Re: Getting text and non-block nodes before any block

Yes, in my tests, the original for-each-group will select whitespace-only
nodes as the first group, which is not good. Using the
*|text()[matches(.,'\S')] selector seems to work, including preserving
whitespace between non-block elements.

The for-each-group solution is definitely better than my initial approach.

Cheers,

E.
bbbbb
Eliot Kimber, Owner
Contrext, LLC
http://contrext.com




On 9/15/14, 8:40 AM, "Wendell Piez wapiez@xxxxxxxxxxxxxxx"
<xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx> wrote:

>Hi again,
>
>I should have added that when grouping to induce structure, you need
>to keep in mind that select="node()" will select whitespace-only text
>nodes, comments and PIs along with the stuff you probably want, and
>these can split your groups. So select="*|text()[matches(.,'\S')]" (or
>a similar adjustment) may better when using this technique.
>
>Cheers, Wendell
>
>
>On Mon, Sep 15, 2014 at 9:33 AM, Wendell Piez wapiez@xxxxxxxxxxxxxxx
><xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx> wrote:
>> Hi Eliot,
>>
>> I would treat this as a special application of a more general problem,
>> namely grouping loose children that might have been wrapped, but
>> weren't. Such as splitting paragraphs around lists or other constructs
>> not permitted in paragraph analogues in the target.
>>
>> So, select="node()" group-adjacent="cp:isBlock(.)" will return groups
>> that are either all blocks, or all not blocks.
>>
>> Inside for-each-group, position() returns the position of the group
>> within all the groups, so you can use it to determine the first group.
>>
>> But with all the groups given, plus
>> when/@test='current-grouping-key()" to determine whether it is (or
>> isn't) a group of blocks, much more is also possible.
>>
>> Cheers, Wendell
>>
>>
>>
>> On Fri, Sep 12, 2014 at 8:10 PM, G. Ken Holman g.ken.holman@xxxxxxxxx
>> <xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx> wrote:
>>> At 2014-09-12 21:50 +0000, Eliot Kimber ekimber@xxxxxxxxxxxx wrote:
>>>>
>>>> I think this is an FAQ but my situation may be slightly different. In
>>>>any
>>>> case, if it's an FAQ, maybe my asking it will help.
>>>>
>>>> This is in a DITA context, where I can know for any element whether
>>>>it is
>>>> a block or not a block (because in DITA all elements have a base type
>>>>that
>>>> is inherently block or not block).
>>>>
>>>> I have the situation where a wrapper element may contain any
>>>>combination
>>>> of text, non-block elements, or block elements.
>>>>
>>>> I need to find either all the nodes up to the first block or determine
>>>> that there are none and get the first block.
>>>
>>>
>>> I think this will work:
>>>
>>>   select="node()[. &lt;&lt; current()/*[lc:isBlock(.)][1]"
>>>
>>> But it is XSLT 2.0 only.
>>>
>>> For XSLT 1.0 I might do something like:
>>>
>>>   select="node()[generate-id(following-sibling::*[lc:isBlock(.)][1]) =
>>>                  generate-id(current()/*[lc:isBlock(.)][1])]"/>
>>>
>>>> Where the lc:isBlock() function returns true if the element provided
>>>>is a
>>>> DITA block element.
>>>>
>>>> This seems to pass all my tests but I'm wondering if there's a better
>>>>way
>>>> to make this check?
>>>>
>>>> I also realized that I can use the $textBeforeBlocks variable to then
>>>> process any remaining blocks (because I will have already output the
>>>>text
>>>> before blocks):
>>>>
>>>> <xsl:when test="$hasTextBeforeBlocks">
>>>>   <xsl:apply-templates select="node() except($textBeforeBlocks)"/>
>>>>         </xsl:when>
>>>
>>>
>>> That includes the block element ... you might try using ">>" in a way
>>> similar to how I used "<<" above if you didn't want to include the
>>>block
>>> element itself.
>>>
>>> I hope this helps!
>>>
>>> . . . . . . Ken
>>>
>>>
>>> --
>>> Contact us for world-wide XML consulting and instructor-led training |
>>> Free 5-hour lecture: http://www.CraneSoftwrights.com/links/video.htm |
>>> Crane Softwrights Ltd.            http://www.CraneSoftwrights.com/s/ |
>>> G. Ken Holman                   mailto:gkholman@xxxxxxxxxxxxxxxxxxxx |
>>> Google+ profile:      http://plus.google.com/+GKenHolman-Crane/about |
>>> Legal business disclaimers:    http://www.CraneSoftwrights.com/legal |
>>>
>>>
>>> ---
>>> This email is free from viruses and malware because avast! Antivirus
>>> protection is active.
>>> http://www.avast.com
>>>
>>>
>>
>>
>>
>> --
>> Wendell Piez | http://www.wendellpiez.com
>> XML | XSLT | electronic publishing
>> Eat Your Vegetables
>> _____oo_________o_o___ooooo____ooooooo_^
>>
>
>
>
>--
>Wendell Piez | http://www.wendellpiez.com
>XML | XSLT | electronic publishing
>Eat Your Vegetables
>_____oo_________o_o___ooooo____ooooooo_^