WelcomeWelcome | FAQFAQ | DownloadsDownloads | WikiWiki

Author Topic: non-ASCII chars bring problems eg. in TextToSpeech  (Read 3726 times)

Offline labeas

  • Sr. Member
  • ****
  • Posts: 266
non-ASCII chars bring problems eg. in TextToSpeech
« on: April 15, 2017, 06:06:40 AM »
We see that TC has also adopted the absurd fad of using 3 bytes for
characters that have had adequate ASCII bytes since Mose's days.

echo "Tiny Cores" |hexdump -C ==
00000000  54 69 6e 79 20 43 6f 72  65 e2 80 99 73 0a        |Tiny Core...s.|

Or was it the <pdftotext TheTCbook> that introduced this garbage?
Since <pdftotext> HAS an option to render ASCII, does it read <blob>,
and then deliver either 3-bytes or Char(') depending on it's option-setting.
Ie. is it possible that Char(') was given to <textTopdf>, and 3bytes are
delivered by <pdftotxt>. So it's my <pdftotxt> that's creating the garbage?


Offline coreplayer2

  • Hero Member
  • *****
  • Posts: 3020
Re: non-ASCII chars bring problems eg. in TextToSpeech
« Reply #1 on: April 15, 2017, 04:31:53 PM »
TC is what you make it..

there are plenty of pdf viewers in the repo, including lightweight viewers capable of rendering a pdf as it was intended

mupdf from the x86 repo for example

Offline labeas

  • Sr. Member
  • ****
  • Posts: 266
Re: non-ASCII chars bring problems eg. in TextToSpeech
« Reply #2 on: April 21, 2017, 04:19:22 AM »
> there are plenty of pdf viewers in the repo, including lightweight viewers capable of
>   rendering a pdf as it was intended
.
 I'm talking about BEYOND as it was intended, to UNintended consequences.
 If my explanation was difficult to understand, consider:
 you're lying down, listening the TextToSpeech of some complex material.
 All the old stuff <talks correctly>. But the new smart-arse material
 says "they sea aye en tea see" instead of "they can't see".
.
>   mupdf from the x86 repo for example
Well my post wasn't about getting a pdf renderer, but since you named
one and I need one on TC, I tried to fetch it; but it's not listed in
my TC64-appsbrowser.


Offline polikuo

  • Hero Member
  • *****
  • Posts: 758
Re: non-ASCII chars bring problems eg. in TextToSpeech
« Reply #3 on: April 21, 2017, 04:30:34 AM »
I need one on TC, I tried to fetch it; but it's not listed in
my TC64-appsbrowser.

How about evince, flaxpdf, qpdf ?

Offline coreplayer2

  • Hero Member
  • *****
  • Posts: 3020
Re: non-ASCII chars bring problems eg. in TextToSpeech
« Reply #4 on: April 21, 2017, 11:10:54 AM »
Quote
...mupdf from the x86 repo for example
Well my post wasn't about getting a pdf renderer, but since you named
one and I need one on TC, I tried to fetch it; but it's not listed in
my TC64-appsbrowser.
It was an example because I'm not a mind reader...  The onus is on you to provide relevant information.  To that end, please specify which TC version and architecture you are referring too thank you

If the desired extension is missing from your specific repository, you're welcome to request an extension in the appropriate forum thread or compile it and submit for others who may be looking for the same program.

:)

Offline labeas

  • Sr. Member
  • ****
  • Posts: 266
Re: non-ASCII chars bring problems eg. in TextToSpeech
« Reply #5 on: April 28, 2017, 04:16:33 AM »
Since I initiated this thread in TC64 forum, it's assumed that I need TC64 extentions?
]]    I need one on TC, I tried to fetch it; but it's not listed in
]]     my TC64-appsbrowser.

]   How about evince, flaxpdf, qpdf ?

==> Log of install & test flaxpdf:--------
.................
.....
libpoppler is already installed!
libtiff is already installed!
libxcb is already installed!
libxshmfence is already installed!
lzo is already installed!
pcre is already installed!
pixman is already installed!
udev-lib is already installed!
=?=> imagine the redundancy for users who d/l all for very boot, as suggested?

tc@box:~$ sudo mc
# flaxpdf OneLaptop.pdf
libpng warning: iCCP: known incorrect sRGB profile
Visual doesn't match our expectations

# flaxpdf ch02.pdf
libpng warning: iCCP: known incorrect sRGB profile
Visual doesn't match our expectations

# flaxpdf ch02.pdf
libpng warning: iCCP: known incorrect sRGB profile
Visual doesn't match our expectations

???????????

Actualy <pdftotext> is more important than a pdf-renderer

Offline coreplayer2

  • Hero Member
  • *****
  • Posts: 3020
Re: non-ASCII chars bring problems eg. in TextToSpeech
« Reply #6 on: April 28, 2017, 10:48:17 AM »
Since I initiated this thread in TC64 forum, it's assumed that I need TC64 extentions?
You might have a point there,  appreciate that many folks don't actually use a web browser to view forum posts

Quote
]]    I need one on TC, I tried to fetch it; but it's not listed in
]]     my TC64-appsbrowser.

]   How about evince, flaxpdf, qpdf ?

==> Log of install & test flaxpdf:--------
.................
.....
libpoppler is already installed!
libtiff is already installed!
libxcb is already installed!
libxshmfence is already installed!
lzo is already installed!
pcre is already installed!
pixman is already installed!
udev-lib is already installed!
=?=> imagine the redundancy for users who d/l all for very boot, as suggested?
No-one has ever suggested folks download an extension and all it's dependencies on every reboot.   The intent of everyone's efforts is quite the opposite.   It's encouraged  to download into persistent storage which prevents having to download on every reboot.    Even those who use a CD to boot from are encouraged to remaster an ISO/CD to prevent this download again on reboot mayhem.

C'mon get with the program already and stop being resistant.  TC is not that difficult, do the right thing and download one time to persistent storage.      When you find a more suitable extension do some clean up of old extensions, deleting any unneeded items from your tce store and onboot/ondemand list.

Quote
=?=>
Here we go again..  I'm lost here on the meaning.


Quote
Actualy <pdftotext> is more important than a pdf-renderer
Really? To you maybe..  But consider this, the sole intent of a "Portable Document File" is to allow unbroken document formatting across multiple platforms.     Any compatible pdf reader will display the document as the creator intended and that is the most important aspect of any and all PDF's.
If you choose to deconstruct a pdf with pdf2text and it's not working as desired,  I suggest you take it up with it's creator.  Is pdf2text an extension?? 

Or even is pdf2text related to TextToSpeech??

« Last Edit: April 28, 2017, 11:12:25 AM by coreplayer2 »

Offline Rich

  • Administrator
  • Hero Member
  • *****
  • Posts: 11694
Re: non-ASCII chars bring problems eg. in TextToSpeech
« Reply #7 on: April 28, 2017, 11:13:32 AM »
Hi coreplayer2
Quote
Is pdf2text an extension??
It's part of  poppler-bin.tcz.

Offline coreplayer2

  • Hero Member
  • *****
  • Posts: 3020
Re: non-ASCII chars bring problems eg. in TextToSpeech
« Reply #8 on: April 28, 2017, 11:35:35 AM »
Cool, we have a tool for everything huh?

Offline labeas

  • Sr. Member
  • ****
  • Posts: 266
Re: non-ASCII chars bring problems eg. in TextToSpeech
« Reply #9 on: April 29, 2017, 11:03:04 PM »
>  Here we go again..  I'm lost here on the meaning.

OK, I'll try to simplify. I've raised several related problems.
Perhaps I should try to limit the scope?

>] Actualy <pdftotext> is more important than a pdf-renderer

> Really? To you maybe..  But consider this, the sole intent.....

Yes, for me.

> If you choose to deconstruct a pdf with pdf2text and it's not
> working as desired, I suggest you take it up with it's creator.
> Is pdf2text an extension??

I don't know. Perhaps not if all the FB/twitter kiddies don't
realise that ASCII-bytes and <pdf images> are as different as
a banana and a photo-of-a-mirror-image of a banana.

>
>  Or even is pdf2text related to TextToSpeech??

If you need the banana, the photo-of-a-mirror-image won't do.
======================
Perhaps I've already mentioned that I'd installed flaxpdf.

# flaxpdf OneLaptop.pdf
libpng warning: iCCP: known incorrect sRGB profile
Visual doesn't match our expectations

# flaxpdf ch02.pdf
libpng warning: iCCP: known incorrect sRGB profile
Visual doesn't match our expectations

# flaxpdf ch02.pdf
libpng warning: iCCP: known incorrect sRGB profile
Visual doesn't match our expectations
================
This relates to deep physics of hardware/color-pixels .........

Next time I want to speak with the engineer, not the sales clerk!