fs and notes

[HelenaSabel]

Hello! Let's see if we like how the notes are represented as feature structures at this point. If we have a "regular" note (that is, with no translated content in it) it looks like:

<fs corresp="#M2_p2_c90">
               <f name="southey" n="14" ana="-0.07">
                  <string ana="start">He was called the </string>
                  <string ana="note">Footnote: We have no word for the Donzel
                     of the Spaniards, and Damoisel of the French.</string>
                  <string ana="end">
                  Child of the Sea, for so they had named
                     him,</string>
               </f>
               <f name="montalvo" n="13">
                  <string>el cual el Donzel del Mar se llamava que assí le pusieron nombre:</string>
               </f>
               <f name="type" select="indefinite">
                  <string>Comments</string>
               </f>
            </fs>

If we have a note that contains direct references to Montalvo, we embed a <f name="note"> with a <fs> for each match.

<fs corresp="#M1_p1_c43">
               <f name="southey" n="23" ana="0.61">
                  <string ana="start">in whom as he beheld her by the light of the three torches, he thought all the beauty of the world was
                  </string>
                  <fs>
                     <f name="note">
                        <fs>
                           <f name="southey" n="8">
                              <string>Here, says the Spanish author, we should learn,</string>
                           </f>
                           <f name="type" select="add">
                              <string>Comments</string>
                           </f>
                        </fs>
                        <fs corresp="#M1_p1_c52">
                           <f name="southey" n="10" ana="0.2">
                              <string>that when women turn aside their thoughts from worldly things,</string>
                           </f>
                           <f name="montalvo" n="12">
                              <string>que assí como las mugeres apartando sus pensamientos de las mundales cosas:</string>
                           </f>
                           <f name="type" select="indefinite">
                              <string>Comments</string>
                           </f>
                        </fs>
                        <fs corresp="#M1_p1_c53">
                           <f name="southey" n="10" ana="0">
                              <string>despising the great beauty with which Nature has endowed them,</string>
                           </f>
                           <f name="montalvo" n="10">
                              <string>despreciando la gran hermosura de que la natura las dotó:</string>
                           </f>
                           <f name="type" select="indefinite">
                              <string>Comments</string>
                           </f>
                        </fs>
                       <!--  Many other <fs> as the previous ones -->
                     </f>
                  </fs>
                  <string ana="end"> centered.</string>
               </f>
               <f name="montalvo" n="37">
                  <string>que a la lumbre de tres hachas que en la cámara Ardían la mirava paresciéndole que toda la hermosura del mundo en ella era junta teniéndose por muy bienaventurado en que Dios a tal estado lo truxera:</string>
               </f>
               <f name="type" select="indefinite">
                  <string>Comments</string>
               </f>
            </fs>

Do you think this is coherent?

[ebeshero]

@HelenaSabel @setriplette Just a quick observation: It's easier to read this if you click on the pencil icon (as if you were going to edit Helena's comment), so you can see the nesting a little more easily. It's complicated, but I think it's readable, and we can pull the notes out of this code to study them all together when we're ready to do that. We could read it most easily with the outline view in oXygen open, so we can readily see where Southey's notes are positioned.

[ebeshero]

@HelenaSabel @setriplette I think this works, and I hope we can use it.

[ebeshero]

@HelenaSabel @setriplette Greetings Helistaceans! Okay--here's a long message--so you may want to grab some coffee and read this at a leisurely pace. :-) I'm reopening this issue, because we need to repair the XSLT we're using to generate the XML files we're using to produce our HTML tables. We've all been busy with other projects, so it took us a while to figure out what we really need here, and what might cause problems. Now, I think, we can see things a little more clearly. And if we can take a look at this in the next two weeks, I think it'll help with the presentation we're planning to deliver (or at least my part of it)! I'm fine with taking the lead on this if you're busy, Helena, but if you can, it'll be really helpful for you to revisit your XSLT for generating the fs.xml files, and I'll help test and comment on my end! That way I'm not mucking about with your carefully-crafted XSLT scripts.

@HelenaSabel here's what going on: Stacey and I are virtually co-presenting here in a couple of weeks. This is a two-day methods in translation symposium in Hong Kong, and we're the only TEI/XML-based project represented. We're both digging back into Amadis after some time "away", though Stacey has been working on Amadis sort of all along: She's had a student assistant contribute some transcriptions of Montalvo (most of which don't have counterparts ready in Southey yet b/c I'm backlogged). Those transcriptions aren't really encoded and Stacey says they need lots of corrections, which she'll work on eventually. It's also a moment after Stacey has finished and submitted her book manuscript (recently accepted for publication--congrats, Stacey!), in which she devoted some space to writing about Southey's translation of Montalvo. Stacey and I have been meeting to talk about the project a couple of times this past week, and she has some new clarity on the project and on the classification of translations. As we reviewed the fs.xml files we're using for that work, we had much clearer ideas of what those should be doing--and we definitely want to change their structure to address some problems--including (especially) how we're handling Southey's notes.

First, as a quick review for everyone, the fs files should give us a means to classify the units of translation we're processing. So, we pull out in fs format a structure that lets us pair units from Montalvo with corresponding units from Southey. Where Southey makes additions, those appear alone, and are classified as additions. Where there are corresponding units, we add f elements with attributes carrying our classifications. That's all good--except that we run into problems when we try to process Southey's notes. If we follow the structure of sentences with annotations literally, the annotations are usually positioned in the middle of Southey's sentences, and this caused us to try to nest them in our fs files in between the units of a phrase in the main text. Our strategy here outputs basically invalid TEI, which doesn't need to to bother us, except it's also kind of difficult to follow and, we decided today after reviewing a few of these together and looking at the source texts, really adding an unnecessary complication.

Consider:

• It'll help us to keep this as simple as possible, and not to consider the fs output literally as an edition of the two texts, and that means it's fine if a small amount of disordering takes place in order to study the translation units.

• What Helena is outputting now from those fs XML files are our HTML tables. These are terrific: they are quickly legible on the web to the outside world and easy to share our process with others. See our published HTML table for Southey's Ch. 2: Where Southey has performed translation in a note, we show the annotation level of text as italics. That's good. And the color coding of the HTML output is perfect. That shouldn't change.

• Stacey and I agreed that we might want a small symbol or something to appear on a translation unit that Southey has transposed in Montalvo, to quickly signal any clause-unit that he has moved out of its sequence in Montalvo. But we aren't concerned about changes to the order of words within the clause-unit, and Stacey says, these really do not require our markup at all (since an English sentence is just about always going to change the subject-verb-object order of a given sentence in Spanish anyway and is to be expected). Really, what matters more here is signaling transposition. (Basically, if the @corresp attribute points to something that is NOT + 1 from the previous @corresp, I think we can signal this, though we'll want to observe our results carefully.)

• What needs to change in the fs.xml is the way we are holding Southey's notes: We don't need to be so literal about where these appear in relation to Southey's main text. ( I realize I'm the one who probably imposed this b/c I wanted to preserve textual hierarchies somehow back in 2015. We're thinking differently about this now! The tables aren't an edition, but they're just a way of studying translation units.) Let's simplify this now, and make good TEI, too. To simplify it, for the purposes of preparing the fs units, let's pretend as if the notes appear at the ends of Southey's sentences, and output those after the complete unit of main text is processed. (I realize this may be complicated to rearrange in the XSLT, and am ready to help!)

• There are other changes we need to make, too, to bundle multiple "sets" together in the fs files. I repaired a couple of these files by hand in this commit, and you can see the changes I've made to the code there. Basically, we need to be using <vColl> to combine multiples of <string> or <fs> in an fs hierarchy, and we'll need to add that to the XSLT outputting these files.

Note: I applied the TEI-all schema to fs_S2.xml while in the throes of repairing it to find out if the errors I was seeing were due to my ODD-generated schema. They weren't (the errors are the same whether we're using TEI-all, or our schema). I also figure, we'll re-output these tables once we've updated the XSLT anyway. We do want to use our ODD-generated schema.

• Finally, Stacey and reviewed the schema Helena wrote for the fs files. That schema is now fully incorporated into our ODD, but we left comments there just as we were thinking about our classification system. Stacey observes that we should simplify this a great deal now--and to read more about that, let's take a look at the comments we left here, marked 2017-05-27.

Okay--that's a lot! Helena, I'll be happy to Skype with you about all this if it'll help, though we'll need to find a time. I'm traveling a lot this week and next, and working on two other projects at the same time (ahh, multitasking!), so that may be hard. Anyway, let's discuss further at least here! Thanks both of you!

[HelenaSabel]

Hello! I updated the XSLT transformations in order to create the vColl elements when needed. Inside the folder 'tables' there is a new folder'correction' with the new FS files. I didn't delete the previous ones because we had begun the analysis of the translations and we might want to look back at them. Regarding the notes, I kept things as they were (I think). In new Chapter1.html (folder html/correction) there is an inline note (that it does not contain translated content): the cell that contains the note is underlined and the note contents are readable with mouseover. When the note contains translated materials, the contents of the note appear as table rows in italic. Do you agree with this management of the notes?

I tried associating the ODD generated schema to the new FS files and I didn't get any errors, but I should, because right now we haven't included any translation types in those files... May you check that, @ebeshero ? I read your comments about the simplification of categories and it makes a lot of sense. Is it worth it to include the old types into the new files or is it better to wait for the new classification?

[ebeshero]

@HelenaSabel Thanks for the speedy XSLT work--that's impressive! :-) Validation took a few minutes to complete for me on the files, but I tested fs_S23.xml and eventually saw many errors of the kind we expect to see re valid values for type of translation, etc. fs_S1.xml turned out just fine, valid according to our schema. (That's what we'd expect, I think, since that one was completed?)

I am not sure why the validation is so slow! These are just testing attribute values prescribed in Relax NG.

[ebeshero]

@HelenaSabel As to your second question: whether to include the old types/subytpes in the new files, or wait for new classifications, I'm not sure. I'm inclined to wait, but let's ask @setriplette !

[HelenaSabel]

Indeed you are right @ebeshero : I think I failed to notice the “Validation in progress” message and just saw the green square that was kept from the tei.allschema. It took a while too for the validation process to be over.

[setriplette]

I think let's wait for new classifications.

[ebeshero]

@HelenaSabel I'm viewing the HTML tables now, and yes, I think this output is good--it's consistent with what we've been doing, and yes, outputting the notes (when they are translation units) as their own table entry with text in italics works. And where notes don't contain translation (but are sort of at a "meta" level discussing the translation) I think it's appropriate to reveal those on mouseover. Thank you!

I'll paste preview links here to help Stacey quickly see the output tables: Ch, 1 Ch. 2 Ch. 21 Ch. 23

[ebeshero]

Good morning @setriplette ! :-)