Обсуждение: Possible to display Unicode symbols in our SGML docs?
I'm currently looking at the latest psql-unicode-output patch, which
includes this bit of proposed documentation:
+ Newlines in data are shown using a carriage return symbol
+ (<literal>↵</literal>) in the right-hand margin.
+ Wrapped data uses an ellipsis symbol
+ (<literal>…</literal>) in the right-hand margin of a
+ wrapped line, and in the left-hand margin of the following
+ continuation line.
Unfortunately this markup seems to be pure speculation on the author's
part:
openjade -wall -wno-unused-param -wno-empty -wfully-tagged -D . -D . -c
/usr/share/sgml/docbook/dsssl-stylesheets/catalog-d stylesheet.dsl -t sgml -i output-html -V html-index postgres.sgml
openjade:ref/psql-ref.sgml:1797:22:E: "8629" is not a character number in the document character set
openjade:ref/psql-ref.sgml:1799:22:E: "8230" is not a character number in the document character set
make: *** [HTML.index] Error 1
Is there a way around that, or should I just delete the attempts
to show the specific symbols?
Note that this is in psql's reference page, which means that anything
we want to do here also has to translate to man-page output. That
might kill the idea even if it would work otherwise.
regards, tom lane
On lör, 2009-11-21 at 17:05 -0500, Tom Lane wrote: > I'm currently looking at the latest psql-unicode-output patch, which > includes this bit of proposed documentation: > > + Newlines in data are shown using a carriage return symbol > + (<literal>↵</literal>) in the right-hand margin. > + Wrapped data uses an ellipsis symbol > + (<literal>…</literal>) in the right-hand margin of a > + wrapped line, and in the left-hand margin of the following > + continuation line. I don't think there is a strong need to document this at all. Do we say anywhere that columns are separated by vertical lines? > Unfortunately this markup seems to be pure speculation on the author's > part: > > openjade -wall -wno-unused-param -wno-empty -wfully-tagged -D . -D . -c /usr/share/sgml/docbook/dsssl-stylesheets/catalog-d stylesheet.dsl -t sgml -i output-html -V html-index postgres.sgml > openjade:ref/psql-ref.sgml:1797:22:E: "8629" is not a character number in the document character set > openjade:ref/psql-ref.sgml:1799:22:E: "8230" is not a character number in the document character set > make: *** [HTML.index] Error 1 > > Is there a way around that, or should I just delete the attempts > to show the specific symbols? DocBook SGML is limited to something like Latin 1. So delete it.
On Sat, Nov 21, 2009 at 11:33 PM, Peter Eisentraut <peter_e@gmx.net> wrote: > On lör, 2009-11-21 at 17:05 -0500, Tom Lane wrote: >> I'm currently looking at the latest psql-unicode-output patch, which >> includes this bit of proposed documentation: >> >> + Newlines in data are shown using a carriage return symbol >> + (<literal>↵</literal>) in the right-hand margin. >> + Wrapped data uses an ellipsis symbol >> + (<literal>…</literal>) in the right-hand margin of a >> + wrapped line, and in the left-hand margin of the following >> + continuation line. > > I don't think there is a strong need to document this at all. Do we say > anywhere that columns are separated by vertical lines? I think it's not intuitively obvious which is which here. It would be nice to mention it. >> Is there a way around that, or should I just delete the attempts >> to show the specific symbols? > > DocBook SGML is limited to something like Latin 1. So delete it. You could always write it out longhand "carriage return (code point 8269)" -- greg
Greg Stark <gsstark@mit.edu> writes:
> On Sat, Nov 21, 2009 at 11:33 PM, Peter Eisentraut <peter_e@gmx.net> wrote:
>> I don't think there is a strong need to document this at all. �Do we say
>> anywhere that columns are separated by vertical lines?
> I think it's not intuitively obvious which is which here. It would be
> nice to mention it.
If it were easy to do, I'd be for doing it.
>> DocBook SGML is limited to something like Latin 1. �So delete it.
> You could always write it out longhand "carriage return (code point 8269)"
I don't think mentioning the numeric code point is going to do anything
for anybody. If we can't actually show the symbol, then just saying
"carriage return symbol" seems like the best bet. It might be worth
making sure that the phrase exactly matches the Unicode standard's name
for the glyph, but otherwise I won't sweat it.
regards, tom lane
On lör, 2009-11-21 at 23:27 -0500, Tom Lane wrote: > Greg Stark <gsstark@mit.edu> writes: > > You could always write it out longhand "carriage return (code point 8269)" > > I don't think mentioning the numeric code point is going to do anything > for anybody. If we can't actually show the symbol, then just saying > "carriage return symbol" seems like the best bet. It might be worth > making sure that the phrase exactly matches the Unicode standard's name > for the glyph, but otherwise I won't sweat it. 8629 = U+21B5 = downwards arrow with corner leftwards - may indicate a carriage return or new line -> 23CE return symbol 8230 = U+2026 = horizontal ellipsis