Hi Charlie,<br><br><div class="gmail_quote">On Mon, Sep 27, 2010 at 8:01 PM, Charlie Brej <span dir="ltr"><<a href="mailto:plymouth@brej.org">plymouth@brej.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"> Sorry about the delay. I'll try and answer as many questions aimed for me as I find, but I will probably have missed some.</blockquote><div><br></div><div>No problem. </div><div><br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"> <div class="im"><br> <br> On 09/25/2010 11:50 AM, Jerome Martin wrote:<br> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> - Is there a way to call up an external binary from plymouth, like a<br> system() function ? If not, I guess this would not be much to add, I can<br> do it if you give me a proper pointer in your code (hints on the proper<br> parse token to use, etc.).<br> </blockquote> <br></div> Yes possible, although I would recommend using the on keystroke execute command feature rather than letting a plugin run anything it likes. At least that way you can be sure what state the system is in.</blockquote><div> <br></div><div>Well, this will be under control :-)</div><div>I just need a small menu (not a problem to implement) that asks a couple of questions and pass the answers to a couple of config scripts. The system is a cluster node that boots from a live system image, reads a persistent config partition and autoconfigures itself. But for seeding the cluster, or in corner cases, I need some interaction with the user to help seed the config, and I want to use plymouth for that.</div> <div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div class="im"><br> <br> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> - I desperately miss a couple of possibilities regarding strings, like<br> determining a string length, and addressing strings as lists<br> (a="foobar"; c=a[2]), thus making it possible to do strings<br> transformations, iterate on strings, etc. Did I miss something ? Would<br> that be hard to implement ?<br> </blockquote> <br></div> Ah! So there are two functions that do this. "CharAt" and "SubString". Somehow, I forgot to add these to the wiki.<br> <br> "String".CharAt(2) => "r"<br> "String".SubString(2, 4) => "rin"<br> <br> I think it is indexed starting at zero and the SubString takes inclusive start and end pointers.</blockquote><div><br></div><div>Yes, meanwhile, as I said in a later email, I noticed those in the code. I will add a String.Len() method for my needs an be all set.</div> <div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div class="im"><br> <br> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> - Last, but related to the previous, I currently keep an index on lists<br> I create to knopw their length before iterating on them, but it would be<br> nice to have a mylist.Len() method.<br> </blockquote> <br></div> Are you just appending to a list by adding elements at increasing indices? list[0] = "blah", list[1]="sdf"...<br> In which case you can just add a length element and keep that up to date when you add/remove elements. (list.length++ or list["length"]++)</blockquote><div><br></div><div>This is what I do right now, I do the housekeeping in my script for an separate length variable. But now that I have added a couple of additional methods to plymouth script modu;e, I know how trivial it is to add that kind of thing, so I'll just do it when I do not feel to lazy/when the need is immediate.</div> <div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div class="im"><br> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> Can any of the two previous points be done with a "list | []" method of<br> sorts ?<br> </blockquote> <br></div> Not really. That's a rather complex one to explain.<br></blockquote><div><br></div><div><nod> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"> <br> > Things using "\033"<br> <br> There is no escaping supported in the scripted system apart from \n \" and \0.<br> <br> Ok, sorry, just found your patch. I'm split between \123 and \x123.<br></blockquote><div><br></div><div>Mhhh, I hesitated too but settled for C-like behavior.</div><div>We can make it nicer and easily support at the same time \0, \XYZ with XYZ in octal, even \XY if we add a lookahead function for the n next chars that do not shift the current char pointer.</div> <div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"> <br> Are there other questions I have missed?</blockquote><div><br></div><div>I think that's all, actually I solved most of it meanwhile :-)</div><div>The code is pretty clear when you get into it. It only lacks a bit of user docs.</div> <div><br></div><div>What do you think about having a switch to hand ctrl-V, esc, etc either to the script or to the main.c code depending on the user needs ? that would allow me to submit something that could go in your git repo for my ANSI key handling logic. I could even add some examples from my code, like a log viewer that supports scrolling with up/down, things like that.</div> <div><br></div><div>Best Regards,</div></div>-- <br>Jérôme Martin<br>