Compress df file output

[lasconic]

See https://github.com/BoboTiG/ebook-reader-dict/issues/1161#issuecomment-1024989236

• Who uses the df file ?
• Do they use it directly in a device that couldn't use the compressed version ?
• Can these users handle the decompression process ?

[Moonbase59]

I don’t have access to a Kobo. Maybe we can gather some info on that on MobileRead (English) and E-Reader Forum (German)?

These two seem to be the most frequented forums for people who wish to optimize their devices. Both also have Kobo subforums (is .df a Kobo-only format?).

Might also be a way to make people aware of this project. I, for instance, found it only by accident, even though I work on hyphenation and dictionary stuff for quite a while now.

---

For my understanding: I read about dictutil. Is it true that a .df dictfile is just a "Markdown-like" input file to generate these odd dicthtml.zip dictzip files from? And does the reader handle .df or only the dictzip files?

[lasconic]

df is not a "kobo format", meaning Kobo ereaders can't read it. It's only an input format for other scripts I believe. It is a text file indeed, and I guess it's only used by users as an input in pyglossary to generate other formats (like we do for stardict).

[BoboTiG]

@lasconic is totally right. We introduced the DictFile format only to be able to generate the StarDict, because we could not use "the dicthtml" Kobo file as input.

We could compress that file if PyGlossary can read it, it will save a good amount of bytes.