This flag, if set to false, will re-use data in indexedDB if present. If not present, it will fetch as normal.

They are here: https://git.door43.org/unfoldingWord/en_translation-annotations.

for example, Titus questions are:
https://git.door43.org/unfoldingWord/en_translation-annotations/src/branch/master/TIT/TIT_tq.tsv

This is for aligned texts. See ULT (aligned) word count accuracy for the other cases.

Learned yesterday that aligned texts have no limit on how deeply the children attributes are nested in the USFM-JS format.

At present a depth of 5 is hard-coded, and may not be sufficient. This would cause the words to be missed and the book to be under counted.

Convert both aligned ULT and UST to use a recursive algorithm instead.

This is for unaligned texts. See UST (aligned) word count accuracy for the other case.

The tQ component only has word counts. There are no details or links to the individual chapters (folders in Github) or verses (markdown files).

Should there be?

Similar to allow a range of chapters can be provided, it would be excellent if we could provide a range of books.

For example, if I want to see what is included in Titus, Luke, and Acts, it would be great to be able to specify:

<BookPackageRollup bookId='tit,luk,act' chapter='' />

and then get the combined numbers for all of those.

Add them...

... using try/catch

This issue is for aligned texts. See ULT (unaligned) word count accuracy for others.

The header for TA says "Translation Articles" but it should be "Translation Academy"

Do not include XML comments in word counts. Example from Strong's article H0835.md (use raw mode to view):

<!-- Status: S2="NeedsEdits" -->
<!-- Lexica used for edits:   -->

Unnecessary with localforage and makes it so you can't view properly and expand using the dev tools.

Select a book and manually count and compare counts. Perhaps Jude would be a nice test case that would be easy to manually count?

For Psalm 1, the number of notes 565 and the word count is only 28.

Translation Notes for "PSA" and Chapters 1
Total number of notes: 565
Distinct number of words in notes: 23
Total number of words in the notes: 28

This is for aligned texts. See UST (unaligned) word count accuracy for the other case.

Currently the TA information is embedded in the TN card but it needs to be represented distinctly as it's own component.

When this RCL is used by the Book Package App, the header title looks like this:
Package Rollup for "1JN, 2JN, 3JN" and Chapters (ALL).

While the RCL can restrict the results to specified chapters, this is not possible in the app.

So change the RCL to omit the chapters bit when no restriction is specified. Thus the title for the app will look like this: Package Rollup for "1JN, 2JN, 3JN"

Due to limitations of localStorage, use NPM component local forage instead.

Here is a short list of cosmetic changes to make on the "cards":

• Refer to word counts as "Total word count:"
• Refer to unique words as "Unique words:"
• Place "Total word count" first

The Gitea API has this:
https://git.door43.org/api/swagger#/repository/repoGetContents

For given a book of the bible, say "tit", this URL returns all the chapter folders:
https://git.door43.org/api/v1/repos/unfoldingword/en_tq/contents/tit

Then for each chapter folder, "03", this URL returns all the question verse files:
https://git.door43.org/api/v1/repos/unfoldingword/en_tq/contents/tit/03

Then for each verse file found, here "01.md", this URL will return metadata and actual content of file (base64 encoded):
https://git.door43.org/api/v1/repos/unfoldingword/en_tq/contents/tit/03/01.md

This will avoid "guessing" and eliminate the 404 errors.

Similar to unfoldingWord-box3/hello-world-react-component-library#2, setup cypress to run the tests in this repo.

Add a grand total of all the words across all the resources.

There are a number inconsistencies due learning things during development on box3. Examples:

• Names of the components
• How data is passed around (sometimes in database, sometimes in returned objects, sometimes both)
• Probably some other things I haven't noticed yet...

For example, if a team chooses to do Titus first and then Luke after that, how many articles from Translation Words and how many modules from Translation Academy would be a part of the Luke package, given that they've already translated a handful in the Titus package?

We need one more top level component that that lists out each book of the Bible. This would probably be the meta element that would be used as a visual method of selecting a book.

Each book name should be clickable, revealing underneath it the Book Package contents.

The new books JSON has two new attributes:

• Title: the name of the book, such as "Titus"
• USFM: the base name of the book in USFM format, such as "57-TIT"

For book oriented data, we only need to retain total word counts for entire book. For example, for Titus, we need to retain the counts for ULT, UST, UTQ, and UTN.

For reference oriented data (to wit, UTA and UTW), we need to preserve:

• the article type, i.e., UTA or UTW
• the article name or URI
• the counts for the article
• the books which reference the article

All this is known to the RCL currently, but is not retained. Retention will enable optimization to be done.

@jag3773 Other thoughts?

As an example, consider translation questions. In advance I do not know whether any questions exists for a given verse. So I have guess and attempt to fetch the questions. When they don't exist, I get one or two errors for each such failed attempt. I suspect it might be faster to request all the files in a repo and use it as a whitelist to avoid failed attempts.

Failed to load resource: the server responded with a status of 404 (Not Found)
:6060/#/BookPackage?id=bookpackagetn:1 

Access to XMLHttpRequest at 'https://git.door43.org/unfoldingWord/en_uhal/raw/branch/master/content/H0834a.md' from origin 'http://localhost:6060' has been blocked by CORS policy: No 'Access-Control-Allow-Origin' header is present on the requested resource.

Switch to NPM Material Table component.

Question: do we need the word order table that shows all words in the lexical order found? If not just show the word frequency table.

Another option is to put both tables into their own collapsible section.

BP Information

Re-flow the BP information using the following as an example, using 1 and 2 John. I'm thinking of something like Tree View https://material-ui.com/components/tree-view/. Note the order of the resources below too.

> Book Packages Total Word Count: 72,932

When expanded, it would reveal these indented:

⌄ Book Packages Subtotals (<- expanded by default)

• ULT Word Count: 3,245
• UST Word Count: 4,821
• UTA Word Count: 10,031
• UTW Word Count: 2,426
• UTN Word Count: 33,783
• UTQ Word Count: 19,088

> 1 John Book Package Word Count: 61,135

> 2 John Book Package Word Count: 20,668

When those are expanded, they would reveal each component (indented), which would also be a dropdown. I've expanded them in the example below so I don't have to repeat them twice.

⌄ ULT Word Count: 319

[Word Frequency Table]

⌄ UST Word Count: 498

[Word Frequency Table]

⌄ Translation Academy Word Count: 7,373

Linked modules: 7 unique, 14 total links
[Table of links]

⌄ Translation Words Word Count: 11,254

Linked articles: 37 unique, 69 total links
[Table of links]

⌄ Translation Notes Word Count: 936

Number of Notes: 33
[Word Frequency Table]

⌄ Translation Questions Word Count: 11,254

Number of Questions: 11
[Word Frequency Table]

Currently, filenames in the repos may not be named per their embedded references in other files. For example, a UTA article may begin with a capital letter when it should be lowercase.

These mistakes result in lower word counts and should be captured so they can be corrected.

See go-gitea/gitea#9482

This would simplify the code by avoiding recursion in the code to traverse the directory structures in a repo.

At present this package has an older version of word count code. Replace with NPM component uw-word-count