Outlines may have malformed descriptions

[xristy]

In several outlines the is represented in the content of the element instead of the @type, for example in O4JW333:

<outline:description type="location">vol:1,ff.1r-311r(pp.1-621)</outline:description>
<outline:description type="authorship">
t. Sarvajñādeva (k); Vidyākaraprabha; Dharmākara (ka0; Dpal gyi lhun po; r. Vidyākaraprabha; Dpal brtsegs
</outline:description>
<outline:description>sde dge number 1</outline:description>
<outline:description>stog number 1</outline:description>
<outline:description>snar thang number 1</outline:description>
<outline:description>lha sa number 1</outline:description>
<outline:description>urga number 1</outline:description>
<outline:description type="toh">1</outline:description>
<outline:description type="sde dge number">1</outline:description>

this also illustrates duplication of the sde dge number.

[eroux]

Only 3 outlines seem to suffer from this problem, here are the fixed versions of these: fixed-outlines.zip can you quickly check them before I (or you) upload them to exist?

I figured it was easier to fix them directly than having tedious code to handle all the various cases like the spelling mistakes and cases where the description has a type like

<outline:description type="location">lha sa number 1</outline:description>

Note that I did not fix puzzling things like

<outline:description>snar thang number</outline:description>

(with no number), that's garbage data, but well, at least the interesting data will get transferred correctly...

[xristy]

Excellent! I'll upload these three. Thanks so much.

[xristy]

Corrections uploaded

[xristy]

O5JW1143 also has malformed descriptions - it was a clone of O4JW333

[eroux]

O5JW1143.xml.zip here's the fixed version

[eroux]

A few others I just spotted:

• O2DB75712 (ex: O2DB757121KG72517)
• O00CR0008 (ex: O00CR00084CZ281547)
• O1PD181215 (ex: O1PD1812154CZ133521)

[eroux]

The fixed outlines, can you upload them?

fixed-outlines.zip

[xristy]

I've uploaded these.

I noticed that in a number of cases there were sde dge elements but no numbers which I assume will simply be filtered out via xml2ld

I also noted some section names in O1PD181215 that have some sort of invalid character next to a chinese character

[eroux]

Thanks for the upload! For the ignore:

<description type="sde dge number"></description>

is ignored, but

<description>sde dge number</description>

is not (hence part of the changes I made to these outlines)... I didn't really look for invalid characters, do you have a node ID in which it appears?

[xristy]

O1PD1812154CZ135987, O1PD1812154CZ136047, O1PD1812154CZ136352, etc The section names directly under the skabs gsum pa/ rgyu mtshan nyid theg pa'i skor/ (ka-go) section

[eroux]

indeed, it's also an u+fffd as in the other encoding problems...