How to Install and Run Untranslated Visual Novel Text Hooking Tools

[sanahtlig]

My edict2 is not gzipped.  Don't think a compressed file will work.

If the archive won't decompress it's probably corrupted.  That or security software / settings are interfering.

Edited December 3, 2015 by sanahtlig

[Mr. Meogii]

My edict2 is not gzipped.  Don't think a compressed file will work.

If the archive won't decompress it's probably corrupted.  That or security software / settings are interfering.

it's probably corrupted. But what should I do? 

[Tyr]

it's probably corrupted. But what should I do? 

Uh, ... maybe ... download it again?

[Zalor]

I feel that TA+ITH is a bit outdated. Ever since I switched to ChiitransLite, I've been enjoying reading VNs a lot more. I would recommend it over TA. For those interested, here is a tutorial on how to install it. 

ChiitansLite ITH tutorial

[sanahtlig]

I feel that TA+ITH is a bit outdated. Ever since I switched to ChiitransLite, I've been enjoying reading VNs a lot more. I would recommend it over TA. For those interested, here is a tutorial on how to install it. 

ChiitansLite ITH tutorial

I'm still not convinced.  The tools seem equivalent except for fullscreen capability, which I generally don't need.  Where is this improved user experience stemming from?

[Zalor]

I feel that TA+ITH is a bit outdated. Ever since I switched to ChiitransLite, I've been enjoying reading VNs a lot more. I would recommend it over TA. For those interested, here is a tutorial on how to install it. 

ChiitansLite ITH tutorial

I'm still not convinced.  The tools seem equivalent except for fullscreen capability, which I generally don't need.  Where is this improved user experience stemming from?

1) full screen compatibility is really nice for me.

2) Transparency mode is nice. It's nice to have my parser above (or in some cases below the actual text). As apposed to always having to look to the left (or right) of the VN. It's just annoying looking back and forth all the time.

3) I find that Chiitans deals with conjugations better (as well as with names). In conjugation heavy sentences, it's really nice how Chiitrans will change the color of grammatical extensions. It makes it easier (and faster) for me to determine how words are being used. All around I think it organizes the parsed sentences better in a way that helps my reading speed.

Too be honest with you, a part of me always kind of disliked reading VNs with TA, and it felt more like work than anything else. TA feels more like a study tool to me than an actual tool to leisurely read untld VNs with. With Chiitrans, I actually enjoy reading untld VNs, and it doesn't feel like im studying (even though I kind of am). 

[sanahtlig]

3. I suppose that's useful for beginners.  It certainly doesn't hurt (unless it's wrong).

Based on Tyr's response in the other thread, seems like the main advantage of TA is the ability to use 2 parsers at once.  Not sure how to enable multiple dictionaries, but Tyr seems to think it's possible.

I've tried overlays before and they're pretty clunky, since they interfere with the game interface.  That's a problem in gameplay titles (many of the VNs I play).

[Tyr]

In conjugation heavy sentences, it's really nice how Chiitrans will change the color of grammatical extensions. It makes it easier (and faster) for me to determine how words are being used.

In which way? My Chiitrans colors words exactly like JParser. Did I miss a feature? O.o

Not sure how to enable multiple dictionaries, but Tyr seems to think it's possible.

Yes, I use a Jp-->En and a Jp-->Ger dictionary at the same time. It's useful because both dictionaries have a different source thus one might cover idioms or something the other does not. I didn't found a Jp-->Jp dictionary, though. That's why I asked you in the other thread.

I noticed that JParser will always search every dictionary and then use the most greedy match (means the one with the most characters) while Mecab only uses simple word forms. Additionally, JParser will look for conjugations. Those conjugations are defined in the Conjugations.txt file. Both, Chiitrans and TA use this file, so if you think one of them is better at finding conjugations, you can just copy this file to the other tool and have the same algorithm there. Chiitrans uses an older version of this file, while the new TA has a more complex one. If you like the newer one, but  you want to use Chiitrans, just copy the new file included in the new TA over the old one in your Chiitrans data folder.

[sanahtlig]

 Yes, I use a Jp-->En and a Jp-->Ger dictionary at the same time. It's useful because both dictionaries have a different source thus some might cover idioms or something the other does not. I didn't found a Jp-->Jp dictionary, though. That's why I asked you in the other thread.

I found a JP->JP dictionary a few weeks ago, but I couldn't get it to work with Translation Aggregator.  I was attempting to to make JParser use edict2, and MeCab use this one.

https://en.osdn.jp/projects/naist-jdic/

 

Edited December 5, 2015 by sanahtlig

[Tyr]

That's not how TA works, I'm afraid. TA needs a dictionary in EDICT format. What dictionaries your Mecab installation uses is irrelevant. Mecab is only used for parsing.
If you want this dictionary to work with TA, you need to convert the CSV format into EDICT.

All the dictionaries you have imported into TA will be used at the same time for both parsers.

[Zalor]

In conjugation heavy sentences, it's really nice how Chiitrans will change the color of grammatical extensions. It makes it easier (and faster) for me to determine how words are being used.

In which way? My Chiitrans colors words exactly like JParser. Did I miss a feature? O.o

The only way I could explain it is with a picture. So here is a reference:

http://i938.photobucket.com/albums/ad226/Thezalor/Text hooker comparison 1.png

Chiitrans could understand that "掃除してくれていた" was all part of one word. But TA broke up that one word into three parts: "掃除して" "くれて" "いた". Which is just a pain in the ass. It is in this sense that I think that TA is weaker with compound conjugations.

[Tyr]

Indeed, that's interesting. Thank you for this example.

I wonder if Chiitrans knows that -te kureru and -te iru are conjugations with a special meaning. The popup only says "-te-Form --> -te-Form --> Past" which is true, but it does not mention how the kureru and the iru change the meaning of souji suru. The definitions of kurete and ita which you can look up in TA do tell you, however. (You wont understand it though, if you're Japanese is not even on the level that you know this already).

This actually affirms me in my conclusion that Chiitrans is very handy for experienced learner of Japanese. Having the whole conjugation instead of every part of it marked, is certainly easier to the eyes and better to read, even if you don't have this special meaning you probably already know anyway.
The best solution would be however, if Chiitrans would also give you the definition of -te kureru and -te iru, though. That's why I wondered if this parsing is hardcoded for this special forms. If it is, it shouldn't be hard to add this information which would make this look-up even better.

[sanahtlig]

This actually affirms me in my conclusion that Chiitrans is very handy for experienced learner of Japanese. Having the whole conjugation instead of every part of it marked, is certainly easier to the eyes and better to read.
The best solution would be however, if Chiitrans would also give you the definition of -te kureru and -te iru, though. That's why I wondered if this parsing is hardcoded for this special forms. If it is, it shouldn't be hard to add this information which would make this look-up even better.

So Chiitrans doesn't translate the conjugation?  That's actually one of my favorite aspects of JParser.  I never really learned grammar because JParser handles it for me or I glean it from context.

Edited December 6, 2015 by sanahtlig

[Tyr]

In this example, Chiitrans gives you only the translation for souji suru, not for kureru or iru. It only tells you that the whole conjugation consists of -te --> -te --> past. There is no way as far as I see to look up kureru or iru (and with this to look up the special -te meaning of these words) when it's part of the conjugation of another verb.

I think the conjugations you like in JParser are the ones defined in the Conjugations.txt. I think you can add even more, like for example specifically define a "-te iru" with a definition like "progressive form" or something and JParser would then use this and add the "iru" to the conjugation of the -te-verb and not as its own word.

[Down]

This actually affirms me in my conclusion that Chiitrans is very handy for experienced learner of Japanese. Having the whole conjugation instead of every part of it marked, is certainly easier to the eyes and better to read, even if you don't have this special meaning you probably already know anyway.

The point is a bit moot, if you're an experienced learner you already know the conjugations (and if you don't, relying on a parser to tell you how they work doesn't seem like the greatest idea anyway) and you don't need to look at your hooker.

[Tyr]

I would argue that even though you're experienced in the language, many people probably look at the hooker nonetheless, because reading nicely parsed and colored Japanese with furigana is easier than looking at the original text.

[Down]

Idk, I feel like you'd want to get rid of the crutches as soon as you can. One of the reason I had a hard time reading VNs with ITH/TA at first was because I spent too little time looking at the actual screen of the game. Chiitrans does give you the option of having a window superimposed on it, but there's still other issues. Parsers often fail on various things, get you the wrong furigana on a lot of words, etc... I can hardly see them as more convenient for reading.

Besides, the kind of parsing tools do is kinda trivial: parsing difficulties usually have to do with long groups of words, which parser tools can't help you with.

[Mephisto]

I just read the text in the text box. Having sentences broken down in to different color schemes seems kind of like a headache, and would affect how you read the lines. I like running with TA open in the background somewhere for looking up words I don't know and actually want to know. I can see how it might be useful if you aren't comfortable with parsing sentences yourself, but the sooner you do that the more enjoyable reading becomes. Playing without machine assistance also feels a lot different than playing with, it allows you to focus much more easily on the game. 

My guess on why it shows up differently in TA/chiitrans is that whatever dictionary TA is using simply doesn't recognize that conjugation (doesn't have an entry for it), while chiitrans does. It would be a lot of work to write an entirely different parsing algorithm, and you can also see in that picture that for some random reason TA has an entry for ショックだろう while chiitrans does not. (God knows why there's an entry for that) 

[Tyr]

My guess on why it shows up differently in TA/chiitrans is that whatever dictionary TA is using simply doesn't recognize that conjugation (doesn't have an entry for it), while chiitrans does. It would be a lot of work to write an entirely different parsing algorithm, and you can also see in that picture that for some random reason TA has an entry for ショックだろう while chiitrans does not. (God knows why there's an entry for that) 

With "God" I guess you mean me?
Because I explained this two times already in my posts above.

This has nothing to do with dictionary entries. Dictionaries don't include conjugations. Conjugations are defined in the Conjugation.txt file. For darou, we have for example:

"Name" : "v-ta-stem",
    "Part of Speech" : "Verb",
    "Tenses" : [
      {
        "Formal" : false, 
        "Negative" : false, 
        "Suffix" : "だろう", 
        "Tense" : "Past Conjectural\/Presumptive"
      }
    ]

or

"Name" : "adj-ta", 
    "Part of Speech" : "Adj", 
    "Tenses" : [
      {
        "Formal" : false, 
        "Negative" : false, 
        "Suffix" : "だろう", 
        "Tense" : "Volitional"
      }
    ]

If JParser finds the "Suffix" with the correct "Part of Speech", it will assume it's a conjugation of this "Tense".
 

[Rea_x_Chihiro]

Took me awhile to realize that this was fuwanovel. I thought i opened a random forum guide .

Btw, thanks for the guide. I really need this to learn japanese language a lot faster

 

[VirginSmasher]

Edict2 isn't working for me. It downloaded fine, but it gives me this if I try to extract it.

[sanahtlig]

Make sure the directory path isn't too long.  Try decompressing it on your desktop.

[VirginSmasher]

33 minutes ago, sanahtlig said:

Make sure the directory path isn't too long.  Try decompressing it on your desktop.

Decompressing the file didn't work. Also, what's different about decompressing it in the desktop?

 

I saved it as a .gz file because it was the only option. Does WinRar work with those type of files as well?

[Akirooge]

For some reason .7z archives doesn't open in my PC 

[sanahtlig]

WinRAR can handle GZip-compressed files, yes.  Not sure what to tell you except download it to a different directory and try again.