There is an extra argument now being passed to the elehandler, the
last argument of dirty is set to 1 if the pap/chp can not be
guaranteed to be exactly the same as the base style. Though it is possible
that a pap/chp could have two sprms which cancel each other out to leave the
pap/chp exactly the same as the original style. In this case the element will
be marked (erroneously ?) as being dirty, this kind of nonsense is likely only
to happen in fastsaved files, which really are the spawn of the devil.
This extra arg certainly requires a source change in abi, sorry about that.
(while we're at it a few more rohr warnings have been removed)
the next task is to fix out of sequence styles.
To document the stylesheet for you, heres what we have.
the wvParseStruct contains a STSH member, this has a STSHI
header with the number of styles mentioned, followed by an
array of STD elements for each style.
The sgc member of each style tells you if its a paragraph or
standalone character style. Now every paragraph style also
contains a character style. I dont really grok standalone
character styles. I dont believe that a section of text can
really be a standalone character style, as it would have to
override the paragraph style, and then it would have a "dirty"
element.
The vast majority of styles are paragraph styles that also include
a character style.
Now each paragraph style has a grupe array with two members, the
first is a PAP, and the second is a CHP. So you could quite easily
look through the std list and extract the properties of the para
and chp for the style, and map them nicely to an abi style. The
name of the style is in unicode in the xstzName member of the STD,
according to the spec this can be a list of possible names, ive
yet to see this so i cant comment on what the delimimator character
might be.
Each character style has one grupe member, consisting of a CHPX, the
X means that its not a CHP, but a delta from a CHP. Id currently reccomend
that you ignore these character styles for a while, as it consists of sprms
that you have to parse to understand. I have to have a think about them to
see what way they can be munged into something readable.
So given that abi has style support you can parse the STSH stucture and
map the styles that have a sgc == sgcPara into abi styles, then when the
elehandler has a 0 set for dirty the PAP or CHP pointed to by the istd (which I
should point out is a simple index into the std array) is exactly the same as
the PAP or CHP of the paragraph in question.
Im not sure that im making a lot of sense, but this is some documentation to get
anyone starting on a style mapping project going. I suppose this should make
importing quicker as well, as there will now be frequent CHP and PAP's that dont
have to be parsed at all to know what they are.
C.
Real Life: Caolan McNamara * Doing: MSc in HCI
Work: Caolan.McNamara@ul.ie * Phone: +353-86-8790257
URL: http://www.csn.ul.ie/~caolan * Sig: an oblique strategy
Use an old idea