Hi Ian, I’ve finally had a chance to look at your code.
I noticed a problem with nested spans. I created with four words, "Three four five six", marking "four" as bold, and "five" as italic. OpenOffice saved this as follows:
<text:p text:style-name="Standard">
Three
<text:span text:style-name="T1">four</text:span>
<text:span text:style-name="T2">five</text:span>
six
</text:p>
ODFParagraphContentGet only handles text nodes right now. I guess you're probably aware of the above issue, but just figured I'd mention it as it's the first problem I ran into.
I can confirm that changing text node directly inside a paragraph causes it to be updated in the odt file after a put operation.
I also tried adding a new paragraph to the HTML file, but that wasn't reflected after an update.
There's quite a few colour macros used here, e.g.:
printf(CYAN "ODFParagraphContentGet\n" RESET);
These colours could perhaps be integrated with Gabriela's logging framework, so they're not explicitly specified by filter code, and can be turned on/off or customised easily. We could perhaps have one colour per level, or alternatively an additional parameter to specify the colour. Gabriela - any thoughts on this?
There were also three files I found that had an ASF license header:
DocFormats/api/src/Operations.c
DocFormats/filters/odf/src/lenses/ODFDocument.c
DocFormats/filters/odf/src/lenses/ODFHeader.c
On the corinthia-node suggestion - I would recommend putting it into corinthia-webapp repository instead. Franz has also been working on a node-based server, so perhaps the two of you could collaborate on this.
(fingerprint 5435 6718 59F0 DD1F BFA0 5E46 2523 BAA1 44AE 2966)