Tab file specification

[boolbag]

Can someone please point me out to the spec for dictionaries in tab file format? It looks like a simple format for me to generate, to then be converted by pyglossary. Did a simple test which worked great.

Looking inside some files, it looks like there are some special tags that can be parsed by pyglossary, e.g.:

• <k > → key
• <c> → color

Dying to see the full spec.

In particular, I need to know how to link to sound and image resources.

In advance, big thanks.

[ilius]

HTML tags are not specific to Tab file or the input file format.
We only detect if the content of an entry is HTML, Plain text or XDXF (in case of Tabfile or other formats that can have be any of these).
And having <k> makes it be detected as XDXF.
Having html tags makes it be detected as HTML.

But which HTML tags will work for you depends on which dictionary application you use.

You have a point about having a format specification, because we do some escaping/unescaping on tabs, newlines and backslashes.

[boolbag]

Thank you, @ilius

In the meantime, saw that GoldenDict will not read images and sound in Stardict files.
They do read sound and images in Lingvo files.
I can generate dsl, but since there are about 200,000 tiny image and sound files, the ideal would be to compile them to Lingvo's compiled lsd format.

It looks like the Lingvo compiler isn't around anymore. Pyglossary doesn't output lsd, right?

[ilius]

No worries.

LSD is a proprietary/closed binary format so takes time to reverse engineer. And Lingvo is pretty abandoned, so that's a no.

[boolbag]

OK.
The good news regarding dsl is that I learned it's possible to place all the media files in a zip file as long as this naming convention is respected. Works on GoldenDict for Android.
your_dictionary.dsl
your_dictionary.dsl.files.zip
Not compiled, but that's all I need.

Thank you for pyglossary, I've been playing with it the past few days and finding it terrific.