HWPF current status

[Themistoklis Dakanalis <da...@csd.uoc.gr>]

Hi,

i have some fundamental questions about the HWPF and the word elements it
supports. Particularly, I am interesting in the  following word elements:

   1. Text (headings, paragraphs,...)
   2. Tables
   3. Images
   4. Headers & Footers
   5. Footnotes

Could you tell me what is the current support of HWPF concerning the above
elements?

Thanks in advance!

-- 
Themistoklis Dakanalis

Re: HWPF current status

[Themistoklis Dakanalis <da...@csd.uoc.gr>]

Hi Rainer,

yes i think that this is the level of detail I am interested in. As i had in
my mind when i wrote the previous message, i will have  a lot of work (code)
to do. My final goal is to transform word documents into XML (DITA
specifications). So, I hope to have the support of this community during my
effort.

Thanks a lot!!!


On 9/5/07, Rainer Schwarze <rs...@admadic.de> wrote:
>
> Hi,
>
> Themistoklis Dakanalis wrote:
> > i have some fundamental questions about the HWPF and the word elements
> it
> > supports. Particularly, I am interesting in the  following word
> elements:
>
> General limitation: HWPF can read a lot of real life Word files, but not
> write them. (I seem to remember that "fast saved" files cannot be read
> reliably.)
>
> >    1. Text (headings, paragraphs,...)
>
> Text can be retrieved, YMMV with 2-byte characters.
> Headings should be identifiable by reading the paragraph style and
> comparing its id with the predefined heading style numbers.
> (I think that these codes are not yet declared conveniently in HWPF, so
> you need to look at the Word spec if you need them [get yourself a nice
> comfy chair when reading the Word spec and put away all heavy and
> dangerous objects in case you become furious while reading it :-) ].)
>
> Formatting information (bold, italic, ...) can be retrieved. If you need
> very accurate formatting information, you need to create a few relevant
> test cases and check that everything works as expected. I remember that
> HWPF sometimes uses other defaults than Word and may for instance
> sometimes say font size is "12pt" when it actually is "10pt".
>
> >    2. Tables
>
> There is a getTable method in Range which retrieves a Table instance for
> the table starting at the supplied Paragraph. I don't know how well that
> works. You should expect that you get most, but not all Table
> properties, because several property codes occuring in table information
> are not documented in the public Word spec.
>
> >    3. Images
>
> Simple images should work, Office Art is not supported.
> I doubt that images referenced by fields only ( {INCLUDEPICTURE} ) are
> supported (Fields are not supported). If the image is cached in the Word
> file, it may be accessible via the PicturesTable class, but I did not
> check that.
>
> >    4. Headers & Footers
> >    5. Footnotes
>
> Those are not fully supported. Trouble increases with 2-byte characters,
> fields and bookmarks in these parts.
>
>
> I hope that this is roughly the detail level you were interested in. If
> you need more details, just ask :-)
>
> If anybody has a different view on these topics, please correct me.
>
> Best wishes, Rainer
> --
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@poi.apache.org
> For additional commands, e-mail: user-help@poi.apache.org
>
>


-- 
Themistoklis Dakanalis

Re: HWPF current status

[Rainer Schwarze <rs...@admadic.de>]

Hi,

Themistoklis Dakanalis wrote:
> i have some fundamental questions about the HWPF and the word elements it
> supports. Particularly, I am interesting in the  following word elements:

General limitation: HWPF can read a lot of real life Word files, but not
write them. (I seem to remember that "fast saved" files cannot be read
reliably.)

>    1. Text (headings, paragraphs,...)

Text can be retrieved, YMMV with 2-byte characters.
Headings should be identifiable by reading the paragraph style and
comparing its id with the predefined heading style numbers.
(I think that these codes are not yet declared conveniently in HWPF, so
you need to look at the Word spec if you need them [get yourself a nice
comfy chair when reading the Word spec and put away all heavy and
dangerous objects in case you become furious while reading it :-) ].)

Formatting information (bold, italic, ...) can be retrieved. If you need
very accurate formatting information, you need to create a few relevant
test cases and check that everything works as expected. I remember that
HWPF sometimes uses other defaults than Word and may for instance
sometimes say font size is "12pt" when it actually is "10pt".

>    2. Tables

There is a getTable method in Range which retrieves a Table instance for
the table starting at the supplied Paragraph. I don't know how well that
works. You should expect that you get most, but not all Table
properties, because several property codes occuring in table information
are not documented in the public Word spec.

>    3. Images

Simple images should work, Office Art is not supported.
I doubt that images referenced by fields only ( {INCLUDEPICTURE} ) are
supported (Fields are not supported). If the image is cached in the Word
file, it may be accessible via the PicturesTable class, but I did not
check that.

>    4. Headers & Footers
>    5. Footnotes

Those are not fully supported. Trouble increases with 2-byte characters,
fields and bookmarks in these parts.


I hope that this is roughly the detail level you were interested in. If
you need more details, just ask :-)

If anybody has a different view on these topics, please correct me.

Best wishes, Rainer
-- 

---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@poi.apache.org
For additional commands, e-mail: user-help@poi.apache.org