hi the effort to get to a "modular" extraction of the text and its formatting was more painful than i thought and the weather was way too hot...
so, with some delay, the current state before the next weekend: - text gets exported. i have a basic test for italic, bold, superscript, subscript and paragraph styles and they seem to be correctly exported. - multiple frames are exported sequentially, but the order is wrong (i have to fix the sorting routine... a question of static functions that do not behave as a i would hope). - the epub file does open and display, but the "mimetype" file is not correctly put at the beginning of the file itself (i will have to check this: it seems to be an error in the new zip library scribus is using...). the epubchecker should not show other warnings. i have created three test files that can be found at https://github.com/aoloe/scribus-plugin-export-epub/tree/master/test/samples it would be helpful if the community could provide more test cases. here again some rules: https://github.com/aoloe/scribus-plugin-export-epub#samples finally, after having struggled so much with the scribus internals (and its undocumented changes) i've refactored all the access to the scribus document and its elements in a few high level classes that i'd like to see promoted as some sort of API for plugins. the plan is to make an own plugin out of it. i would love it, if somebody could have a look at it and review it (and check if it makes sense...). https://github.com/aoloe/scribus-plugin-export-epub/tree/master/src/module if other people think that this can be useful, i would appreciate some help in finding the best way to add a documentation for it (http://readthedocs.org?). the plans for the near future are: - fix the ordering of the frames. - adding the scripter bindings, so that i can automatically create multiple epubs starting from the existing test cases. - start to refactor in the code related to the image frames. a further quest for help: i could need some help in setting up a python script that compares some reference epub files with the ones freshly generated... some sort of unit tests that acts on files. does anybody have any hint? anybody wanting to work on this? i wish you a wonderful weekend! a.l.e
