Dear Daniel,
I think you nailed it, it's obviously a filtering problem !
One more issue regarding json comma crossed my head just now.
Example being this:
{
<#list mylist1 as item>
'${item}': '${item}'<#sep>,</#sep>
</#list>,
<#list mylist2 as item>
'${item}': '${item}'<#sep>,</#sep>
</#list>
}
If mylist1 is empty, then it will result in
{
,
'mylist2item1': 'mylist2item1',
'mylist2item2': 'mylist2item2'
}
And if mylist2 happens to be empty, then it'll result in
{
'mylist2item1': 'mylist2item1',
'mylist2item2': 'mylist2item2'
,
}
If both are empty, then
{
,
}
So the ugly and not reusable way to deal with this would be:
{
<#list mylist1 as item>
'${item}': '${item}'<#sep>,</#sep>
</#list><#if mylist1?has_content && mylist2?has_content>,</#if>
<#list mylist2 as item>
'${item}': '${item}'<#sep>,</#sep>
</#list>
}
The reusable approach is maybe sth like this
{
<@optionalNested commaSuffix=true>
<#list mylist1 as item>
'${item}': '${item}'<#sep>,</#sep>
</#list>
</@optionalNested>
<#list mylist2 as item>
'${item}': '${item}'<#sep>,</#sep>
</#list>
}
where optional nested only adds a comma at the end when it's
trimmed-<#nested> is not empty.
Are there any other solutions that i might miss ?
On Mon, Jun 13, 2016 at 1:55 PM, Daniel Dekany <[email protected]> wrote:
> That's a work around for the separator, but this problem is deeper
> that that. Things like x?index, x?hasNext, x?itemParity, etc. don't
> work either if you do filtering like that. So I plan to add closure
> support (if it's technically possible...) and then you can do
> something like:
>
> <#list hobbies?filter(i -> i != 'walk') as hobby>
>
> Though you meant to filter lists before passing them to the
> data-model, so it will also support bad practices... But I know that
> in real world filtering has good uses too.
>
>
> Monday, June 13, 2016, 3:55:29 AM, Albert Kam wrote:
>
>> Hello Daniel,
>>
>> Thanks for the great updates !
>>
>> I have something in mind concerning the <#sep> directive. Is it
>> possible to add a startSeparator only when the previous loop produces
>> content ?
>> This happens a lot when printing json hash entries. Example :
>>
>> <#local hobbies=['read', 'code', 'walk']>
>> {
>> <#list hobbies as hobby>
>> <#if hobby != 'walk>
>> 'prop${hobby_index}': '${hobby}'<#sep>,<#sep>
>> </#if>
>> </#list>
>> }
>>
>> This will result in
>> {
>> 'prop0': 'read',
>> 'prop1': 'walk',
>> }
>> Notice the extra comma after 'walk'.
>>
>> I have a workaround for this issue with adding the comma at the head
>> if already had an entry:
>> <#local hasEntry=false>
>> <#list hobbies as hobby>
>> <#if hobby != 'walk>
>> <#if hasEntry>,</#if>
>> 'prop${hobby_index}': '${hobby}'
>> <#local hasEntry=true>
>> </#if>
>> </#list>
>>
>> This will result in
>> {
>> 'prop0': 'read'
>> ,
>> 'prop1': 'walk'
>> }
>>
>> The ideal directive to emultate the workaround above would look
>> something like this:
>> <#list hobbies as hobby>
>> <#if hobby != 'walk>
>> <#head_sep>,<#head_sep>
>> 'prop${hobby_index}': '${hobby}'
>> </#if>
>> </#list>
>> Where <#head_sep> renderes the comma only if the previous loop had an entry.
>>
>> Just a suggestion.
>>
>> Thanks !
>>
>>
>> On Mon, Jun 13, 2016 at 3:32 AM, Daniel Dekany <[email protected]> wrote:
>>> I plan to start a voting here on 2016-06-14, for releasing 2.3.25. So
>>> if anyone wants to add anything before that, inform me.
>>>
>>> This is what we have so far:
>>> - Version history:
>>> http://freemarker.org/builds/2.3.25-nightly/documentation/_html/versions_2_3_25.html
>>> - Last build: http://freemarker.org/builds/2.3.25-nightly/freemarker.jar
>>>
>>> --
>>> Thanks,
>>> Daniel Dekany
>>>
>>
>>
>>
>
> --
> Thanks,
> Daniel Dekany
>
--
Do not pursue the past. Do not lose yourself in the future.
The past no longer is. The future has not yet come.
Looking deeply at life as it is in the very here and now,
the practitioner dwells in stability and freedom.
(Thich Nhat Hanh)
