20 replies to this topic

[ebulerdo]

Hi Brent and group

At last I have learnt how to prepare a Bible module using Tooltip Tool 3.23.

However, I have noticed there's a problem with Spanish special characters. Maybe it's related to the fact Brent mentioned that T3 doesn't work in Unicode.

This is the sample text I'm importing:

#define abbreviation=TEST
#define description=TEST
#define comments=TEST
#define font=DEFAULT
#define ot=1
#define nt=0
#define apocrypha=0
#define strong=0
#define version=1

÷[Gen]
1:1 En el principio creó Dios los cielos y la tierra.
1:2 Y la tierra estaba desordenada y vacía, y las tinieblas estaban sobre la faz del abismo, y el Espíritu de Dios se movía sobre la faz de las aguas.
1:3 Y dijo Dios: Sea la luz; y fue la luz.
1:4 Y vio Dios que la luz era buena; y separó Dios la luz de las tinieblas.
1:5 Y llamó Dios a la luz Día, y a las tinieblas llamó Noche. Y fue la tarde y la mañana un día.

I save that as a .bblx file and I get no error messages. Now when I open my module with e-Sword, all the special characters (áéíóúñ) are gone:

Gén 1:1 En el principio cre Dios los cielos y la tierra.
Gén 1:2 Y la tierra estaba desordenada y vaca, y las tinieblas estaban sobre la faz del abismo, y el Espritu de Dios se mova sobre la faz de las aguas.
Gén 1:3 Y dijo Dios: Sea la luz; y fue la luz.
Gén 1:4 Y vio Dios que la luz era buena; y separ Dios la luz de las tinieblas.
Gén 1:5 Y llam Dios a la luz Da, y a las tinieblas llam Noche. Y fue la tarde y la maana un da.

When I open the module with SQLite Manager, I see the accented letters have been replaced by weird characters:

1 1 1 1 En el principio cre¢ Dios los cielos y la tierra.
2 1 1 2 Y la tierra estaba desordenada y vac¡a, y las tinieblas estaban sobre la faz del abismo, y el Esp¡ritu de Dios se mov¡a sobre la faz de las aguas.
3 1 1 3 Y dijo Dios: Sea la luz; y fue la luz.
4 1 1 4 Y vio Dios que la luz era buena; y separ¢ Dios la luz de las tinieblas.
5 1 1 5 Y llam¢ Dios a la luz D¡a, y a las tinieblas llam¢ Noche. Y fue la tarde y la ma¤ana un d¡a.

Do you see the funny ¢ ¡ ¤ symbols? I have tried using both ANSI and UTF8 text. Is there an option to fix this? Or if not, could it please be fixed in a future version? As it is now, it's not usable to build modules in Spanish. :-(

Thank you!

Carlos

[BH.]

Are you using the "Plain Text" way of making a BBLX or the "RTF" ? If you are using "Plain Text", then please try "RTF".

[ebulerdo]

Are you using the "Plain Text" way of making a BBLX or the "RTF" ? If you are using "Plain Text", then please try "RTF".

Hi Brent

Thank you for your help, I'm afraid I'm driving you crazy with my questions. O:-)

I was trying the "plain text" option since my text didn't contain any RTF codes. Following your suggestion I have tried the RTF option, and the results are better but not perfect: Special letters are converted to RTF codes, e.g. ó becomes \'f3. This looks fine since e-Sword correctly interprets those codes and displays the right characters. But when you try and search a word such as "creó", e-Sword won't find it. I had to type "cre\'f3" in order for e-Sword to find it. Obviously this is not an option. :-)

The problem lies in the RTF converter. Most RTF editors I've seen convert these letters to codes instead of keeping them.

I don't know if it would be possible to fix that. Or maybe it would be easier to fix the "plain text" option.

Thank you very much!

Carlos

[BH.]

Hi Brent

Thank you for your help, I'm afraid I'm driving you crazy with my questions. O:-)

I was trying the "plain text" option since my text didn't contain any RTF codes. Following your suggestion I have tried the RTF option, and the results are better but not perfect: Special letters are converted to RTF codes, e.g. ó becomes \'f3. This looks fine since e-Sword correctly interprets those codes and displays the right characters. But when you try and search a word such as "creó", e-Sword won't find it. I had to type "cre\'f3" in order for e-Sword to find it. Obviously this is not an option. :-)

The problem lies in the RTF converter. Most RTF editors I've seen convert these letters to codes instead of keeping them.

I don't know if it would be possible to fix that. Or maybe it would be easier to fix the "plain text" option.

Thank you very much!

Carlos

There is an option in T3 to view the RTF code. It is under "Options".

What you are saying is that e-Sword is displaying the characters correctly, but you can easily search, is that right???

If you have a sample file with the issue, perhaps I can look at it and see what my code is doing to it. It is a disadvantage to not be a native Spanish speaker and not have direct access to a Spanish version of Windows, etc... Also, T3 being ASCII based does not help, even though I've been able to make it preserve Unicode characters.

Understand that T3 started out just as tool to find and tooltip scripture references. I started out making modules just using e-Sword doing copy and paste. Time consuming, and fraught with errors. When e-Sword v9 came out, using SQLite3 databases, that is when I thought I'd try making module making easier. All my source documents I was working with were in English, so no problem. Then, after being pestered enough, I added Spanish as an "overlay" on my current code. I'm telling you this because T3's current state was one of small additions over time, and not a planned product from the very beginning. That fact that others find it useful is gratifying, but understand that programming is not my profession, and I've learned a lot just doing T3. I will try to continue to make "fixes" as I can. Having samples to work with is always most helpful...

BH.

[BH.]

Carlos - I took the data you have posted above, and created a BBLX in v3.24 of T3. And then generated an Installer (new feature in T3). I am posting it below. It will install a Bible, whose abbrevation is "TEST". It is only the first 5 verses of Genesis. The Spanish characters are preserved.

Before is the text I used in T3 to generate the BBLX file (which I called TEST.BBLX) and the extra code to create the Installer.

#define abbreviation=TEST
#define description=TEST
#define comments=TEST
#define font=DEFAULT
#define ot=1
#define nt=0
#define apocrypha=0
#define strong=0
#define version=1

#define IssModuleType=BBLX
#define IssExeName=TEST - BBLX
#define IssModuleName=TEST

÷[Gen]
1:1 En el principio creó Dios los cielos y la tierra.
1:2 Y la tierra estaba desordenada y vacía, y las tinieblas estaban sobre la faz del abismo, y el Espíritu de Dios se movía sobre la faz de las aguas.
1:3 Y dijo Dios: Sea la luz; y fue la luz.
1:4 Y vio Dios que la luz era buena; y separó Dios la luz de las tinieblas.
1:5 Y llamó Dios a la luz Día, y a las tinieblas llamó Noche. Y fue la tarde y la mañana un día.

[ebulerdo]

Carlos - I took the data you have posted above, and created a BBLX in v3.24 of T3. And then generated an Installer (new feature in T3). I am posting it below. It will install a Bible, whose abbrevation is "TEST". It is only the first 5 verses of Genesis. The Spanish characters are preserved.

#define abbreviation=TEST
#define description=TEST
#define comments=TEST
#define font=DEFAULT
#define ot=1
#define nt=0
#define apocrypha=0
#define strong=0
#define version=1

#define IssModuleType=BBLX
#define IssExeName=TEST - BBLX
#define IssModuleName=TEST

÷[Gen]
1:1 En el principio creó Dios los cielos y la tierra.
1:2 Y la tierra estaba desordenada y vacía, y las tinieblas estaban sobre la faz del abismo, y el Espíritu de Dios se movía sobre la faz de las aguas.
1:3 Y dijo Dios: Sea la luz; y fue la luz.
1:4 Y vio Dios que la luz era buena; y separó Dios la luz de las tinieblas.
1:5 Y llamó Dios a la luz Día, y a las tinieblas llamó Noche. Y fue la tarde y la mañana un día.

Hi Brent

Thank you very much for your help.

I tried the above code and, as you say, Spanish characters are visually kept. But actually Tooltip Tool has converted them to RTF codes. E.g. á (accented a) has become \'e1. This looks fine on screen, but when you search for a word like "Jehová", it will not match "Jehov\'e1".

Recently I tried converting a French bible and I stumbled upon the same problems. Eventually I had to use BeST, which has some issues of its own.

Since you wrote that answer one month ago I have been doing a lot of testing with T3, and I find it an extremely useful piece of software. But sadly it's not very helpful for building modules in other languages, unless some issues are fixed:

- When using plain text, special characters are converted to funny symbols.
- When using RTF text, they are converted to RTF codes.
- Tooltiping Bible references can find false positives when preceded by a unicode character (Remember "niños 1" tooltiped to "niñHos_1"?)

I have some ideas and I would like very much to contribute to fixing those issues. I will gladly provide any sample texts needed, character lists with equivalences, or whatever thing I can help with. I guess the solution would be to have it support unicode instead of ANSI, but maybe this is not that easy. Of course, I understand T3 is an amateur software, and even then these issues are not the main concern, so it's clear they won't be fixed at once. But I would like to help if you decide to have a look at them.

Thank you!

Carlos

[BH.]

The problem is that the tools I used do not support unicode. I have added some functions on my own, but I don't have the expertise for working with non-English character sets.

You said, "When using RTF text, they are converted to RTF codes". Yes, that is how the unicode is represented in RTF. The issue is not the RTF code per se (I think), but the "code points" chosen to display the characters. If you were to copy and paste Hebrew of Greek in a dictionary in e-Sword into T3, then look at the RTF code, you will see a similar appearance as with the Spanish characters T3 generates.

You said, "Tooltiping Bible references can find false positives when preceded by a unicode character (Remember "niños 1" tooltiped to "niñHos_1"?". Yes, this is a problem with not supporting Unicode 100%. One thing to di is to user your own Abbreviations file, and remove those abbreviations that are causing the false hits. That is a possible work around to the inabilities of T3.

I will continue to look at these issues, but I make no guarantees outside of English. Since the tools I use do not fully support Unicode, it may be that T3 will never work well with non-English text. The Bible tells us why we have these problems in Genesis 11.

[BH.]

BTW - The Spanish Bible I used to compare my generated text with was the SRV. And from what I see, it also can not be searched in e-Sword using the same characters. I will try another Spanish version.

***edit***

I just downloaded and installed the LBLA. I still can not search for the same characters. So what is going on? Carlos, please try the search in these versions.

Here i what I tried, in LBLA or SRV, selected "creó". I then clicked the Search button, which will show the word in the search dialog box. I then click the search button on the dialog box, and get nothing. So, the Spanish Bible produced by T3 is no different on my system. What am I doing wrong?

BTW, if you copy and paste Spanish text from SRV or LBLA, into T3, and look at the RTF code, it is the same as that produced by T3.

[patchworkid]

Hi please check this out http://www.omicronla...ngla-fonts.html try using this font and the software that goes with it. When I use different languages I use the fonts that goes with it. (Put the font in C:Windows - Fonts folder.)

thanks
Patchworkid

[BH.]

Hi please check this out http://www.omicronla...ngla-fonts.html try using this font and the software that goes with it. When I use different languages I use the fonts that goes with it. (Put the font in C:Windows - Fonts folder.)

thanks
Patchworkid

I'm using the fonts that e-Sword uses. Also, understand that the specific font used does not change the underlying code points, only the display of those characters.