USFM markup interferes with Biblical Terms Tool (and thus Glossary automatization)

Question

I’m using PT 9.0.100.87.

Our project marks all occurrences of proper names (\pn), geographical names (\png), and names of Deity (\nd). This facilitates publishing in ways peculiar to our project.

Sometimes we have multi-word Biblical Terms that include a name. Most importantly, our keyterm for “kingdom of heaven” is as follows:

គ្លើយ \nd អ៊ើយឋៃ\nd* វ៉ៃ-ឡាំងហ៍
= matter \nd god\nd* rule

I have tried the following combinations in the Biblical Terms tool, and none will find/match the expression in the text:

• គ្លើយ \nd អ៊ើយឋៃ\nd* វ៉ៃ-ឡាំងហ៍ <- literal
• គ្លើយ ** វ៉ៃ-ឡាំងហ៍ <- replace \nd អ៊ើយឋៃ\nd* with ** [see image below]
• គ្លើយ ** អ៊ើយឋៃ* វ៉ៃ-ឡាំងហ៍
• គ្លើយ ** អ៊ើយឋៃ* វ៉ៃ-ឡាំងហ៍
• គ្លើយ ** អ៊ើយឋៃ** វ៉ៃ-ឡាំងហ៍
• គ្លើយ ** អ៊ើយឋៃ** វ៉ៃ-ឡាំងហ៍

image1086×242 56.1 KB

The problem is clearly coming from the final asterisk in the marker \nd*. We can see this with the Key Term “Son of David”, where the final marker \pn* comes at the end of the term instead of the middle (making it different from “Kingdom of God”):

ណាគ ច ពតាវ \pn ដាវីត\pn*
= child grandchild king \pn david\pn*

If I place the string ណាគ ច ពតាវ \pn ដាវីត\pn* in the Biblical Terms tool, it finds the term, but note that it does not interpret the final * in \pn* as being a part of the Key Term.

image974×405 137 KB

So apparently, the tool can’t read (or skip over, or move beyond) an asterisk.

This has consequences for linking occurrences of the term to the glossary entry. (When I try to make the link for “Son of David”, nothing gets linked, even though the Biblical Terms tool is identifying the Key Term as present.)

Does anyone know of a way to use valid USFM markup within Key Terms, have the Key Terms tool recognize the term, and have the Glossary linking tool markup those Terms? Thanks! (It would be sufficient to have a trick to get the tool to skip over asterisks!)

较早的帖子 - 以原始语言显示

Answer 1

This is a known issue because, in the rendering, the asterisk is considered to be “match any word-forming characters” instead of a literal asterisk. Since the asterisk is not normally a word-forming character, the asterisk as part of the end-marker does not match.

However, there is a hack to get around it. Although I tried it out on a test project and it didn’t seem to cause any issues with the Inventories, Checks, Wordlist, etc., I still feel a disclaimer is needed:

WARNING: This is a hack and can cause unexpected behavior in other parts of Paratext. Use at your own risk.

Add an asterisk (*) to your word-medial punctuation characters in your project’s language settings. This works because when the wildcard (asterisk) is encountered, it will now match the asterisk in the end-marker. Because the checks, inventories, Wordlist, etc. ignores most markers, it doesn’t seem to create any problems in my brief testing.

较早的帖子 - 以原始语言显示

Comments

I’ll try this and then report whether we notice any undesirable consequences. Thanks very much, anon291708!

较早的帖子 - 以原始语言显示

Answer 2

No problems so far - a few hours in. (Except in PT 8, I get an error when I open the software. I can close the error and continue into the program. No error notification in PT 9.)

较早的帖子 - 以原始语言显示