Hi Romain ! Le 15/06/2011 15:28, Romain Beauxis a écrit : > Hi Martin! > > 2011/6/15 Martin Hamant<[email protected]>: >> Le 15/06/2011 10:48, David Baelde a écrit : >>> Hi Martin, >>> >>> Using your terminology to clarify, you could write: >>> >>> On Wed, Jun 15, 2011 at 10:37 AM, Martin Hamant<[email protected]> wrote: >>>> out = output.icecast( >>>> fallible=true, >>>> start=false >>>> ) >>> out = >>>> if stream_format == "mp3" then >>>> out(%mp3) >>>> end >>> # here you also need the else branch to say what out becomes if >>> stream_format != "mp3". >> Yes, as I am using mp3 for testing purpose, I haven't wrote further. >> >> >>> Perhaps one way to clarify things is to avoid calling "outputs" a >>> "global variable" that is "initialized". A definition has only one >>> value over the course of its life, it never changes. You may hide it >>> in some part of the script with another definition, but this is really >>> different: >>> x = 1 >>> if true then >>> x = 2 >>> # here x is 2 >>> print(x) >>> end >>> # x is 1 again >>> print(x) >>> # in the if-then-block, you could as well replace all occurrences of "x" >>> by "y" >>> >>> Your "dyn_sources" is closer to what is usually called a (global) >>> variable. It has an initial value (dyn_source = ref<initial value>) >>> but its value can change over time (dyn_source :=<new value>). >>> >>> I know these differences can be confusing. I hope that my explanations >>> help you see the difference, which should be useful when writing >>> scripts. >>> >> Gasp this is very hard to me. I understand the life of x in the little snip >> you provided. >> I think the confusing is about the partial construction. >> here, we speak about a definition "out = icecast.output()", that gives a >> function out() which is precisely designed to deffer some parameters >> definition. (isn't it ?). Which is called "partial application". >> >> I am sorry, you are certainly going to repeat yourself :( >> >> Here is the full first part of my script, I have annotated examples which >> works and some don't: >> >> ========== >> # First, we create a list referencing the dynamic sources: >> dyn_sources = ref [] >> >> # This is our icecast output. >> # It is a partial application: the source needs to be given! >> out = output.icecast( >> fallible=true, >> start=false >> ) >> >> # Now we write a function to create >> # a playlist source and output it. >> def create_playlist(uri) = >> # The playlist source >> >> param = string.split(separator="\|",uri) >> source_device = list.nth(param,0) >> stream_host = list.nth(param,1) >> stream_port = list.nth(param,2) >> stream_password = list.nth(param,3) >> stream_id = list.nth(param,4) >> stream_genre = list.nth(param,5) >> stream_url = list.nth(param,6) >> stream_description = list.nth(param,7) >> #stream_format = list.nth(param,8) >> >> s = input.alsa(device="#{source_device}") >> >> # metadata telnet handler >> s = server.insert_metadata(id="#{stream_id}-meta", s) >> >> >> >> # the following doesn't work. >> # for testing purpose >> #stream_format = "mp3" >> >> #out = >> #if stream_format == "mp3" then >> # out(%mp3) >> #end >> >> # the following works but hey why ? >> >> out = >> out(%mp3) >> >> #The following don't work >> >> #out(%mp3) > I think you can try to see this in terms of types. Here's a simplified > presentation. At first: > out = output.icecast() > will be of type format -> source -> source, i.e. its a partial > application which needs at least a format and a source. Is it right to say, output.icecast() take a format and a source as args, "returning" a new source ? So what about the type (format, source) => source I have big problems reading types... it looks very complicated to me. I understand there is no "this function returns something" concept, but something more like "this function gives the populated/modified source"
> Then you do: > out = output.icecast(%mp3) > Now out is of type source -> source, i.e. you have applied one more > argument, thus the type has changed. Here I understand that out has now its format, so its new type says that it wait for what remains to work: the source, and will provide a new source. ? > But if you do: > out(%mp3) > Then the variable out is still of type format -> source -> source > because the above was not kept.. can you define "not kept" (from ? why ?) > Thus, if later you want to do: > out(source) > Then this will only work if out is of type source -> source, i.e. in > the first case above.. This is not really clear but I swear I am trying... My abstractive brain needs time so I am going to give it a break :) I still don't understand how I can update a source in a "if" iteration... I think that the better would be that you provide a working snippet, and I'll try to understand what's happening why it works. :( ------------------------------------------------------------------------------ EditLive Enterprise is the world's most technically advanced content authoring tool. Experience the power of Track Changes, Inline Image Editing and ensure content is compliant with Accessibility Checking. http://p.sf.net/sfu/ephox-dev2dev _______________________________________________ Savonet-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/savonet-users
