funderburkjim
commented
7 years ago This <UL/> element in pwg.xml comes from the pseudo-xml form <UL> in the pwg.txt digitization.
In other words, it was part of the digitization, so that's why it was included.
It was marked 'undocumented' since I didn't understand what it was. It was just waiting for someone (you!) to investigate.
Based on your investigation, I think I agree that <UL> actually serves no useful purpose and should be
removed.
I am not sure about removing the preceding line break. Will have to look at the pwg.txt instances first. Will post them here so we can all take a look. Then, will plan to generate corrections.
BTW: Nice presentation of the situation.
fxru
gasyoun
The element
ULis – according to pwg.dtd – undocumented.<UL/>occurs six times in pwg.xml all with the path/pwg/H1/body/ULand it is handed through as<UL>with a preceding line break into the interface:AkarRay: ...
</s><UL/><s>...KagaYja: ...
</ls> <UL/></body><tail>...Gaw: ...
<divm type="n" n="1">1)</divm> <UL/><divm type="e" n="a">...Jara: ...
</s> hinzu; vgl. <UL/> <ls>...wakkara: ...
Musik verstan</i>- <UL/> -<i>den und am...baka: ...
von Besonnenheit, aber <UL/>auch von Schelmerei...The occurrences are all at a column break at the end the first column, but only when it coincides with the bottom line of the page but does not jump up to the top of the page again, because of a preceding end of another letter-part. (see UL_pwg_baka.jpg below) I didn’t verify whether it occurs in all of these situations. There is no further pattern in realtion to the other markup:
in
AkarRay, it basically cuts an<s>entity in half.in
KagaYja, it is completely oddly placed, as the entry doesn’t actually continue and<UL/>occurs directly before the closing<body>tagin
Gaw, it is positioned between two<dvim>notesin
Jarait is placed sowhere inside the running text of<body>in
wakkara, it cuts an<i>-element – and in fact a word (verstanden ‘understood’) in it – into two partsin
baka, it is again located somewhere in the running text of<body><UL/>thus simply indicates the end of the first column, and a continuation of the text somewhere below the top of the page. While it doesn’t relate to anything in the<body>element in particular, it can effectively separate what would be one element into two, as is the case with the<i>elements inwakkara. The current treatment in the interface is probably not optimal and removing it (and the preceding line break) would be a simple fix that would improves the situation.(relevant XPATH to the entries:
/pwg/H1/h/key1[ancestor::H1/body/UL])