Well, Freemarker, Velocity. Tapestry Offline all require some dev knowledge.
And I did not say you give your users a standard dataset, I said give them a
standard way to describe it like json or xml. Thus transformation is like:
array --> T5 loop (table, ol, ul)
elements --> div's or whatever
attributes --> maybe bootstrap col serrings aso
There are even json to html form conversion tools which do a great job like
jeremydorn.com/json-editor
Jens
But if freemarker solves it: fine!
Jens
Von meinem iPhone gesendet
> Am 01.03.2016 um 02:25 schrieb David Diaz <d...@viddiaz.com>:
>
> Hi Jens,
>
> Yes, not all users will have software development backgrounds (of course).
> However, our users will not be the ones developing these renderer
> classes/templates, and thus it's fine if doing this requires some SDLC and
> Tapestry knowledge. I know it sounds a bit outlandish, but requirements are
> requirements.
>
> It would be nice we could get our users to give us standardized data sets
> that matched the same data each time, but it's just not realistic in this
> case.
>
> Thiago's suggestion of Freemarker seems like it can do the job for me just
> fine, and although having two templating systems isn't ideal, it's a lot
> better than the current system that is in place.
>
> Thanks,
> David.
>
> On Tue, Mar 1, 2016 at 12:05 PM, mailingl...@j-b-s.de <mailingl...@j-b-s.de>
> wrote:
>
>> Hi David!
>>
>> Thanks for the explanation. Unfortunately I doubt your approach is the way
>> to go. You are forcing all your users to get familiar with Tapestry, so
>> they are all really aware of software development, java, Testing, the dev
>> pipeline and everything? Who makes sure the tml and fictious jave class
>> works smoothly together?
>>
>> I believe either they have to provide valid static html (no dynamic part,
>> but renderable by tapestry via outputraw). Or you provide base classes and
>> tmls and they send back derived version (tml / java inheritance, see T5
>> website), but again there must be some awareness for software dev and the
>> toolchain, which is not worth the effort as you still need to know all
>> dynamic parts in advance.
>>
>> What about accepting e.g xml / json and render the data in a generic way
>> on your side? Based an certain elements or attributes you should be able
>> even to render complex layouts and your users only need to know which
>> elements/attributes you support? Yes I know this sounds like the early
>> 80ies jsp tag libs or good old xml-dynamic-swing-ui systems-frameworks.
>>
>> Jens
>>
>> Von meinem iPhone gesendet
>>
>>> Am 01.03.2016 um 00:03 schrieb David Diaz <d...@viddiaz.com>:
>>>
>>> Hi all,
>>>
>>> Thanks for the replies! I checked out your link Barry - it doesn't seem
>> to
>>> be exactly what I want. (I'll explain later in the reply... I'm still new
>>> to mailing lists).
>>>
>>> Thiago & Jens:
>>> How I'm doing fetching-code-outside is someone writes a .java file that
>>> implements an interface. Say it has function a, b and c they have to
>>> implement. They then compile this into a class, put it into a program and
>>> link it to the set of data they want to render. Then, when using the web
>>> interface the application will detect when its loading the set of data
>> and
>>> load it into its own class (not overwriting), create a new instance and
>>> then use functions a, b and c to render the data into the webpage. This
>> is
>>> already working - but currently how it is done is that the user has to
>>> hardcode their HTML into the .class file and do things like \"<div>" +
>>> escapeHTML(data) + "</div>";\ instead of it having a template system
>> where
>>> they could go <div>${data}</div> and then the application would use the
>>> tapestry template/component system.
>>>
>>> An example usage for this would be, for example, a library that wants to
>>> digitize its pages and then have it available online to be viewed through
>>> the application. The library could store the files in an XML format like
>>> so: http://da.viddiaz.com/example.xml and then use the following
>>> hypothetical code and .tml to render it: http://da.viddiaz.com/code.txt
>>> http://da.viddiaz.com/code.tml
>>>
>>> Then this library could have different sets of data that wouldn't match
>>> this exact set of data - it might have sets of emails (so then you could
>>> write an email renderer), MARC data and so on. These examples are basic -
>>> in the real world the data is a lot more complex! I hope this makes a bit
>>> more sense why I would need to take this approach.
>>>
>>> In regards to the Dynamic component, I must have messed up using it.
>> Still
>>> though, the problem would be getting it linked up to the .class file that
>>> is loaded so that the user could then use tapestry in there (otherwise
>> the
>>> ComponentResources would be the .tml file that loaded it, not the
>>> dynamically loaded one).
>>>
>>> The stack trace I was getting is here, although I don't think it's really
>>> useful: http://da.viddiaz.com/stacktrace.txt
>>>
>>> Thanks,
>>> David.
>>>
>>> On Tue, Mar 1, 2016 at 9:09 AM, mailingl...@j-b-s.de <
>> mailingl...@j-b-s.de>
>>> wrote:
>>>
>>>> I see. So a tapestry TML without a JAVA class is not considerably
>> dynamic
>>>> as it has no business logic at all. Either your 3rd party delivers
>>>> TML+CLASS (only components from my point of view) than you can
>> integrate it
>>>> as module. But this is far away from "dynamic". If you provide the
>> logic,
>>>> the TML can not change anything nor add something new, so its about
>>>> rearanging HTML or changing the style, but this is not what you want
>>>> either, right? Because this might be a pure CSS related topic.
>>>>
>>>> Can you give us a real piece of code you get from your 3rd party for
>>>> integration? Maybe it's really an iframe, an external template engine
>> (to
>>>> render raw output) or something completely different like angular...
>>>> I am still confused about the requirements...
>>>>
>>>> Sorry
>>>>
>>>> Jens
>>>>
>>>> Von meinem iPhone gesendet
>>>>
>>>>> Am 29.02.2016 um 09:05 schrieb David Diaz <d...@viddiaz.com>:
>>>>>
>>>>> Hi Jens,
>>>>>
>>>>> Yes I have external pages that I need to integrate. And by code I mean
>>>> the
>>>>> actual .java page that would (hopefully) power the .tml.
>>>>>
>>>>> The problem I have is that I need to (obviously) render content in a
>>>>> specific way, but the application I am writing needs to load the
>>>> templates
>>>>> dynamically.
>>>>>
>>>>> Thanks,
>>>>> David.
>>>>>
>>>>> On Mon, Feb 29, 2016 at 6:37 PM, mailingl...@j-b-s.de <
>>>> mailingl...@j-b-s.de>
>>>>> wrote:
>>>>>
>>>>>> Hi David!
>>>>>>
>>>>>> What do you mean by "external code"? Are you talking about external
>>>>>> Tapestry pages/components and tml's you need to integrate? Because you
>>>>>> mentiond HTML not TML? Does "code" refer to Javascript? Is the
>> external
>>>>>> code self contained? What about an ugly but simple "iframe"?
>>>>>>
>>>>>> Jens
>>>>>>
>>>>>>
>>>>>> Von meinem iPhone gesendet
>>>>>>
>>>>>>> Am 29.02.2016 um 05:18 schrieb David Diaz <d...@viddiaz.com>:
>>>>>>>
>>>>>>> Hi all,
>>>>>>>
>>>>>>> I have been using Tapestry for a little bit now and I have run into
>> one
>>>>>>> problem that I can't figure out how to solve.
>>>>>>>
>>>>>>> In my application I need to be able to load external code & HTML to
>>>>>> display
>>>>>>> to the user from a trusted source at run-time. Currently I have
>> gotten
>>>>>> code
>>>>>>> to load fine and I am using outputraw to render the result from the
>>>> code.
>>>>>>> This is pretty bad though since I have to code the HTML within the
>> Java
>>>>>>> file instead of having it templated.
>>>>>>>
>>>>>>> I saw that Tapestry has a "Dynamic" component that allows loading
>>>>>> templates
>>>>>>> from an external source but I tried using it - it tries loading a
>> file
>>>>>> from
>>>>>>> the file system at page creation time... this is not suitable for my
>>>> use
>>>>>>> case since I need to load it from a String and I need to do it at
>>>> render
>>>>>>> time.
>>>>>>>
>>>>>>> I also tried messing around with DynamicTemplateParser & MarkupWriter
>>>>>> but I
>>>>>>> couldn't get it to bind to my properties/functions since the code is
>>>>>> loaded
>>>>>>> at runtime and is not defined at compile time (I would hit a NPE when
>>>> the
>>>>>>> PropBindingFactory would try and locate the component).
>>>>>>>
>>>>>>> If anyone could suggest any way of accomplishing this (or if it's
>>>> futile
>>>>>>> and I'm wasting my time), it would be really appreciated.
>>>>>>>
>>>>>>> Thanks,
>>>>>>> David.
>>>>>>
>>>>>> ---------------------------------------------------------------------
>>>>>> To unsubscribe, e-mail: users-unsubscr...@tapestry.apache.org
>>>>>> For additional commands, e-mail: users-h...@tapestry.apache.org
>>>>
>>>> ---------------------------------------------------------------------
>>>> To unsubscribe, e-mail: users-unsubscr...@tapestry.apache.org
>>>> For additional commands, e-mail: users-h...@tapestry.apache.org
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscr...@tapestry.apache.org
>> For additional commands, e-mail: users-h...@tapestry.apache.org
>>
>>
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@tapestry.apache.org
For additional commands, e-mail: users-h...@tapestry.apache.org
