Re: [JIRA] Created: (FOR-505) Allow the retrieval of raw html files

[Ross Gardler <rg...@apache.org>]

David Crossley wrote:
> Ross Gardler wrote:
> 
>>[ snip ]
>>
>>We could add a new skinconf property to define the behaviour for *.html 
>>files in xdocs we can tell people affected by this issue to make their 
>>embedded content ihtml and their raw content html. i.e.
>>
>><!-- HTML files in the XDoc directory can  wither be passed unprocessed 
>>to the client, or they can be skinned like any other content. If you set
>>this setting to false, you can include skinned files by giving them the
>>ihtml extension. -->
>>
>><skinHTMLSources>true</skinHTMLSources>
>>
>>The result is, setting skinHTML to false will give the same behaviour as 
>>0.6 We could make this the default in order to minimise the upgrade 
>>behaviour.
> 
> 
> If we implement this workaround, then i reckon that the default
> should be "true". It would be easier to explain the "false" case
> and it would be a smaller number of people affected.
> 
> However, it forces them to change their filename extensions.
> That would be difficult for the main use-case, which is to include
> a set of existing *.html docs as part of the site.
> 
> The thread "RT: RAW content" that is linked from issue FOR-505
> ended with an interesting idea about using site.xml to specify
> exclude/include patterns.
> http://marc.theaimsgroup.com/?l=forrest-dev&m=110172196321097
> It seems to be independent from the "locationmap" ideas discussed
> earlier in that thread. Would that be better to implement now,
> rather than the skinHTMLSources=true workaround?

Wow, that is the most focused rambling thread I have read for a long 
time, all that discussion to come around to such an elegant solution. 
Kudos to everyone involved (and thank goodness for mail archives, I had 
completely forgotten the whole thing). A tip for the wary read the first 
few and the last email, skip everything in between it goes round many 
houses to get to the last mail.

Anyway, +1 for implementing it instead of  the skinHTMLSources=true 
workaround.

I'm game for having a go at it, but I'm not sure when I will be able to 
so others should feel free to jump in.

Ross

Re: [JIRA] Created: (FOR-505) Allow the retrieval of raw html files

[Ross Gardler <rg...@apache.org>]

Ferdinand Soethe wrote:
> 
> 
> 
> Ross Gardler wrote:
> 
> 
>>>And would I still have an additional separate entry in the sitemap to
>>>have /old_site_1/index.htm on the menu or would this have to be
>>>within the rawContent element.
> 
> 
>>Do you mean "would I still have an additional separate entry in the 
>>*skins.xml*"?
> 
> 
> Sorry, I was writing bullshit. Of course I meant

LOL - so was I, at least we understood what we intended to say!

Ross

Re: [JIRA] Created: (FOR-505) Allow the retrieval of raw html files

[Ferdinand Soethe <sa...@soethe.net>]

Ross Gardler wrote:

>> And would I still have an additional separate entry in the sitemap to
>> have /old_site_1/index.htm on the menu or would this have to be
>> within the rawContent element.

> Do you mean "would I still have an additional separate entry in the 
> *skins.xml*"?

Sorry, I was writing bullshit. Of course I meant

> And would I still have an additional separate entry in the site.xml to
>                                                            ^^^^^^^^
> have /old_site_1/index.htm on the menu or would this have to be
> within the rawContent element.

But your reply has already made this clear this: I do need to
reference a file to show it one a menu since rawcontent will only
provide the processing hints.

Thanks

--
Ferdinand Soethe

Re: [JIRA] Created: (FOR-505) Allow the retrieval of raw html files

[Ross Gardler <rg...@apache.org>]

Ferdinand Soethe wrote:
> Ross Gardler wrote:
> 
> 
>>Anyway, +1 for implementing it instead of  the skinHTMLSources=true
>>workaround.
> 
> 
> Pardon my ignorance but is this
> 
> 
>>A third possibility that is a marrying of the two is elements in
>>site.xml to make it much more like the Ant fileset idea, like this:
>>
>><!-- everything in old_site directory -->
>><rawContent dir="/old_site/**"/>
>>
>><!-- everything in 0.6_docs, except the forums -->
>><rawContent dir="/0.6_docs/**">
>>   <exclude name="forums/**"/>
>></rawContent>
>>
>><!-- everything in 0.6_docs, except all files in forums except the 
> 
> index.html file -->>
> 
>><rawContent dir="/0.6_docs/**">
>>   <exclude name="forums/**"/>
>>   <exclude name="forums/**/index.html"/>
>></rawContent>
>>
>>This leaves site.xml easily readable, allows such content to be provided 
>>in an external file if there is no way of generating it from another 
>>type of site descriptor file and allows us to describe pages not linked 
>>to in site.xml.
> 
> 
> what we are going to implement now?

Yes, if the concept can be brought to a working design. Of course that 
is the hard part.

> And if so, what would it look like if I had raw content in more than
> one branch? For example could I have
> 
> <rawContent dir="/old_site_1/**"/>

> <rawContent dir="/old_site_2/**"/>

Just add an element for each of the matches that represent raw content, 
each with its own include and exclude patterns.

> And would I still have an additional separate entry in the sitemap to
> have /old_site_1/index.htm on the menu or would this have to be
> within the rawContent element.

Do you mean "would I still have an additional separate entry in the 
*skins.xml*"?

If so, yes, there would be no need to change any other functionlaity in 
Forrest.

What would happen is that when we receive a request for /old_site_1/**
the sitemap would first look in the config to see if it is raw content. 
If it is it would read the content and serve it. If it isn't processing 
would proceed as normal.

When I find the time I'll write up a proposal about how to implement this.

Ross

Re: [JIRA] Created: (FOR-505) Allow the retrieval of raw html files

[Ferdinand Soethe <sa...@soethe.net>]

Ross Gardler wrote:

> Anyway, +1 for implementing it instead of  the skinHTMLSources=true
> workaround.

Pardon my ignorance but is this

> A third possibility that is a marrying of the two is elements in
> site.xml to make it much more like the Ant fileset idea, like this:
> 
> <!-- everything in old_site directory -->
> <rawContent dir="/old_site/**"/>
> 
> <!-- everything in 0.6_docs, except the forums -->
> <rawContent dir="/0.6_docs/**">
>    <exclude name="forums/**"/>
> </rawContent>
> 
> <!-- everything in 0.6_docs, except all files in forums except the 
index.html file -->>
> <rawContent dir="/0.6_docs/**">
>    <exclude name="forums/**"/>
>    <exclude name="forums/**/index.html"/>
> </rawContent>
> 
> This leaves site.xml easily readable, allows such content to be provided 
> in an external file if there is no way of generating it from another 
> type of site descriptor file and allows us to describe pages not linked 
> to in site.xml.

what we are going to implement now?

And if so, what would it look like if I had raw content in more than
one branch? For example could I have

<rawContent dir="/old_site_1/**"/>

and

<rawContent dir="/old_site_2/**"/>


And would I still have an additional separate entry in the sitemap to
have /old_site_1/index.htm on the menu or would this have to be
within the rawContent element.

Thanks

--
Ferdinand Soethe