Rework used encodings

[tobiasdiez]

I think JabRef sometimes uses encodings which are a bit counter-intuitive or where a better alternative might be available. Triggered by JabRef/JabRef-Browser-Extension#33.

In this issue, I try to collect the current usage of encodings so that we get an overview. Feel free to edit this list.

Operation	Currently used	Suggestion
Import from cmd (--importToOpen)	File -> Default	Argument -> File -> DB -> UTF8
Fetcher	?	Fetcher decides -> File -> UTF8
Export	?	?
?	?	?

Legend:

• File: as JabRef comment %Encoding: ....
• Default: JabRefPreferences.getDefaultEncoding()
• Argument: via the command line or by asking the user
• DB: the encoding of the DB where the entries are imported to / exported from

[lenhard]

One thing that we talked about earlier with regard to the BibtexParser: The encoding supplied in the bib file (%Encoding) is currently completely ignored by JabRef. The parser explicitly ignores it upon reading a file and it is stored nowhere. On writing the file, the encoding set in the preferences is used.

I want to note that this is nothing we "broke", JabRef has always done that as far as I can recall.

[tobiasdiez]

Its ignored by the parser but not by the importer:

jabref/src/main/java/net/sf/jabref/logic/importer/fileformat/BibtexImporter.java

Line 43 in ee08252

public ParserResult importDatabase(Path filePath, Charset defaultEncoding) throws IOException {

[lenhard]

I see. My knowledge seems to be confined to the deeper parts of our parsing stack ;-)

[tobiasdiez]

I don't think that we currently have open issues related to the encoding. Although the used encoding is still sometimes a bit mystic, I will close this issue now.