[prev in list] [next in list] [prev in thread] [next in thread] 

List:       libreoffice
Subject:    [GSoC] QuarkXPress import, Week 3
From:       Alex P <alex.pantec () gmail ! com>
Date:       2017-06-21 19:28:31
Message-ID: CAANYnKedpKAVsBx9z2j82JBm-8CJt_yiYs+-B8tKmRhEmCQCyQ () mail ! gmail ! com
[Download RAW message or body]

[Attachment #2 (multipart/alternative)]


During this week we continued reverse engineering of QuarkXPress 3-4 format
and had some substantial progress:
- Completed object type identification (shape type, content type).
- Refactored object parsing and improved size identification (QXP does not
store sizes/offsets for object records), which should cover most cases now.
Currently only for QXP 3.3, but it should not be difficult to implement in
QXP 4, it's very similar.
- Finished parsing of paragraph styles, rules, tabs.
- Refactored and improved parsing of document structure, block chains. This
should be the ultimate version of it: we may identify new records and
details, but there probably will not be any major changes in document
parsing.
- Identified some object properties: color shade, line/frame style, gap
color, text inset, vertical alignment, baseline offset, ... Looks like we
identified most of the object properties except some more advanced such as
picture cropping.
- Parsing of QXP 4 object properties (they are mostly the same, sometimes
different order or size).
- Linked text boxes.
- Object groups.
- Some progress with QXP 4 deobfuscation, but there are still major issues.

[Attachment #5 (text/html)]

<div dir="ltr">During this week we continued reverse engineering of QuarkXPress 3-4 \
format and had some substantial progress:<div>- Completed object type identification \
(shape type, content type).</div><div>- Refactored object parsing and improved size \
identification (<span style="font-size:12.8px">QXP does not store sizes/offsets for \
object records</span>), which should cover most cases now. Currently only for QXP \
3.3, but it should not be difficult to implement in QXP 4, it&#39;s very \
similar.</div><div>- Finished parsing of paragraph styles, rules, tabs.</div><div>- \
Refactored and improved parsing of document structure, block chains. This should be \
the ultimate version of it: we may identify new records and details, but there \
probably will not be any major changes in document parsing.</div><div>- Identified \
some object properties: color shade, line/frame style, gap color, text inset, \
vertical alignment, baseline offset, ... Looks like we identified most of the object \
properties except some more advanced such as picture cropping.</div><div>- Parsing of \
QXP 4 object properties (they are mostly the same, sometimes different order or \
size).</div><div>- Linked text boxes.</div><div>- Object groups.</div><div>- Some \
progress with QXP 4 deobfuscation, but there are still major \
issues.</div><div><br></div></div>


[Attachment #6 (text/plain)]

_______________________________________________
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice


[prev in list] [next in list] [prev in thread] [next in thread]