MakePlecoDict - Formatting?

[ipsi]

Is it possible to specify formatting when using this? I'm using it to create a custom dictionary out of my flashcards (I'm trying to learn the vocab from a text book, and I don't really want the user dictionary full of the thousand or so entries...). It works fine, and it works as I expect, but for the lack of formatting and the inability to recognise punctuation in the headwords (e.g. my text specifies 块（钱）, with the （） letting you know that 钱 is optional. However, Pleco doesn't like that. Is there anything that can be done about that? Or do I just have to put up with removing them for now and leaving them in the Pinyin definition (Which seems to work fine)?

The formatting I'm looking for is Bold Face (as in the Oxford), and maybe also the numbers in Circles (as in the ABC). Can this be done with MakePlecoDict? Or am I just totally out of luck?

 

[mikelove]

It can be done, but you need either a hex editor or a program that lets you insert custom / undefined font symbols. Unicode character 0xEAB2 starts a boldfaced range, 0xEAB3 ends it, and the circled numbers are the standard Unicode enclosed digits range (0x2460-0x246F).

There's no way to get it to handle parentheticals correctly, though, at least not in the current version.

 

[ipsi]

Ok, that's vaguely annoying, but I'm sure I can manage. Thanks.

Also, totally off topic, why isn't there a "Report this message" button for PMs? I've just received some spam...

 

[ipsi]

Actually, another question: When switching from Simplified to Traditional Characters, how is the conversion performed? I know that there is not a one-to-one mapping from Trad to Simp, so how does the software do it?

When writing a dictionary, would it be a better idea to put any characters in the definition in Traditional Characters to ensure the meaning is preserved? (If done in simplified characters, it may not translate perfectly to traditional).

Also, I note that the range of enclosed alpha-numerics goes to 24FF. Do you support up to that, or only up to 246F?

EMEditor lets you insert special characters. That's definitely a wonderful editor, and I'm glad I've got even the free version.

 

[ipsi]

Fun times. Seems to work fine. Java's quite nice about inputting Unicode it would seem.

Mostly what I wanted to know this for is to make a program to play around with flashcard lists on the PC (merging, splitting, converting to dict, etc.), as I figured if I'm going to have my flashcards be a dictionary, I might as well try and make them be a nice dictionary.

The results look something like this. EDIT: 上 for specific examples of the stuff I just asked about.

The code looks like this. Yes. The code is ugly as sin. I haven't gotten around to tidying it up yet. I also haven't provided a JAR file because I'm lazy, and it's not really ready for use by anyone other than me.

 

[mikelove]

Not sure why phpBB doesn't include a "Report this message" button, but moot point since we're switching to vBulletin in a month or two anyway.

The simplified-to-traditional converter is pretty dumb, actually; if it sees a simplified character which it knows should always be changed to a specific traditional character, it does, but for ambiguous cases it does nothing unless there's an extra code added after the character specifying the traditional version to use. (0xEABA, then the traditional character, then 0xEABB)

We only include the 2460-2473, 24B6-24B9, and 24D0-24D5 portions of the enclosed alphanumerics range, basically because we had to manually create bitmaps (in 4 different font sizes!) for them and didn't see any point in creating them for characters we weren't using.

Interesting code, spending most of my time programming in straight C I find myself envying anyone who gets to write text manipulation routines with a decent string class at their disposal...

 

[ipsi]

Good fun.

Traditional: Ok then. Probably be best to write everything out in traditional characters then?

Manually creating fonts... Ouch... That must've hurt...

Pure C... Ohh... Never done that, for which I am quite glad. Java's reasonably nice, most of the time. But yeah. I wouldn't want to try writing that in C. I just don't see it as being at all easy (well, compared to Java anyway).

 

[mikelove]

Unfortunately, the converter doesn't work in the other direction (never bothered to make it do so, since all of the dictionaries we license are primarily oriented around simplified) so if you put the whole thing in traditional then it'll only show up in traditional in PlecoDict. As you may have surmised, MakePlecoDict was kind of an afterthought, we really aren't planning to release any serious tools for creating your own dictionaries until the desktop version comes out, so for the moment it's probably best to muddle through with it but keep your text files in a relatively easy-to-reformat state so you can bring them into our new and much better converter later on.