Whereas I have not previously distinguished the 's of the translations between those translating the whole sentence or its subcomponents, this would be beneficial to do in order to facilitate transformation scenarios which extract items from a corpus and then create dictionary entries. Thus it should be as follows:
<spanGrp type="translation">
<span target="#d1e140" xml:lang="en" type="sentence">Shanty came with me to the S.F airport.</span>
<span target="#d1e140" xml:lang="es" type="sentence">Shanty vino conmigo al aeropuerto de S.F.</span>
<span target="#d1e141" xml:lang="en">came</span>
<span target="#d1e141" xml:lang="es">vino</span>
<span target="#d1e147" xml:lang="en">Shanty</span>
<span target="#d1e147" xml:lang="es">Shanty</span>
.......
</spanGrp>
Do this via XSLT:
for-each //spanGrp
where previous sibling is //seg[@type='S']
find select //(span[@xml:lang='en'])[1] and //(span[@xml:lang='es'])[1]
Whereas I have not previously distinguished the 's of the translations between those translating the whole sentence or its subcomponents, this would be beneficial to do in order to facilitate transformation scenarios which extract items from a corpus and then create dictionary entries. Thus it should be as follows:
Do this via XSLT: