What does the parseable filter do?

Question

Paratext 9 has the option of filtering for parseable in the Wordlist. It is unclear what this is doing, and it does not do what I need it to do. I could not find any information in the Help topics.
Because the language I work with has a very complex morphology, the translators will often mark a spelling as correct when it is actually incorrect. In order to check those, I look at the morphology, break the words into morphemes and mark that breakdown as correct. What I need is a filter that will filter out the breakdowns that are marked as correct. There does not seem to be a way to do that. I had hoped that the parseable filter would help, but it does not. it would be helpful if the column could be sorted according to whether a breakdown was marked as correct or not.

舊文章 - 以原文顯示

Answer 1

From the name “exclude parseable” I would expect it to exclude words with a parse.

This happens when I test it. I set up a test project with this phrase, “In the days of King Herod, a priest named Zechariah belonged to the order of Abijah. His wife, whose name was Elizabeth, also did belong to Aaron’s descendants”, then launched the wordlist with morphology showing, when I parse “belonged” as “belong +ed”, Paratext enters the whole word parse /belong/ for belong, but shows it in red. Then if I parse “name” as “nam”, it guesses the parse “nam +ed” for named.

Now if I save and turn on “Exclude Parseable”, all four of these words; belong, belonged, name, named" disappear from the wordlist.

But in another project I have a copy of, with a complex morphology, when I turn on “Exclude Parseable” it does seem to exclude all words where the parse is approved, and also some words where the parse is not approved, but not all of them.

舊文章 - 以原文顯示

Comments

We do not have an association to a Fieldworks project. We do use the Interlinearizer, but I cannot see that this would have any influence.
I would agree with sewhite. When I select Exclude Parseable, the words with an approved parsing are excluded. However, many words with a wrong and unapproved parsing are also excluded. It would be helpful if it would only exclude the words with an approved parsing. The program usually suggests some parsing, and in some cases that parsing is correct, but quite often incorrect. All NT books and a few OT books have now been checked and all morpheme analyses in those books are marked as approved.
The Wordlist window has a Refresh List option which will refresh the list if a change has been made in the spelling column, but not if a change has been made in the morphology column. In order to refresh the morphology, I can uncheck the Exclude Parseable and re-check it, but that does not help. In order to refresh the morphology, I have to close the Wordlist Window and re-open it.

舊文章 - 以原文顯示

Answer 2

Good question. I don’t know. Could it be related to having a Fieldworks project associated with your Paratext project, that has approved parses for the word or similar words?

舊文章 - 以原文顯示

Answer 3

That menu item is supposed to be hidden (you should only see it while holding Shift). It’s actually a bug that it’s showing.

It’s experimental code that Nathan Miles wrote a long time ago. From a comment in the code:

I am testing the theory that it would be useful to see all the words which do not have a “strict” parse.
Strict means

• The existing sequence of prefix strings has been seen in an approved parse.
  e.g. in English we are likely to have approved “+ing +s” but not “+s +ing”.
  This is not very useful for English, but for a language with long strings of small suffixes
  it may be a key distinction.
• The stem has been seen in an approved parse.

I’ll make sure it’s hidden properly again.

舊文章 - 以原文顯示

Comments

Thank you for the explanation. It would actually be very helpful to have such a filter if only it would exclude all the words with an approved parsing. So, let me suggest this as a wish.

舊文章 - 以原文顯示