Loading ...
Sorry, an error occurred while loading the content.

tags and grammatical reversals and other complexities

Expand Messages
  • anthony baldwin
    Just out of curiosity, what do you do with something like this: N ° de SOLICITUD N ° APPLICATION
    Message 1 of 11 , Feb 1, 2008
    View Source
    • 0 Attachment
      Just out of curiosity, what do you do with something like this:

      N</f0><f1>°</f1> de <f2> SOLICITUD</f2>
      <segment 0016> N</f0><f1>°</f1><f2> APPLICATION</f2> <end segment>

      ??

      Obviously (well, at least for those who speak Spanish and English),
      I really want the translated segment to read:

      Application No.

      But, either the tags or the formatting get messed up if I rearrange
      the segment.
      In general, I do as I did above, and then edit the final draft
      in OpenOffice, but, I have to carefully edit, since spellchecking
      doesn't catch stuff like this, and it annoys me, and I'm wondering if
      anyone has a better idea, of if you just deal with it like I do, and, do
      you split up run-on sentences like this one when you are translating,
      since they are also quite annoying?

      Clearly, this is an issue in any language pair that has significant
      differences, even if they are as simple as adjective placement, which is
      an issue in all of my pairs at some point or another
      (PT/EN, ES/EN, FR/EN).

      I have no idea what complexities face those who translate, for instance,
      between English and Chinese or Russian or Arabic face, or other
      languages, but, I can only imagine, the languages being quite different,
      that there are similar issues, and, anyone translating between English
      and any "normal" latin-based language, of course, has precisely these
      same issues I have, in one direction or another, and also probably has
      to deal with these ridiculous, long-winded, run-on sentences that I seem
      to be currently in the habit of writing in e-mails, even though I would
      never write them in any formal document, such an article, or something,
      or an academic paper for a class, which is another story all together,
      isn't it?

      /tony

      --
      Anthony Baldwin
      http://www.BaldwinLinguas.com - Translation & Interpreting

      http://www.LinguasOS.org - Linux for Translators

      Así también, la lengua es un miembro pequeño,
      y se gloría de grandes cosas.
      He aquí, ¡un pequeño fuego ­cuán grande bosque enciende!
    • Richard/g
      ... ... It s a pain. I usually do something like the following: APPLICATION N° Lots of
      Message 2 of 11 , Feb 1, 2008
      View Source
      • 0 Attachment
        On Fri 01 Feb 2008, anthony baldwin wrote:
        >
        > Just out of curiosity, what do you do with something
        like this:
        >
        > N</f0><f1>°</f1> de <f2> SOLICITUD</f2>
        > <segment 0016> N</f0><f1>°</f1><f2> APPLICATION</f2>
        <end segment>
        >
        > ??

        It's a pain. I usually do something like the following:

        <segment 0016> </f0><f1></f1><f2>APPLICATION N°</f2> <end
        segment>

        Lots of fiddling/editing but that's the price of having
        the translation memory.

        > Obviously (well, at least for those who speak Spanish
        and English),
        > I really want the translated segment to read:
        >
        > Application No.
        >
        > But, either the tags or the formatting get messed up if
        I rearrange
        > the segment.
        > In general, I do as I did above, and then edit the
        final draft
        > in OpenOffice, but, I have to carefully edit, since
        spellchecking
        > doesn't catch stuff like this, and it annoys me, and
        I'm wondering if
        > anyone has a better idea, of if you just deal with it
        like I do, and, do
        > you split up run-on sentences like this one when you
        are translating,
        > since they are also quite annoying?

        Sometimes, it is easier to remove most formatting from
        the original and define a style that fits and apply it.

        I guess that is why I do editing for the sidux manual
        directly in the html. Did the original translation in
        OmT but it's easier to do additions directly in the
        file --though no TM.

        > Clearly, this is an issue in any language pair that has
        significant
        > differences, even if they are as simple as adjective
        placement, which is
        > an issue in all of my pairs at some point or another
        > (PT/EN, ES/EN, FR/EN).
        >
        > I have no idea what complexities face those who
        translate, for instance,
        > between English and Chinese or Russian or Arabic face,
        or other
        > languages, but, I can only imagine, the languages being
        quite different,
        > that there are similar issues, and, anyone translating
        between English
        > and any "normal" latin-based language, of course, has
        precisely these
        > same issues I have, in one direction or another, and
        also probably has
        > to deal with these ridiculous, long-winded, run-on
        sentences that I seem
        > to be currently in the habit of writing in e-mails,
        even though I would
        > never write them in any formal document, such an
        article, or something,
        > or an academic paper for a class, which is another
        story all together,
        > isn't it?

        Removing all page formatting except for tables,
        sure makes it nicer to type in the text without dodging
        the tags. Seems twice as fast to type as to edit.

        saludos,
        Richard.


        > /tony
        >
        > --
        > Anthony Baldwin
        > http://www.BaldwinLinguas.com - Translation &
        Interpreting
        >
        > http://www.LinguasOS.org - Linux for Translators
        >
        > Así también, la lengua es un miembro pequeño,
        > y se gloría de grandes cosas.
        > He aquí, ¡un pequeño fuego ­cuán grande bosque
        enciende!
        >
      • Marc Prior
        Tony, I m not sure exactly what the problem is. Is it that you want the order of N° and SOLICITUD to be reversed, but they both have in-line formatting, which
        Message 3 of 11 , Feb 1, 2008
        View Source
        • 0 Attachment
          Tony,

          I'm not sure exactly what the problem is.

          Is it that you want the order of N° and SOLICITUD to be reversed, but
          they both have in-line formatting, which prevents that?

          If so, this should be possible:

          </f0><f2>APPLICATION</f2> N<f1>°</f1>

          The tag validator will throw an error, but the compiled document should
          open OK.

          Or is it that you have the character <f1>°</f1> in the original, which
          is formatted in a certain way (I'm guessing: underlined?), and you don't
          want corresponding formatting in the target? That's easy too: you simply
          delete everything between the relevant tags, so:

          </f0><f1></f1><f2>APPLICATION</f2> No.

          Sometimes it isn't possible to re-arrange the sequence of formatting
          exactly how you want it. German is probably worse than Spanish for
          changing the word order, see here:

          www.crossmyt.com/hc/linghebr/awfgrmlg.html

          If you don't have time to read all of it, just search for "There are ten
          parts of speech".

          If all else fails, I have two workarounds:

          1. Put a marker in the text (I use ##) at points requiring post-editing.
          2. Edit the source text. It's no different really to editing the target
          text, except that you can do it on the fly and don't then have to
          remember to edit the target later. Be careful about editing the source
          text after you've done any significant amount of work: what you've done
          so far won't be lost and will show up as fuzzy matches when you reload
          the project, but the formatting tags will renumber themselves and it can
          be a heck of a lot of work to correct them all again.

          I split and join sentences, yes; if the syntax is broadly similiar, a
          conjunction replacing the full stop (period) or vice-versa, it's no big
          deal, but sometimes the syntax is such that how much of the text goes in
          one segment and how much goes in the other can be fairly arbitrary.
          Still no big deal, really, but it can leave you with some "strange"
          segment pairs in your TM.

          By the way, I have attempted to learn nine foreign languages at various
          parts of my life (not counting Arabic, with which I didn't get much
          further than the pronunciation, which in the book I was using involved
          descriptions like "this is like the sound a camel makes whilst it is
          being loaded"), but if you're wondering why Spanish isn't one of the
          nine, I tell people that it would have been, if only I had been 14 and
          not 43 when I first heard Juanes' "Fotografía". How tragic...

          Marc

          anthony baldwin schrieb:
          > Just out of curiosity, what do you do with something like this:
          >
          > N</f0><f1>°</f1> de <f2> SOLICITUD</f2>
          > <segment 0016> N</f0><f1>°</f1><f2> APPLICATION</f2> <end segment>
          >
          > ??
          >
          > Obviously (well, at least for those who speak Spanish and English),
          > I really want the translated segment to read:
          >
          > Application No.
          >
          > But, either the tags or the formatting get messed up if I rearrange
          > the segment.
          > In general, I do as I did above, and then edit the final draft
          > in OpenOffice, but, I have to carefully edit, since spellchecking
          > doesn't catch stuff like this, and it annoys me, and I'm wondering if
          > anyone has a better idea, of if you just deal with it like I do, and, do
          > you split up run-on sentences like this one when you are translating,
          > since they are also quite annoying?
          >
          > Clearly, this is an issue in any language pair that has significant
          > differences, even if they are as simple as adjective placement, which is
          > an issue in all of my pairs at some point or another
          > (PT/EN, ES/EN, FR/EN).
          >
          > I have no idea what complexities face those who translate, for instance,
          > between English and Chinese or Russian or Arabic face, or other
          > languages, but, I can only imagine, the languages being quite different,
          > that there are similar issues, and, anyone translating between English
          > and any "normal" latin-based language, of course, has precisely these
          > same issues I have, in one direction or another, and also probably has
          > to deal with these ridiculous, long-winded, run-on sentences that I seem
          > to be currently in the habit of writing in e-mails, even though I would
          > never write them in any formal document, such an article, or something,
          > or an academic paper for a class, which is another story all together,
          > isn't it?
          >
          > /tony
          >
        • anthony baldwin
          ... That would be it. ... Hmmm...I didn t know I could do that. Then again, if there is a problem opening the target document, and I ve done this to numerous
          Message 4 of 11 , Feb 1, 2008
          View Source
          • 0 Attachment
            Marc Prior wrote:
            > Tony,
            >
            > I'm not sure exactly what the problem is.
            >
            > Is it that you want the order of N° and SOLICITUD to be reversed, but
            > they both have in-line formatting, which prevents that?

            That would be it.

            >
            > If so, this should be possible:
            >
            > </f0><f2>APPLICATION</f2> N<f1>°</f1>
            >
            > The tag validator will throw an error, but the compiled document should
            > open OK.

            Hmmm...I didn't know I could do that.
            Then again, if there is a problem opening the target document, and I've
            done this to numerous segments, running the tag validator won't be much
            help in resolving the issue.

            Understand, I'm not saying that this is in anyway a problem with OmegaT,
            just a headache when working with languages with different ideas on
            grammatical structure.
            I was curious how other folks dealt with it.
            It never occurred to be to think of it as a software shortcoming.


            >
            > Sometimes it isn't possible to re-arrange the sequence of formatting
            > exactly how you want it. German is probably worse than Spanish for
            > changing the word order, see here:
            >
            > www.crossmyt.com/hc/linghebr/awfgrmlg.html
            >
            > If you don't have time to read all of it, just search for "There are ten
            > parts of speech".


            See...I knew others had similar, or worse, headaches to deal with...

            >
            > If all else fails, I have two workarounds:
            >
            > 1. Put a marker in the text (I use ##) at points requiring post-editing.

            You know, I do that in other instances.
            When I have < or > as a math symbol and it's messing with tags,
            I replace it with GREATHERTHAN!!!, which I can easily find in the target
            text. Actually, with that, I can run a simple find/replace in OOo.


            > 2. Edit the source text. It's no different really to editing the target
            > text, except that you can do it on the fly and don't then have to
            > remember to edit the target later. Be careful about editing the source
            > text after you've done any significant amount of work: what you've done
            > so far won't be lost and will show up as fuzzy matches when you reload
            > the project, but the formatting tags will renumber themselves and it can
            > be a heck of a lot of work to correct them all again.
            >
            > I split and join sentences, yes; if the syntax is broadly similiar, a
            > conjunction replacing the full stop (period) or vice-versa, it's no big
            > deal, but sometimes the syntax is such that how much of the text goes in
            > one segment and how much goes in the other can be fairly arbitrary.
            > Still no big deal, really, but it can leave you with some "strange"
            > segment pairs in your TM.
            >
            > By the way, I have attempted to learn nine foreign languages at various
            > parts of my life (not counting Arabic, with which I didn't get much
            > further than the pronunciation, which in the book I was using involved
            > descriptions like "this is like the sound a camel makes whilst it is
            > being loaded"), but if you're wondering why Spanish isn't one of the
            > nine, I tell people that it would have been, if only I had been 14 and
            > not 43 when I first heard Juanes' "Fotografía". How tragic...\


            Así es la vida....

            The Puerto Rican girls here in my home state were enough motiviation for
            me to learn Spanish...

            Thanks,
            /tony




            --
            Anthony Baldwin
            http://www.BaldwinLinguas.com - Translation & Interpreting

            http://www.LinguasOS.org - Linux for Translators

            Así también, la lengua es un miembro pequeño,
            y se gloría de grandes cosas.
            He aquí, ¡un pequeño fuego ­cuán grande bosque enciende!
          • Marc Prior
            ... Generally, tags should be preserved in the same order, but there is some wiggle room . The manual provides details. The tag validator will flag segments
            Message 5 of 11 , Feb 2, 2008
            View Source
            • 0 Attachment
              anthony baldwin schrieb:
              >
              > Hmmm...I didn't know I could do that.
              > Then again, if there is a problem opening the target document, and I've
              > done this to numerous segments, running the tag validator won't be much
              > help in resolving the issue.

              Generally, tags should be preserved in the same order, but there is some
              "wiggle room". The manual provides details. The tag validator will flag
              segments in which the source and target segment tags differ; it's left
              to you though to decide whether a deviation is deliberate or not. You
              may have deviated deliberately in one segment but have accidentally
              deleted or corrupted a tag in another segment, and it's the latter that
              is causing the problem. You won't know whether your modification to the
              tag order in a particular segment is acceptable to OOo until you've
              corrected unwanted tag errors in all the other segments. So in that
              respect, the tag validator is essential.

              > You know, I do that in other instances.
              > When I have < or > as a math symbol and it's messing with tags,
              > I replace it with GREATHERTHAN!!!, which I can easily find in the target
              > text. Actually, with that, I can run a simple find/replace in OOo.

              What version of OmT are you using? I thought that particular bug had
              been fixed.

              > The Puerto Rican girls here in my home state were enough motiviation for
              > me to learn Spanish...

              ;-)

              Marc
            • andrius
              ... IMO, the folks simply need to remember what has been stated long ago in GNU gettext manual: don t split sentences (the exact wording may be different:-).
              Message 6 of 11 , Feb 2, 2008
              View Source
              • 0 Attachment
                Pn, 2008 02 01 22:18 -0500, anthony baldwin rašė:

                > Understand, I'm not saying that this is in anyway a problem with
                > OmegaT,
                > just a headache when working with languages with different ideas on
                > grammatical structure.
                > I was curious how other folks dealt with it.
                > It never occurred to be to think of it as a software shortcoming.

                IMO, the folks simply need to remember what has been stated long ago in
                GNU gettext manual: don't split sentences (the exact wording may be
                different:-). Then, if the document is cleanly formatted, it's easy to
                reorder, delete or duplicate tag pairs, at least in OmegaT.

                However, word processors and peolpe using them have amazing abilities to
                produce tag mess (redundant formatting). OmegaT hides some of it from
                you, but it cannot hide it all or fix it. If I get such a document, I
                first clean it directly, e.g. in content.xml of an ODF file.
                There is some OpenOffice plugin for this, called "slyle overload
                cleaner" or something (French interface), but I haven't used it.

                For EN->LT, I often reorder, delete and sometimes duplicate tag pairs.
                So, in heavily though cleanly formated translation, OmegaT validation
                tool may report about one hundred tag errors, most of them "false
                positives".
                I think it would be nice to have a "light" validation option, that
                doesn't report deleted, added or reordered tag pairs. Pardon me if it's
                already implemented or there is an RFE – haven't followed the latest
                development.

                Andrius
              • Marc Prior
                ... This sounds very useful, Andrius - do you have a link for it? Marc
                Message 7 of 11 , Feb 2, 2008
                View Source
                • 0 Attachment
                  andrius schrieb:

                  > There is some OpenOffice plugin for this, called "slyle overload
                  > cleaner" or something (French interface), but I haven't used it.

                  This sounds very useful, Andrius - do you have a link for it?

                  Marc
                • anthony baldwin
                  ... That s what I mean...If I ve deliberately changes tags, and made errors in other segments, I could spend a day checking the tags to find the errors.... ...
                  Message 8 of 11 , Feb 2, 2008
                  View Source
                  • 0 Attachment
                    Marc Prior wrote:
                    > anthony baldwin schrieb:
                    >> Hmmm...I didn't know I could do that.
                    >> Then again, if there is a problem opening the target document, and I've
                    >> done this to numerous segments, running the tag validator won't be much
                    >> help in resolving the issue.
                    >
                    > Generally, tags should be preserved in the same order, but there is some
                    > "wiggle room". The manual provides details. The tag validator will flag
                    > segments in which the source and target segment tags differ; it's left
                    > to you though to decide whether a deviation is deliberate or not. You
                    > may have deviated deliberately in one segment but have accidentally
                    > deleted or corrupted a tag in another segment, and it's the latter that
                    > is causing the problem. You won't know whether your modification to the
                    > tag order in a particular segment is acceptable to OOo until you've
                    > corrected unwanted tag errors in all the other segments. So in that
                    > respect, the tag validator is essential.

                    That's what I mean...If I've deliberately changes tags, and made
                    errors in other segments, I could spend a day checking the tags
                    to find the errors....

                    >
                    >> You know, I do that in other instances.
                    >> When I have < or > as a math symbol and it's messing with tags,
                    >> I replace it with GREATHERTHAN!!!, which I can easily find in the target
                    >> text. Actually, with that, I can run a simple find/replace in OOo.
                    >
                    > What version of OmT are you using? I thought that particular bug had
                    > been fixed.


                    Well, I got into that habit using 1.6.4 (I think).
                    I am currently using 1.7.2, since that's what's in the PCLOS repos.
                    I suppose it's no longer necessary then?

                    >
                    >> The Puerto Rican girls here in my home state were enough motiviation for
                    >> me to learn Spanish...
                    >
                    > ;-)
                    >
                    > Marc
                    >
                    >

                    Do you remember...I wrote to this list just a little over a year ago
                    after having first downloaded OmegaT, confused, and not quite sure how
                    using OmegaT was going to be useful to me.
                    I thought I was working more efficiently just editing documents in OOo.

                    At this juncture, I'm not quite sure how I got along without OmegaT.
                    The increase in efficiency contributed deeply to the fact that my income
                    in 2007 was DOUBLE what is was in 2006 (other factors contributed, too,
                    but, OmegaT was HUGE factor).
                    OmegaT, OOo, Firefox and Thunderbird are my four most used applications.
                    Then Evince(pdfviewer), medit, xmms, and, of course, kshisen...

                    I'm helping out with l10n, writing scripts, and trying to learn java and
                    python to become more useful to the project (and for my own ends, as
                    well). The java is going slowly...I like python better.


                    thanks for your assistance, as always
                    tony

                    --
                    Anthony Baldwin
                    http://www.BaldwinLinguas.com - Translation & Interpreting

                    http://www.LinguasOS.org - Linux for Translators

                    Así también, la lengua es un miembro pequeño,
                    y se gloría de grandes cosas.
                    He aquí, ¡un pequeño fuego ­cuán grande bosque enciende!
                  • Marc Prior
                    ... How often do you actually need to change the order of formatting in a text? And how often do you make tag errors? I can probably go months without needing
                    Message 9 of 11 , Feb 2, 2008
                    View Source
                    • 0 Attachment
                      anthony baldwin schrieb:
                      >
                      > That's what I mean...If I've deliberately changes tags, and made
                      > errors in other segments, I could spend a day checking the tags
                      > to find the errors....

                      How often do you actually need to change the order of formatting in a
                      text? And how often do you make tag errors?

                      I can probably go months without needing to change the order of
                      formatting. I'm currently working on the second draft of a 100,000-word
                      job, and I think there's just one instance of formatting reversal in the
                      whole text. A quick Ctrl-T shows that I've got 26 segments with tag
                      errors, so when I get round to correcting them, I'll have to make sure
                      that the one intentional instance is left untouched. But I'm unlikely to
                      "correct" it by mistake; my tag errors rarely involve tag inversion, so
                      when they do I notice it. My tag errors tend to be omitted tags,
                      duplicate tags, or damaged tags (e.g. snipping off one of the brackets
                      by mistake).

                      Marc
                    • anthony baldwin
                      ... Frequently And how often do you make tag errors? Rarely ... Really, it s not a big problem, just a minor annoyance, and I wonder what others did. And,
                      Message 10 of 11 , Feb 2, 2008
                      View Source
                      • 0 Attachment
                        Marc Prior wrote:
                        > anthony baldwin schrieb:
                        >> That's what I mean...If I've deliberately changes tags, and made
                        >> errors in other segments, I could spend a day checking the tags
                        >> to find the errors....
                        >
                        > How often do you actually need to change the order of formatting in a
                        > text?

                        Frequently

                        And how often do you make tag errors?

                        Rarely

                        >
                        > I can probably go months without needing to change the order of
                        > formatting. I'm currently working on the second draft of a 100,000-word
                        > job, and I think there's just one instance of formatting reversal in the
                        > whole text. A quick Ctrl-T shows that I've got 26 segments with tag
                        > errors, so when I get round to correcting them, I'll have to make sure
                        > that the one intentional instance is left untouched. But I'm unlikely to
                        > "correct" it by mistake; my tag errors rarely involve tag inversion, so
                        > when they do I notice it. My tag errors tend to be omitted tags,
                        > duplicate tags, or damaged tags (e.g. snipping off one of the brackets
                        > by mistake).
                        >
                        > Marc
                        >


                        Really, it's not a big problem, just a minor annoyance, and I wonder
                        what others did. And, like I said, a linguistic issue, not a software bug.
                        Translating the segment "out of order" and then just editing it in OOo
                        is simple enough, I suppose.

                        Thanks!
                        /tony

                        >
                        >
                        >


                        --
                        Anthony Baldwin
                        http://www.BaldwinLinguas.com - Translation & Interpreting

                        http://www.LinguasOS.org - Linux for Translators

                        Así también, la lengua es un miembro pequeño,
                        y se gloría de grandes cosas.
                        He aquí, ¡un pequeño fuego ­cuán grande bosque enciende!
                      • andrius
                        ... Yes. http://www.indesko.com/telechargements/eradicateur_de_surch/
                        Message 11 of 11 , Feb 3, 2008
                        View Source
                        • 0 Attachment
                          Št, 2008 02 02 14:05 +0100, Marc Prior rašė:
                          > andrius schrieb:
                          >
                          > > There is some OpenOffice plugin for this, called "slyle overload
                          > > cleaner" or something (French interface), but I haven't used it.
                          >
                          > This sounds very useful, Andrius - do you have a link for it?

                          Yes.

                          http://www.indesko.com/telechargements/eradicateur_de_surch/
                        Your message has been successfully submitted and would be delivered to recipients shortly.