JPEXS Free Flash Decompiler Issue Tracker

Fonts in this SWF file are written upside-down, are way bigger than they should be (about 20 times) and texts have vertical flip text matrix and heights 20 times smaller which places glyphs into correct readable position. Unfortunately, in PDF file it causes thin selection lines. If you import another font into DefineFont2 tag using Embed button, the font will not be upside down, but the texts vertical flip matrix + heights stay in place, so you will need to modify the texts, which can be pretty tedious work if you got many of them. Additionally, each text has outline from a DefineShape3 placed over it, making it bolder, which probably complicates changing a font to other a bit more. You could probably write a program in Java, which will use FFDecs library, and takes the SWF as input and flips + resize the fonts and texts to normal size/position, then it can be exported to PDF with proper selection.

Is there any way to write pdf from swf using pymupdf in python?

I am not really into using python in Java. The convertor we use is already pretty robust. The problem with fonts/texts needs to be solved first before conversion to PDF. Could you please provide us more samples of such SWFs? Maybe I can figure out something...

Any method to export json with bounding box, text, etc. for all text lines, images, shapes, etc,?

There's no such method. You can export XML from SWF. I may be able to implement a fix to FFDec which will handle such fonts/texts and generate PDF with proper selections. You can help me if you provide more SWFs with such problem, I want to be sure it works in all cases, not this specific single one SWF.

Great, will wait for your updated program. Meanwhile, I want to use the pdf to convert to pptx properply, or is there any way to convert swf to pptx?

I sent a zip file for 600 swf files, can you open?

Yes, I got your zip file, it is okay. I don't know a way how to convert SWF to PPTX.

Acrobat can save pdf as pptx, but existing pdf caused bad pptx result, maybe due to wrong thin text lines.

Any update?

Yes, the fix will be possible, just be patient, I need to do some tests... I also have other regular fulltime job, so can't work on it 24/7...

Great. Will wait for update to test.

Please try nightly 3308. I added font/text normalization. It can be executead manually from context menu on SWF file, and is also automatically executed on PDF and SVG exports. It will solve your problem with thin selection in PDF.

It looks good. Thank you. Is it possible to replace font before writing?

You mean to replace font with another one before exporting PDF? Why would you do that? You can execute the "Normalize fonts" from context menu of the SWF, then you can go to the font (marked with bold as changed) and hit Embed button and select characters from another installed font or from a TTF file. But you may encounter a problem that around each text in this SWF, there is a thin outline made of shape. If you change the font of the text, then the outline will not match and it will look weird. You'll need to delete the outlines from the file first. This is not something that can be done automatically.

Ok. I will use pymupdf to replace font