French SI Brochure feedback from BIPM (pages 61-80) about bipm-si-brochure HOT 13 CLOSED

from bipm-si-brochure.

61 | side note: "IAU Trans." in italics | adoc

63 | Resolution 3: no text in bold | adoc

63 | Resolution 3: side notes (1)and (2) no line break between value and unit. | adoc

67 | 2nd line: isolated » at beginning of line | adoc

72 | bullet point (a): l in italics | adoc

73 | 1st side note: Metrologia in italics | adoc

73 | 2nd side note: J. Radiol. Prot. In italics | adoc

73 | last bullet point: isolated » at beginning of line | adoc

74 | 3rd line below equation: "decide" in bold | adoc

74 | CIPM 2003: Metrologia in italics | adoc

75 | CIPM 2005: list needs to be split in two, following "considérand" and "décide" | adoc

77 | 3rd bullet point: Hg not in italics | adoc

77 | 4rd bullet point: Yb not in italics | adoc

80 | CIPM 2009: isolated » at beginning of line | adoc

All fixed.

63 | Resolutio 10: change "remarque" to a side note | adoc

I have moved the note to a footnote inside the table:

[cols="<,<,<",options="unnumbered"]
|===
| Facteur par lequel l’unité est multipliée | Préfixe | Symbole
| stem:[10^(15)] | peta footnote:[En français, il est d’usage courant de mettre un accent à péta.] | stem:["P"]
| stem:[10^(18)] | exa | stem:["E"]
|===

70 | change "remarque" following table into side note | adoc

Same as above.
Relocate note block position to a footnote inside the table. Fixed.

75 | CIPM 2005: group numbes following the decimal separator in groups of three | adoc

76 | 1st bullet point : group numbes following the decimal separator in groups of three | adoc

78 | Resolutio 10:group numbes following the decimal separator in groups of three | adoc

Suposed to be handled automatically.

71 | Resolution 12, second bullet point: μ not in italics | adoc

73 | list following "recommade": μ not in italics | adoc

74 | CIPM 2003: 2nd bullet point from end: μ not in italics | adoc

76 | REcommandation3: μ not in italics | adoc

77 | Recommandation 1: μ not in italics | adoc

Not fixable by markup.

61 | last side note: wrong reference, should be Page 61 (the same page) | adoc

@Intelligent2013, I believe this is a problem of cross-reference definition.

...
. de donner une définition du Temps atomique international;
footnote:[La définition du TAI a été donnée par le CCDS en 1970 (maintenant renommé CCTF), _voir_ <<ccds-tai-definition>>]
...

I've redifined <<ccds-tai-definition>> as <<ccds-tai-definition_s2,Recommandation S 2>> which points directly to ==== Recommandation S 2 instead of === Définition du TAI (PV, 38, 110-111 et _Metrologia_, 1971, 7, 43). Would that be enough to display the correct page number?

77 | 2nd bullet point following "recommande": change to 5s 2S1/2– 4d 2D5/2 | adoc

77 | 3rd bullet point: change to 5d10 6s 2S1/2(F = 0) – 5d9 6s2 2D5/2(F = 2) | adoc

77 | 4th bullet point: change to 6s 2S1/2(F = 0) – 5d 2D3/2(F = 2) | adoc

77 | 5th bullet point: change to 5s2 1S0 – 5s 5p 3P0 | adoc

80 | 1st bullet point following "recommande": change to 5s2 1S0 – 5s 5p 3P0 | adoc

80 | 2nd bullet point: change to 5s2 1S0 – 5s 5p 3P0 | adoc

80 | 3rd bullet point: 4s 2S1/2– 3d 2D5/2 | adoc

80 | 4th bullet point: 2S1/2 (F = 0) – 2F7/2(F = 3, mF = 0) | adoc

80 | 5th bullet point: 6s2 1S0 (F = 1/2) – 6s 6p 3P0 (F = 1/2) | adoc

@opoudjis, maybe this has to do with your concern of using Chemistry symbols in math expressions. These equations are rendering fine in HTML but not in PDF. For example in the first case,

the markup used is:

stem:[5"s" " "^2"S"_(1//2) – 4"d" " "^2"D"_(5//2)]

which renders in HTML like:

and, in PDF:

Now, a workaround (that @ronaldtse doesn't like) is to profane the semantics and write 5s ^2^S~1/2~ - 4d ^2^D~5/2~, or, I think I could make a mixture between inline stems and normal text, but it would look really messy.

from bipm-si-brochure.

This is doing my head in. :-)

If the same input looks ok in HTML and not PDF, that indicates that jEuclid is having trouble with the existing MathML. So I will need input from @Intelligent2013. But although all our lives would be easier avoiding AsciiMath/MathML for these expressions, that is not the direction we are going. So waiting for feedback from Alex on this.

from bipm-si-brochure.

Regarding the first case:

77 | 2nd bullet point following "recommande": change to 5s 2S1/2– 4d 2D5/2 | adoc

In resulted PDF I see:

Specified location:

Source xml:

<math xmlns="http://www.w3.org/1998/Math/MathML">
	<mn>5</mn>
	<mtext>s</mtext>
	<msup>
		<mrow>
			<mtext> </mtext>
		</mrow>
		<mrow>
			<mn>2</mn>
		</mrow>
	</msup>
	<msub>
		<mrow>
			<mi>S</mi>
		</mrow>
		<mrow>
			<mrow>
				<mn>1</mn>
				<mtext>/</mtext>
				<mn>2</mn>
			</mrow>
		</mrow>
	</msub>
	<mi>–</mi>
	<mn>4</mn>
	<mtext>d</mtext>
	<msup>
		<mrow>
			<mtext/>
		</mrow>
		<mrow>
			<mn>2</mn>
		</mrow>
	</msup>
	<msub>
		<mrow>
			<mtext>D</mtext>
		</mrow>
		<mrow>
			<mrow>
				<mn>5</mn>
				<mtext>/</mtext>
				<mn>2</mn>
			</mrow>
		</mrow>
	</msub>
</math>

As I undestood, jEuclid can't display superscripted 2 if previous char is non-break space or empty:

	<msup>
		<mrow>
			<mtext> </mtext>
		</mrow>
		<mrow>
			<mn>2</mn>
		</mrow>
	<msup>
		<mrow>
			<mtext> </mtext>
		</mrow>
		<mrow>
			<mn>2</mn>
		</mrow>
	</msup></msup>

Can we move preceding <mtext>s</mtext> and <mtext>d</mtext> inside msup/mrow via stem markup?
I.e. change stem for resulted XML:

<math xmlns="http://www.w3.org/1998/Math/MathML">
	<mn>5</mn>	
	<msup>
		<mrow>
			<mtext>s </mtext>
		</mrow>
		<mrow>
			<mn>2</mn>
		</mrow>
	</msup>
	<msub>
		<mrow>
			<mi>S</mi>
		</mrow>
		<mrow>
			<mrow>
				<mn>1</mn>
				<mtext>/</mtext>
				<mn>2</mn>
			</mrow>
		</mrow>
	</msub>
	<mi>–</mi>
	<mn>4</mn>	
	<msup>
		<mrow>
			<mtext>d</mtext>
		</mrow>
		<mrow>
			<mn>2</mn>
		</mrow>
	</msup>
	<msub>
		<mrow>
			<mtext>D</mtext>
		</mrow>
		<mrow>
			<mrow>
				<mn>5</mn>
				<mtext>/</mtext>
				<mn>2</mn>
			</mrow>
		</mrow>
	</msub>
</math>

I so, then in resulted PDF will be:

Second proposal - use MathJax instead of jEuclid in mn2pdf. But I need to check how it's possible at all or not...

from bipm-si-brochure.

Again, semantics is preventing this: (²S_1/2) is a single thing.

Our problem is that we're abusing MathML to encode chemical symbols, and while I'm getting away with it in MathJax, you're not in jEuclid.

Manipulating the MathML in the general case looks tricky: presumably, it's going to be something like:

... sorry, @Intelligent2013, I'm writing this in Nokogiri, because XSLT is something I never got comfortable with in the past 20 years:

doc.xpath('//msup[*[1][normalize-space() = ""]').each do |x|
  if p = x.previous_element
    x.elements.first.replace(p.remove)
  else
    # the <sup>y</sup>x is the first thing in the MathML
    .... ??
  end
end

A MathML expression starting with ^yx is certainly possible if we're doing chemistry in MathML, and moving preceding elements around is not going to deal with it. If there is any blank character that jEuclid will allow to take a superscript, we need it...

from bipm-si-brochure.

I've tried all kind of spaces in <mtext> and no success. jEuclid can't evaluate the height of preceding space and therefore put superscripted text on the baseline. BUT I've found a workaround solution - replace mtext to mspace like this:

• <mtext> </mtext> -> <mspace width="0.5ex" height="1ex"/>
• <mtext/> -> <mspace height="1ex"/>

In this case the resulted PDF is:

I'll try to do it via xslt.
And may be it is possible to fix alignment issue with exponent with using <mspace height="1ex"/> also...

from bipm-si-brochure.

Issue with the preceding superscripted text fixed in metanorma/mn-native-pdf@8bcd034.

Updated equations looks ok:
si-brochure-fr.presentation.mtext.pdf

from bipm-si-brochure.

@Intelligent2013 The proper solution in MathML, I have found, is to use <mphantom/> instead of <mtext/>, and I may end up replacing <mtext/> or <mtext> </mtext> with <mphantom/> in postprocessing. In case I do, could you ensure that mphantom is also processed like mspace in the PDF?

from bipm-si-brochure.

@opoudjis Empty mphantom doesn't processing at all in jEuclid.
This xml:

<math xmlns="http://www.w3.org/1998/Math/MathML">
	<mn>5</mn>
	<mtext>s</mtext>
	<msup>
		<mrow>
			<mphantom/>
		</mrow>
		<mrow>
			<mn>2</mn>
		</mrow>
	</msup>
	<msub>
		<mrow>
			<mi>S</mi>
		</mrow>
		<mrow>
			<mrow>
				<mn>1</mn>
				<mtext>/</mtext>
				<mn>2</mn>
			</mrow>
		</mrow>
	</msub>
	<mi>–</mi>
	<mn>4</mn>
	<mtext>d</mtext>
	<msup>
		<mrow>
			<mphantom/>
		</mrow>
		<mrow>
			<mn>2</mn>
		</mrow>
	</msup>
	<msub>
		<mrow>
			<mtext>D</mtext>
		</mrow>
		<mrow>
			<mrow>
				<mn>5</mn>
				<mtext>/</mtext>
				<mn>2</mn>
			</mrow>
		</mrow>
	</msub>
</math>

shows as:

In this case I have to change <mphantom/> via xslt on-fly to <mspace height="1ex"/>.

from bipm-si-brochure.

Ouch. OK. I will not use mphantom. In fact, it sounds like I will need to universally do that change in PresentationXML, if I do use mphantom in the source XML.

from bipm-si-brochure.

Based on the latest PDF provided by @opoudjis: si-brochure-fr(12-01).pdf

All the issues listed above are fixed.
However, there is an additional issue that wasn't before:

In page 73, Greek letters are used to number list items (as reported in #47).

from bipm-si-brochure.

In page 73, Greek letters are used to number list items (like reported in #47).

This issue was fixed in metanorma/mn-native-pdf#288.

from bipm-si-brochure.

Having this issue fixed, it seems everything is solved. Closing.

from bipm-si-brochure.