Translation process
14 views
Skip to first unread message
Sebastian Wagner
May 4, 2012, 9:22:08 AM5/4/12
to pg...@googlegroups.com
Hi all,
I consider translating the magazine to german as there's no German
translation team yet. I took a look at the translation process
<http://pgmag.org/how_to_translate>, the procedure described there
really shocked my. Do I really have to translate 323 single files
one-by-one? And the even more shocking information is that one: /"Now
copy/paste each translated txt files into the various content frames."/
In my opinion this is a Sisyphean challenge. It would be much easier and
faster to translate the texts directly in Scribus, even if it's /"not
appropriate for major text editing."/
It would be nice to hear from the existing translation teams what they
think about this topic.
King regards,
Sebastian Wagner
I consider translating the magazine to german as there's no German
translation team yet. I took a look at the translation process
<http://pgmag.org/how_to_translate>, the procedure described there
really shocked my. Do I really have to translate 323 single files
one-by-one? And the even more shocking information is that one: /"Now
copy/paste each translated txt files into the various content frames."/
In my opinion this is a Sisyphean challenge. It would be much easier and
faster to translate the texts directly in Scribus, even if it's /"not
appropriate for major text editing."/
It would be nice to hear from the existing translation teams what they
think about this topic.
King regards,
Sebastian Wagner
pirboazo
May 7, 2012, 10:44:37 AM5/7/12
to pg...@googlegroups.com
Hi Sebastian,
I look at how the French team is organise and how the translater of FCM are organize :
It is a manual process. Perhaps we shall think to automate creation.
Scribus format is an openFormat , It'is possible to extract text , #00 182 Textes not 363 :-) so we need to found how write text in Scribus files.
I imagine a translation processus as .
1 extract english text.
2 produce a new text ( german, français, Spanish ... )
3 Produce by program a new version.
4 New version is review by reader for the translation.
5 if text KO
They are modify and return at the step 3.
6 text are OK
The scribus spécialiste do the last modification....
A Dream.
Pierre
Pierre BOIZOT
May 7, 2012, 12:26:07 PM5/7/12
to pg...@googlegroups.com
Somebody know if it possible to tag text in scribus?
Pierre
Sebastian Wagner
May 7, 2012, 12:45:29 PM5/7/12
to pg...@googlegroups.com
Hi Pierre,
Thanks for your response!
The Link to the french wiki seems not very useful to me, at least I
could not find practical information.
I agree with you, that the text extraction and replacing should be done
automatically by a program. Maybe there's already an existing one,
otherwise we have to create it ourselves. I took a look at the Scribus
files and I think (I'm not sure), that every text is in /<ITEXT
CH="...">/ included. We could extract this (to files?) and replace it
afterwards. I asked in #scribus on Freenode if this would work (so, if I
am correctly informed about the sla-files), but haven't got an answer yet.
king regards,
Sebastian
Thanks for your response!
The Link to the french wiki seems not very useful to me, at least I
could not find practical information.
I agree with you, that the text extraction and replacing should be done
automatically by a program. Maybe there's already an existing one,
otherwise we have to create it ourselves. I took a look at the Scribus
files and I think (I'm not sure), that every text is in /<ITEXT
CH="...">/ included. We could extract this (to files?) and replace it
afterwards. I asked in #scribus on Freenode if this would work (so, if I
am correctly informed about the sla-files), but haven't got an answer yet.
king regards,
Sebastian
Pierre BOIZOT
May 8, 2012, 3:26:58 PM5/8/12
to pg...@googlegroups.com
Hi Sebastian,
I discuss with the manager of the French translation of FCM ( FullCirclemagazine ) , all activities,
extraction , translation , worflow management (wiki) , scribus edition
are manual execution.
With PCMAG the extraction is automated :-)
The next step could be the management of translation in a workflow.
Do you have response of #scribus ?
Sebastian Wagner
May 8, 2012, 3:54:45 PM5/8/12
to pg...@googlegroups.com
Hi Pierre,
Did you mean PgMag with "PCMAG"?
One guy in #scribus told me that "it will probably work" ;)
So, my idea is the following (roughly described):
I write a Python-Program, that examines the XML-Document and extracts
(by iterating through all tags) all the text-passages (definition:
dataset) to a file (i hope one dataset per line will work). So we now
have a file with all the texts, that has to be translated line-by-line.
A simple GUI-Interface could help translating by showing 2 Text-boxes,
one with the original english text, and one empty - the translation.
The translated datasets will be of course saved to another file, with
the same structure like the original one. The next step applies the
translation back to the XML, like in the beginning, but now we replace
the text with the new ones from the file.
It's also possible to save often occuring (like the licence text) and
short phrases in a cache to avoid redundant translations.
Theres one disadvantage: One textbox/dataset more or less in the
Scribus-File will break the translation-files and needs manual fixing.
But it's still ways better than doing everything by hand.
Nevertheless this could be avoided if every itext-tag has an unique id.
I just asked in #scribus if automatic setting of ids is possible. If
not, it could be done automatically by a program :D
King regards,
Sebastian
On 2012-05-08 21:26, Pierre BOIZOT wrote:
> Hi Sebastian,
>
> I discuss with the manager of the French translation of FCM (
> FullCirclemagazine ) , all activities,
> extraction , translation , worflow management (wiki) , scribus edition
> are manual execution.
>
> With PCMAG the extraction is automated :-)
>
> The next step could be the management of translation in a workflow.
>
> Do you have response of #scribus ?
>
> Pierre Boizot
>
> G+ <https://plus.google.com/u/0/107377830070954284209/about>
>
>
>
>
> 2012/5/7 Sebastian Wagner <sz...@gmx.at <mailto:sz...@gmx.at>>
Did you mean PgMag with "PCMAG"?
One guy in #scribus told me that "it will probably work" ;)
So, my idea is the following (roughly described):
I write a Python-Program, that examines the XML-Document and extracts
(by iterating through all tags) all the text-passages (definition:
dataset) to a file (i hope one dataset per line will work). So we now
have a file with all the texts, that has to be translated line-by-line.
A simple GUI-Interface could help translating by showing 2 Text-boxes,
one with the original english text, and one empty - the translation.
The translated datasets will be of course saved to another file, with
the same structure like the original one. The next step applies the
translation back to the XML, like in the beginning, but now we replace
the text with the new ones from the file.
It's also possible to save often occuring (like the licence text) and
short phrases in a cache to avoid redundant translations.
Theres one disadvantage: One textbox/dataset more or less in the
Scribus-File will break the translation-files and needs manual fixing.
But it's still ways better than doing everything by hand.
Nevertheless this could be avoided if every itext-tag has an unique id.
I just asked in #scribus if automatic setting of ids is possible. If
not, it could be done automatically by a program :D
King regards,
Sebastian
On 2012-05-08 21:26, Pierre BOIZOT wrote:
> Hi Sebastian,
>
> I discuss with the manager of the French translation of FCM (
> FullCirclemagazine ) , all activities,
> extraction , translation , worflow management (wiki) , scribus edition
> are manual execution.
>
> With PCMAG the extraction is automated :-)
>
> The next step could be the management of translation in a workflow.
>
> Do you have response of #scribus ?
>
> Pierre Boizot
>
>
>
>
>
> 2012/5/7 Sebastian Wagner <sz...@gmx.at <mailto:sz...@gmx.at>>
Pierre BOIZOT
May 9, 2012, 6:27:45 AM5/9/12
to pg...@googlegroups.com
Hi Sebastian,
Hi Pierre,
Did you mean PgMag with "PCMAG"?
Yes PGMAG ,
One guy in #scribus told me that "it will probably work" ;)
So, my idea is the following (roughly described):
I write a Python-Program, that examines the XML-Document and extracts
(by iterating through all tags) all the text-passages (definition:
dataset) to a file (i hope one dataset per line will work).
with a tag like TXT-01-023
I think in the same time we must produce the scribus document name with as a special extension ".tpl " where we remplace text-passages by a tag associate.
So we now have a file with all the texts, that has to be translated line-by-line.
A simple GUI-Interface could help translating by showing 2 Text-boxes,
one with the original english text, and one empty - the translation.
The translated datasets will be of course saved to another file, with
the same structure like the original one. The next step applies the
translation back to the XML, like in the beginning, but now we replace
the text with the new ones from the file.
We replace tag in the tpl file by the translate new text and save file on the new name.
A scribus guy can now modify the new document.
Pierre BOIZOT
May 9, 2012, 2:27:10 PM5/9/12
to pg...@googlegroups.com
Hi sebastian, Hi All
I created a small scribus file.
I analyzed scribus file .
extracted text and try to imagine a file structure ...
The texts translated are in directories, and we have a template file.
if the text was treated like a picture that this would be simple.
but we can create programme to replace the texts into file in place of the TAG.
The condition: the translation should not produce more paragraph.
we can put the text in db and create several attribute to manage the state , and lock ...
But perhaps it 's possible to do that with git?
@+
Sebastian Wagner
May 9, 2012, 2:59:16 PM5/9/12
to pg...@googlegroups.com
Hi Pierre,
Yesterday I met a friendly guy on #scribus, who could help us further,
since he gave me a python script that can be exectued inside Scribus
(you will have to run the script through script > execute script) and
extracts all Text passages. The other one does it the other way round,
it replaced the text from a file. They were originally created for
proofreading outside scribus. I haven't played around with it a lot yet,
but I think they work fine.
In a few months there will be a even better, a ingenious solution,
inside Scribus!
> [20:38] <a-l-e> the idea is to get scribus synchronize with a web
> interface
> [20:39] <a-l-e> it's mostly thought for proof reading, but should also
> be usable for translations.
> [20:39] <a-l-e> but i fear that it won't be ready for the next six months.
But for now, theres only a simple interface for translating and
proofreading missing (editing text files leads to distraction by other
lines).
I guess your solution is very similar to it, except that it might break
paragraph styles (thats why this script uses the scribus internals)
King regards,
Sebastian
On 2012-05-09 20:27, Pierre BOIZOT wrote:
> Hi sebastian, Hi All
>
> I created a small scribus file.
> I analyzed scribus file .
> extracted text and try to imagine a file structure ...
>
> The texts translated are in directories, and we have a template file.
>
> if the text was treated like a picture that this would be simple.
> but we can create programme to replace the texts into file in place of
> the TAG.
>
> The condition: the translation should not produce more paragraph.
>
> we can put the text in db and create several attribute to manage the
> state , and lock ...
> But perhaps it 's possible to do that with git?
>
> @+
>
> Pierre Boizot
>
> G+ <https://plus.google.com/u/0/107377830070954284209/about>
>
>
>
>
> 2012/5/8 Sebastian Wagner <sz...@gmx.at <mailto:sz...@gmx.at>>
> <mailto:sz...@gmx.at <mailto:sz...@gmx.at>>>
Yesterday I met a friendly guy on #scribus, who could help us further,
since he gave me a python script that can be exectued inside Scribus
(you will have to run the script through script > execute script) and
extracts all Text passages. The other one does it the other way round,
it replaced the text from a file. They were originally created for
proofreading outside scribus. I haven't played around with it a lot yet,
but I think they work fine.
In a few months there will be a even better, a ingenious solution,
inside Scribus!
> [20:38] <a-l-e> the idea is to get scribus synchronize with a web
> interface
> [20:39] <a-l-e> it's mostly thought for proof reading, but should also
> be usable for translations.
> [20:39] <a-l-e> but i fear that it won't be ready for the next six months.
But for now, theres only a simple interface for translating and
proofreading missing (editing text files leads to distraction by other
lines).
I guess your solution is very similar to it, except that it might break
paragraph styles (thats why this script uses the scribus internals)
King regards,
Sebastian
On 2012-05-09 20:27, Pierre BOIZOT wrote:
> Hi sebastian, Hi All
>
> I created a small scribus file.
> I analyzed scribus file .
> extracted text and try to imagine a file structure ...
>
> The texts translated are in directories, and we have a template file.
>
> if the text was treated like a picture that this would be simple.
> but we can create programme to replace the texts into file in place of
> the TAG.
>
> The condition: the translation should not produce more paragraph.
>
> we can put the text in db and create several attribute to manage the
> state , and lock ...
> But perhaps it 's possible to do that with git?
>
> @+
>
> Pierre Boizot
>
>
>
>
>
> 2012/5/8 Sebastian Wagner <sz...@gmx.at <mailto:sz...@gmx.at>>
Pierre BOIZOT
May 10, 2012, 3:22:34 AM5/10/12
to pg...@googlegroups.com
Hi Sebastian,
Pierre Boizot
The result is not good....
The import modify too much information.
<PAGEOBJECT OwnPage="0" PTYPE="4" XPOS="720.949921259843" YPOS="131.50125984252" WIDTH="537.5" HEIGHT="101.5" RADRECT="0" FRTYPE="0" CLIPEDIT="0" PWIDTH="0" PCOLOR="None" PCOLOR2="None" COLUMNS="1" COLGAP="0" NAMEDLST="" SHADE="100" SHADE2="100" GRTYP="0" ROT="0" PLINEART="1" PLINEEND="0" PLINEJOIN="0" LOCALSCX="1" LOCALSCY="1" LOCALX="0" LOCALY="0" PICART="1" PLTSHOW="0" BASEOF="0" textPathType="0" textPathFlipped="0" FLIPPEDH="0" FLIPPEDV="0" SCALETYPE="1" RATIO="1" PRINTABLE="1" ANNOTATION="0" ANNAME="" TEXTFLOWMODE="0" TEXTFLOW="0" TEXTFLOW2="0" TEXTFLOW3="0" AUTOTEXT="0" EXTRA="0" TEXTRA="0" BEXTRA="0" REXTRA="0" FLOP="0" PFILE="" PFILE2="" PFILE3="" PRFILE="" EPROF="" IRENDER="1" EMBEDDED="1" LOCK="0" LOCKR="0" REVERS="0" TransValue="0" TransValueS="0" TransBlend="0" TransBlendS="0" isTableItem="0" TopLine="0" LeftLine="0" RightLine="0" BottomLine="0" isGroupControl="0" NUMDASH="0" DASHS="" DASHOFF="0" NUMPO="16" POCOOR="0 0 0 0 537.5 0 537.5 0 537.5 0 537.5 0 537.5 101.5 537.5 101.5 537.5 101.5 537.5 101.5 0 101.5 0 101.5 0 101.5 0 101.5 0 0 0 0 " NUMCO="16" COCOOR="0 0 0 0 537.5 0 537.5 0 537.5 0 537.5 0 537.5 101.5 537.5 101.5 537.5 101.5 537.5 101.5 0 101.5 0 101.5 0 101.5 0 101.5 0 0 0 0 " NUMGROUP="0" GROUPS="" startArrowIndex="0" endArrowIndex="0" OnMasterPage="" ImageClip="" ImageRes="1" Pagenumber="0" isInline="0" fillRule="1" doOverprint="0" gXpos="0" gYpos="0" gWidth="0" gHeight="0" LAYER="0" BOOKMARK="0" NEXTITEM="-1" BACKITEM="-1">
<ITEXT CH="Performance Tuning :"/>
<para PARENT="Front_H1"/>
<ITEXT CH=" To infinity..... and beyond !"/>
<trail PARENT="Front_H1_subtitle"/>
<PageItemAttributes/>
</PAGEOBJECT>
is replace by.
<PAGEOBJECT OwnPage="0" PTYPE="4" XPOS="720.949921259843" YPOS="131.50125984252" WIDTH="537.5" HEIGHT="101.5" RADRECT="0" FRTYPE="0" CLIPEDIT="0" PWIDTH="0" PCOLOR="None" PCOLOR2="None" COLUMNS="1" COLGAP="0" NAMEDLST="" SHADE="100" SHADE2="100" GRTYP="0" ROT="0" PLINEART="1" PLINEEND="0" PLINEJOIN="0" LINESP="15" LOCALSCX="1" LOCALSCY="1" LOCALX="0" LOCALY="0" PICART="1" PLTSHOW="0" BASEOF="0" textPathType="0" textPathFlipped="0" FLIPPEDH="0" FLIPPEDV="0" SCALETYPE="1" RATIO="1" PRINTABLE="1" ANNOTATION="0" ANNAME="" TEXTFLOWMODE="0" TEXTFLOW="0" TEXTFLOW2="0" TEXTFLOW3="0" AUTOTEXT="0" EXTRA="0" TEXTRA="0" BEXTRA="0" REXTRA="0" FLOP="0" PFILE="" PFILE2="" PFILE3="" PRFILE="" EPROF="" IRENDER="1" EMBEDDED="1" LOCK="0" LOCKR="0" REVERS="0" TransValue="0" TransValueS="0" TransBlend="0" TransBlendS="0" isTableItem="0" TopLine="0" LeftLine="0" RightLine="0" BottomLine="0" isGroupControl="0" NUMDASH="0" DASHS="" DASHOFF="0" NUMPO="16" POCOOR="0 0 0 0 537.5 0 537.5 0 537.5 0 537.5 0 537.5 101.5 537.5 101.5 537.5 101.5 537.5 101.5 0 101.5 0 101.5 0 101.5 0 101.5 0 0 0 0 " NUMCO="16" COCOOR="0 0 0 0 537.5 0 537.5 0 537.5 0 537.5 0 537.5 101.5 537.5 101.5 537.5 101.5 537.5 101.5 0 101.5 0 101.5 0 101.5 0 101.5 0 0 0 0 " NUMGROUP="0" GROUPS="" startArrowIndex="0" endArrowIndex="0" OnMasterPage="" ImageClip="" ImageRes="1" Pagenumber="0" isInline="0" fillRule="1" doOverprint="0" gXpos="0" gYpos="0" gWidth="0" gHeight="0" LAYER="0" BOOKMARK="0" NEXTITEM="-1" BACKITEM="-1">
<ITEXT CH="Performance Tuning :"/>
<para LINESP="15"/>
<ITEXT CH=" To infinity..... and beyond !"/>
<trail LINESP="15"/>
<PageItemAttributes/>
So the layout is lost :-) we must modify just the TEXT.
Pierre Boizot
2012/5/9 Sebastian Wagner <sz...@gmx.at>
Reply all
Reply to author
Forward
0 new messages