73 replies to this topic

[macnuts]

Can't you map preferred language with txt subtitles characters? Like: preferred subs language 1 = Estonian ==> use Estonian characters in txt subtitles...

[zeros]

Hi macnuts!
Please read the first post.
Our problem has existed since the time when they came up with enigma2.
Our user eesatfan is trying to fix it at the moment but in my opinion it would be good to retain the option to set the force national characters set from menu.

[macnuts]

I was just wondering if you really need a new option.

 

You might try and use config.autolanguage.subtitle_autoselect1 instead of a new config.subtitles.teletext_subtitles_lang. Then inside void eDVBTeletextParser::setPageAndMagazine(int page, int magazine) body you could map selected language to txt table index in force_national_subset (defaulting to -1 for not supported languages).

 

But if this is not an option for you, go on with new settings

Edited by macnuts, 27 January 2014 - 07:11.

[zeros]

macnuts, Thank you, it might be useful comment for us!

[Rob van der Does]

You are aware that TT already has, and always had, it's own language settings (in TT press menu)?

[macnuts]

Of course. And it has always worked for me. But it does not work for zeros and others apparently.

[Rob van der Does]

I was not referring to the functionality, but to the location of the setting. As in the previous posting there was some discussion about where to locate a (new) localization setting for TT.

[zeros]

SatKiekerd, yes we know, but TT subtitles, they are really ugly.

[zeros]

Automatic mode is also working!!! Our user eesatfan did it and it is actually a completely new approach, without special menus. He has changed six different files (if I am correct) and the correct choice of characters is done automatically on the basis of language.
I'll write more tomorrow after testing. The code is not public at the moment but in Estonian forum is for testing IPK files for different boxes and several users confirms that all is working correctly.
I keep my fingers crossed that everything goes smoothly and that the code reaches to the reproduction.

[eesatfan]

Hi all

My idea is use language code from PMT table subtitle info for teletext pid for selecting correct charset when broadcaster ( in our country Viasat ) is not sending charset selecting triplets in X/28/0 or X/29/0. Select subtitle language is allso cached for every channel together with pid and page and magazine number. When just upgrading enigma2 then subtitle cache (if it exists) does not have language info cached and you should manually reselect subtitle ).

Additionally I implemented cyrillic charsets.

 

Here is the patch allso, hopefully somebody will commit it to git.

 subtitle.patch.gz   4.42KB   29 downloads

Edited by eesatfan, 2 February 2014 - 19:28.

[Persian Prince]

maybe helpful : https://github.com/f...ster/teletext.h

[eesatfan]

Yep, there is something to learn...... thanks

[rimas]

Hi all

My idea is use language code from PMT table subtitle info for teletext pid for selecting correct charset when broadcaster ( in our country Viasat ) is not sending charset selecting triplets in X/28/0 or X/29/0. Select subtitle language is allso cached for every channel together with pid and page and magazine number. When just upgrading enigma2 then subtitle cache (if it exists) does not have language info cached and you should manually reselect subtitle ).

Additionally I implemented cyrillic charsets.

 

Here is the patch allso, hopefully somebody will commit it to git.

subtitle.patch.gz

Please commit this patch. It does everything eesatfan wrote. That means finally normal teletext subtitles for estonians, latvians, lithuanians and russians.

 

Many thanks to eesatfan.

[Dimitrij]

Hi all

My idea is use language code from PMT table subtitle info for teletext pid for selecting correct charset when broadcaster ( in our country Viasat ) is not sending charset selecting triplets in X/28/0 or X/29/0. Select subtitle language is allso cached for every channel together with pid and page and magazine number. When just upgrading enigma2 then subtitle cache (if it exists) does not have language info cached and you should manually reselect subtitle ).

Additionally I implemented cyrillic charsets.

 

Here is the patch allso, hopefully somebody will commit it to git.

subtitle.patch.gz

Please commit this patch. It does everything eesatfan wrote. That means finally normal teletext subtitles for estonians, latvians, lithuanians and russians.

 

Many thanks to eesatfan.

+1

[Taapat]

I can confirm that the eesatfan patch works for Latvian Teletext subtitles. Thank you very much.
Only when compiling for my SH4 receiver I got error:

service/servicedvb.cpp: In member function ‘virtual RESULT eDVBServicePlay::enableSubtitles(iSubtitleUser*, iSubtitleOutput::SubtitleTrack&)’:
service/servicedvb.cpp:3055:8: error: name lookup of ‘ii’ changed for ISO ‘for’ scoping [-fpermissive]
service/servicedvb.cpp:3055:8: note: (if you use ‘-fpermissive’ G++ will accept your code) 

 

That's I correct with as follows:

--- a/lib/service/servicedvb.cpp
+++ b/lib/service/servicedvb.cpp
@@ -3049,7 +3049,8 @@ RESULT eDVBServicePlay::enableSubtitles(iSubtitleUser *user, SubtitleTrack &trac
   m_teletext_parser->setPageAndMagazine(page, magazine, lang.c_str());
   if (m_dvb_service)
   {
-   for (int ii=0; ii < m_teletext_parser->max_id; ii++){
+   int ii;
+   for (ii=0; ii < m_teletext_parser->max_id; ii++){
     if (!memcmp(m_teletext_parser->my_country_codes[ii], lang.c_str(), 3)) break;
    }
    if (ii > m_teletext_parser->max_id-2) ii = 0; 

 

It would be good to correct original patch.

[eesatfan]

I can confirm that the eesatfan patch works for Latvian Teletext subtitles. Thank you very much.
Only when compiling for my SH4 receiver I got error:

service/servicedvb.cpp: In member function ‘virtual RESULT eDVBServicePlay::enableSubtitles(iSubtitleUser*, iSubtitleOutput::SubtitleTrack&)’:
service/servicedvb.cpp:3055:8: error: name lookup of ‘ii’ changed for ISO ‘for’ scoping [-fpermissive]
service/servicedvb.cpp:3055:8: note: (if you use ‘-fpermissive’ G++ will accept your code) 

 

That's I correct with as follows:

--- a/lib/service/servicedvb.cpp
+++ b/lib/service/servicedvb.cpp
@@ -3049,7 +3049,8 @@ RESULT eDVBServicePlay::enableSubtitles(iSubtitleUser *user, SubtitleTrack &trac
   m_teletext_parser->setPageAndMagazine(page, magazine, lang.c_str());
   if (m_dvb_service)
   {
-   for (int ii=0; ii < m_teletext_parser->max_id; ii++){
+   int ii;
+   for (ii=0; ii < m_teletext_parser->max_id; ii++){
     if (!memcmp(m_teletext_parser->my_country_codes[ii], lang.c_str(), 3)) break;
    }
    if (ii > m_teletext_parser->max_id-2) ii = 0; 

 

It would be good to correct original patch.

 

Correcting has minor effect, there is a lot of such declarations in c++ code, I know that sh4 buildsystem does not allow -fpermissive’ flag and has lot less warnings , but openpli does. I have done some of programming for sh4 E2 in TDT times...

[urick]

Hi all

My idea is use language code from PMT table subtitle info for teletext pid for selecting correct charset when broadcaster ( in our country Viasat ) is not sending charset selecting triplets in X/28/0 or X/29/0. Select subtitle language is allso cached for every channel together with pid and page and magazine number. When just upgrading enigma2 then subtitle cache (if it exists) does not have language info cached and you should manually reselect subtitle ).

Additionally I implemented cyrillic charsets.

 

Here is the patch allso, hopefully somebody will commit it to git.

subtitle.patch.gz

 

Hi! How to apply this patch?

THX

[littlesat]

The patch format is incorrect....

The idea looks good...

 

$ git am -s ../subtitle.patch
Patch format detection failed.

 

And I do not understand why this is required...

+  {
+   for (int ii=0; ii < m_teletext_parser->max_id; ii++){
+    if (!memcmp(m_teletext_parser->my_country_codes[ii], lang.c_str(), 3)) break;
+   }
+   if (ii > m_teletext_parser->max_id-2) ii = 0;
+   m_dvb_service->setCacheEntry(eDVBService::cSUBTITLE,((pid&0xFFFF)<<16)|((page&0xFF)<<8)|((ii&0x1F)<<3)|(magazine&0x7));
+  }

Edited by littlesat, 6 February 2014 - 18:48.

[eesatfan]

The patch format is incorrect....

The idea looks good...

 

$ git am -s ../subtitle.patch
Patch format detection failed.

 

And I do not understand why this is required...

+  {
+   for (int ii=0; ii < m_teletext_parser->max_id; ii++){
+    if (!memcmp(m_teletext_parser->my_country_codes[ii], lang.c_str(), 3)) break;
+   }
+   if (ii > m_teletext_parser->max_id-2) ii = 0;
+   m_dvb_service->setCacheEntry(eDVBService::cSUBTITLE,((pid&0xFFFF)<<16)|((page&0xFF)<<8)|((ii&0x1F)<<3)|(magazine&0x7));
+  }

 

Hopefully this time is patch correctly formatted.

 subtitle.patch.gz   4.72KB   4 downloads

 

The code fragment stores used language code to Cache in 5 higher bits of magazine byte ( in teletext mode magazine has only 3 bits).

[eesatfan]

This should be correct now

 0001-This-patch-improves-teletext-subtitleing.patch.gz   5.03KB   20 downloads

 

Please ignore previous patch, this is allso not correct for use in only git way.

Edited by eesatfan, 6 February 2014 - 20:07.