, a quick question: the callgraph needs to be built on th... Constantin Asofiei

10:47 AM Feature #3277: implement call-graph v3

Constantin Asofiei wrote:
> Greg, a quick question: the callgraph needs to be built on the F2's ASTs, correct? And n... Eric Faulhaber

08:37 AM Feature #3277: implement call-graph v3

Greg, a quick question: the callgraph needs to be built on the F2's ASTs, correct? And not the full conversion's modi... Constantin Asofiei

05/12/2017

05:45 PM Feature #3277: implement call-graph v3

Just a heads up: the @Element.property(String)@ getter has changed, it returns a @Property@ instance - I'll add a @cg... Constantin Asofiei

05/11/2017

07:34 AM Feature #3277: implement call-graph v3

> Do we really need in Janus to map the key as String? Because if instead of bld.addKey(prop, Mapping.STRING.asParame... Greg Shah

06:33 AM Feature #3277: implement call-graph v3

Greg Shah wrote:
> > About the index issues: this code in JanusGraphDB.createIndexImpl needs to be changed, as compo... Constantin Asofiei

06:16 AM Feature #3277: implement call-graph v3

> About the index issues: this code in JanusGraphDB.createIndexImpl needs to be changed, as composite indexes don't a... Greg Shah

06:15 AM Feature #3277: implement call-graph v3

Constantin, please rework and extend the run_statements.rules, for each call-site:
* establish a "contains" edge f... Greg Shah

06:07 AM Feature #3277: implement call-graph v3

About the index issues: this code in @JanusGraphDB.createIndexImpl@ needs to be changed, as composite indexes don't a... Constantin Asofiei

05/09/2017

05:56 PM Feature #3277: implement call-graph v3

I've just checked in branch 1514a revision 11161 which is a shift to JanusGraph 0.2.0 and Tinkerpop 3.2.2 plus all th... Greg Shah

04:46 PM Feature #3277: implement call-graph v3

>> add all entry points as full vertices in the database, with enough information to associate them back to the origi... Greg Shah

04:22 PM Feature #3277: implement call-graph v3

From Greg:
> I need help with the expansion and rework of the core graph itself. The biggest areas are:
>
> ad... Constantin Asofiei

01:20 PM Feature #3277: implement call-graph v3

Attached is the JanusGraph 0.2.0 javadoc which is not available on the Internet (at least as known by Google and by t... Greg Shah

05/05/2017

12:55 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> As far as I can tell, the javadoc parameter is probably just a bit misleading. Am I missing some... Constantin Asofiei

10:38 AM Feature #3277: implement call-graph v3

In @JanusGraphDB.createIndex*Impl()@, the @text@ parameter is described as this:... Greg Shah

05/02/2017

05:17 PM Feature #3277: implement call-graph v3

Greg Shah wrote:
> In @link_to_include@ or @print_call_site@ we see this code: @cgw.edgeIterator(out, "include-locat... Constantin Asofiei

04:21 PM Feature #3277: implement call-graph v3

In @link_to_include@ or @print_call_site@ we see this code: @cgw.edgeIterator(out, "include-location", "call-site-id"... Greg Shah

04:22 PM Feature #3225: windows vs linux in p2j.cfg.xml

> use progress-style separators, this way being OS neutral
This seems to make good sense. Greg Shah