feature: Replace (cid:x) values for Latin charset

[krishnasism]

As mentioned in the issue #106, pdfminer.six has problems when faced with (cid:x) values when the font is not embedded.

We can manually overwrite the (cid:x) values based on the charset. For now added only LATIN charset.

[krishnasism]

Thinking about it. What do you think about allowing users to pass their custom charsets 🤔 @aptakhin for example

charset = {128: "@"}

pdf = HotPdf(file, decoder=Decoder(charset))

We can have our internal ones that the user can use or they can pass their custom ones. More freedom, and we don't need to maintain all charsets.

[aptakhin]

Thinking about it. What do you think about allowing users to pass their custom charsets 🤔 @aptakhin for example

charset = {128: "@"}

pdf = HotPdf(file, decoder=Decoder(charset))

We can have our internal ones that the user can use or they can pass their custom ones. More freedom, and we don't need to maintain all charsets.

Yes, I think it might be good that we allow users to make their own Decoder if our standard wasn't so useful.