speedo

I have a jit.uldl object downloading xml from a webpage. I have tried connecting it both directly and through jit.textfile to detox but detox is not responding. Anyone have a working example of how I can get this to function? I basically need to read an rss/xml feed and have it parse various values.

detox-parsing-what-kind-of-message-is-expected-at-input

Try using the "line" message to [jit.textfile] then send this to [detox]. It can't read a whole xml file, so you need to send it line by line. See the help file for an example.

Thanks Luke, that helped a lot and now I have (what I believe to be) properly formatted content in detox's input.

I'm getting some error messages with detox in the Max window and it doesn't seem to be formatting the lines correctly.

Do I have to specify somewhere which tags I want it to read? As you can see from the below example it can read a few of the lines but it isn't formatted correctly. This happens with any xml source. Any thoughts? Am I missing a step?

I'm getting some error messages with detox in the Max window and it doesn't seem to be formatting the lines correctly. 

Do I have to specify somewhere which tags I want it to read? As you can see from the below example it can read a few of the lines but it isn't formatted correctly. This happens with any xml source. Any thoughts? Am I missing a step?


I think the problem is the formatting of the xml. The [detox] object expects very strictly formed code whereas page you're downloading doesn't have each tag start a new line which is a problem. It might be more convenient for you to use [regexp] but this is a major headache if you've not encountered it before. I'm not [regexp] pro but I've used it a fair bit before so maybe I can help with that instead. What are you looking for specifically?

Thanks for checking it out. While I can't divulge too much info on this project as I'm under an NDA with the parent company, I can give you some more specific examples of what I'm trying to accomplish.

It seems that most news feeds and such are equally formatted like this, so perhaps detox is not the way to go. I basically need to use one and sort out common variables from the various tags. So, for instance fruitorange35

Where max would give me "fruit, orange, 35" as a list or something I could work with. When the feed refreshes, it will give me 3 more variables to work with that would be passed on as a list.

Maybe I can find an application that will reformat the xml into more conventional syntax and I can read from that.

Where max would give me "fruit, orange, 35" as a list or something I could work with. When the feed refreshes, it will give me 3 more variables to work with that would be passed on as a list. 

Maybe I can find an application that will reformat the xml into more conventional syntax and I can read from that.


If I just print from jit.str.tosymbol I have a line that contains most of the information I need to extract

urn:world:p1003112T, Puerto Rico2010-01-31T20:46:59Z

Sunday, January 31, 2010 20:46:59 UTCSunday, January 31, 2010 04:46:59 PM

In this case, I would want to have a message or text buffer with (T, Puerto Rico, 2010-01-31T20:46:59Z, Distance 40mi)

urn:world:p1003112T, Puerto Rico2010-01-31T20:46:59Z


If you're searching for something specific then try [jit.str.regexp @re ([^] where you can replace the word "food" with the search term. You can always copy this multiple times in the same object if you want to match more things in a row. For instance your example above would use [jit.str.regexp @re ([^([^([^] which is a bit of a handful but will work. It might be easier to use the first outlet of [jit.textfile] and send it a "bang" to output the entire contents rather than iterating through a line at a time.

Still getting an error on [print] from [jit.str.regexp]

See below, using this forum's home page as the source and as the tag.

See below, using this forum's home page as the source and  as the tag.


Not sure if this is helpful to you, but I've done quite a bit of XML parsing work in Max/MSP/Jitter using the Java library JDOM.

I found it a lot more intuitive than doing any parsing in patcher land - one of the great advantages of having mxj available.

You had your [jit.str.tosymbol] and [jit.str.regexp] in the wrong order. You only want to take it out of jitter matrix form and into the max world right at the end before the [print] object.

Ha! Yes, I see that now. Working great! Thanks.

@JG - thanks for the link, I will check that out (although I can't imagine getting into JAVA right now) ;)

Here's an attempt at parsing the original file so that it has better formatted line breaks. If you're new to regular expressions then this isn't the place to start. There will probably be instances where this doesn't work and causes issues but it seems ok-ish for now. I'd suggest sticking to the other example. This is here just in case you are interested in trying to use [detox] to get the tag structure information.

Detox Parsing - What kind of message is expected at input?