Two or more capitals

[KariRudjord]

Can we implement this new OUP rule for sequences of captial letters (not acronyms)?

When two or more capitals appears in text - in headlines, the first three words in a chapter etc. - use 3 x dot 6 (6-6-6) before (Oh, scary!) and 1 x dot 6 (6) after. This shall include space so that there is only one (6-6-6) in front of the capital words and one (6) after the last word with capitals.

(Today there are 2 x dot 6 (6-6) before capitals, and the (6-6) is repeatet after every space.)

[josteinaj]

@bertfrees do you think this is feasible in liblouis or should we just do it in our block-translate.xsl?

[bertfrees]

Should be feasible yes. We can start with some Liblouis YAML tests in any case. Can one of you do that?

[josteinaj]

Great. We'll create some tests.

@KariRudjord: so this is only for headlines but not in normal paragraphs? Is the following correct?

Headlines

---

Headline: THIS IS A HEADLINE WITH SEVERAL WORDS

Braille: ⠠⠠⠠⠞⠓⠊⠎⠀⠊⠎⠀⠁⠀⠓⠑⠁⠙⠇⠊⠝⠑⠀⠺⠊⠞⠓⠀⠎⠑⠧⠑⠗⠁⠇⠀⠺⠕⠗⠙⠎⠠

---

Headline: This IS A HEADLINE with several words

Braille: ⠠⠞⠓⠊⠎⠀⠠⠠⠠⠊⠎⠀⠁⠀⠓⠑⠁⠙⠇⠊⠝⠑⠠⠀⠺⠊⠞⠓⠀⠎⠑⠧⠑⠗⠁⠇⠀⠺⠕⠗⠙⠎

---

Headline: This is a headline with several words

Braille: ⠠⠞⠓⠊⠎⠀⠊⠎⠀⠁⠀⠓⠑⠁⠙⠇⠊⠝⠑⠀⠺⠊⠞⠓⠀⠎⠑⠧⠑⠗⠁⠇⠀⠺⠕⠗⠙⠎

---

---

Paragraphs

---

Paragraph: THIS IS A PARAGRAPH WITH SEVERAL WORDS

Braille: ⠠⠠⠞⠓⠊⠎⠀⠠⠠⠊⠎⠀⠠⠁⠀⠠⠠⠏⠁⠗⠁⠛⠗⠁⠏⠓⠀⠠⠠⠺⠊⠞⠓⠀⠠⠠⠎⠑⠧⠑⠗⠁⠇⠀⠠⠠⠺⠕⠗⠙⠎

---

Paragraph: This IS A PARAGRAPH with several words

Braille: ⠠⠞⠓⠊⠎⠀⠠⠠⠊⠎⠀⠠⠁⠀⠠⠠⠏⠁⠗⠁⠛⠗⠁⠏⠓⠀⠺⠊⠞⠓⠀⠎⠑⠧⠑⠗⠁⠇⠀⠺⠕⠗⠙⠎

---

Paragraph: This is a paragraph with several words

Braille: ⠠⠞⠓⠊⠎⠀⠊⠎⠀⠁⠀⠏⠁⠗⠁⠛⠗⠁⠏⠓⠀⠺⠊⠞⠓⠀⠎⠑⠧⠑⠗⠁⠇⠀⠺⠕⠗⠙⠎

---

[KariRudjord]

It is in all kinds of text, also paragraphs. And it would save me a lot of meaningless work :-)

[josteinaj]

How about punctuation, numbers and other characters? Do they affect this? Can there be multiple sentences between ⠠⠠⠠ and ⠠?

[KariRudjord]

A whole paragraph or a group of paragraphs can appear in capital letters. Then the group of paragraphes are surronded by only one ... before and . after (the one after comes in addition to a ordinary punctuation). This should not be stopped by a number or other sign inside the text.

[josteinaj]

Ok. I'll update the tests in liblouis. Does these look right?

• Input: Jeg bare MÅ snakke med deg!
• Output: ⠠⠚⠑⠛ ⠃⠁⠗⠑ ⠠⠠⠍⠡ ⠎⠝⠁⠅⠅⠑ ⠍⠑⠙ ⠙⠑⠛⠖

• Input: DEN BESTE BURSDAGEN jeg har hatt var da jeg fylte åtte.
• Output: ⠠⠠⠠⠙⠑⠝ ⠃⠑⠎⠞⠑ ⠃⠥⠗⠎⠙⠁⠛⠑⠝⠠ ⠚⠑⠛ ⠓⠁⠗ ⠓⠁⠞⠞ ⠧⠁⠗ ⠙⠁ ⠚⠑⠛ ⠋⠽⠇⠞⠑ ⠡⠞⠞⠑⠄

• Input: EN ROSE MED TORNER
• Output: ⠠⠠⠠⠑⠝ ⠗⠕⠎⠑ ⠍⠑⠙ ⠞⠕⠗⠝⠑⠗⠠

• Input: EN ROSE! MED 15 TORNER.
• Output: ⠠⠠⠠⠑⠝⠀⠗⠕⠎⠑⠖⠀⠍⠑⠙⠀⠼⠁⠑⠀⠞⠕⠗⠝⠑⠗⠄

• Input: SAK 27/07 SØKNAD OM ØKONOMISK STØTTE FOR 2007
• Output: ⠠⠠⠠⠎⠁⠅ ⠼⠃⠛⠌⠼⠚⠛ ⠎⠪⠅⠝⠁⠙ ⠕⠍ ⠪⠅⠕⠝⠕⠍⠊⠎⠅ ⠎⠞⠪⠞⠞⠑ ⠋⠕⠗ ⠼⠃⠚⠚⠛⠠

[josteinaj]

@bertfrees: liblouis/liblouis#687

[KariRudjord]

EN ROSE! MED 15 TORNER.
It should be 2 x dot 6 at the end of the sentence. The first to close the capital letter thing, the other a punctuation.

If it was: En ROSE! MED 15 TORNER? it should be the same in the end, first dot 6 to close the capital letter thing, then a 26 (question mark)

[KariRudjord]

Didn't meen to close ...

[josteinaj]

Ok, so:

• Input: EN ROSE! MED 15 TORNER.
• Output: ⠠⠠⠠⠑⠝⠀⠗⠕⠎⠑⠖⠀⠍⠑⠙⠀⠼⠁⠑⠀⠞⠕⠗⠝⠑⠗⠠⠠

• Input: EN ROSE! MED 15 TORNER?
• Output: ⠠⠠⠠⠑⠝⠀⠗⠕⠎⠑⠖⠀⠍⠑⠙⠀⠼⠁⠑⠀⠞⠕⠗⠝⠑⠗⠠⠢

Right?

[josteinaj]

@KariRudjord also, how about this case:

SAK 27/07 SØKNAD OM ØKONOMISK STØTTE FOR 2007

Should the dot 6 be after "2007" or after "FOR"?

[josteinaj]

@KariRudjord also: has OUP documented this rule anywhere?

[KariRudjord]

ROSE = correct
SAK = dot after 2007 (in en of sentence)

[KariRudjord]

To eller flere ord med store bokstaver-pf.pdf

[josteinaj]

@bertfrees liblouis/braille-specs#7

[bertfrees]

EN ROSE! MED 15 TORNER.
It should be 2 x dot 6 at the end of the sentence. The first to close the capital letter thing, the other a punctuation.

If it was: En ROSE! MED 15 TORNER? it should be the same in the end, first dot 6 to close the capital letter thing, then a 26 (question mark)

Shouldn't it be dot 3 for the punctuation??

SAK = dot after 2007 (in en of sentence)

What do you mean with "in en(d) of sentence"?

I can't find more details about these cases in the PDF. I wonder if much thought has been put into this? I find it a bit illogical that the number (2007) at the end of the capital block is included (closing mark comes after it), but the punctuation (. or ?) is not included (closing mark comes before it) even though it is more "connected" to the capital block because there is no space between it and the last word in capitals (TORNER).

[josteinaj]

Fixed in liblouis repo on August 25th: liblouis/liblouis@fe5ce6f. Will presumably be included in the next release of liblouis, and when NLB has moved from NLB-PIP to DAISY-PIP, we'll be able to use this change.

[josteinaj]

Shouldn't it be dot 3 for the punctuation??

Yes you're probably right. Will add a fix to the PR branch here: liblouis/liblouis#687

I see here that @KariRudjord points out that the "end of capital letters string" marker should be after the year in the string "SAK 27/07 SØKNAD OM ØKONOMISK STØTTE FOR 2007". I don't know how important this and maybe we'll find more corner cases where this shouldn't apply if we try fixing it in liblouis? So I'd say that the current behavior with the marker after "FOR" is fine for now, and then we can do some testing to see how it works.