Re: [xml2rfc-dev] <contact> element in <t>

[Henrik Levkowetz <[e-mail]>]

Hi Martin,

On 2019-12-11 03:16, Martin Thomson wrote:
> The XML in question is produced from kramdown, but you can see the input here:
> 
> https://github.com/quicwg/base-drafts/compare/contributors?expand=1
> 
> This expands into pretty much what you see below.

I'm afraid that's not too helpful -- it really would make a difference
if you could pull out the XML file which exhibits the problem and send
it to me.

	Henrik

> 
> On Wed, Dec 11, 2019, at 12:39, Henrik Levkowetz wrote:
>> Hi Martin,
>> 
>> On 2019-12-11 02:13, Martin Thomson wrote:
>> > Hi Henrik,
>> > 
>> > I just tested this new element out in xml2rfc 2.37.0.  I think that there are some bugs to work out.
>> > 
>> > I have the following input (in reduced form):
>> > 
>> > <t>The following people helped:
>> > ... (ascii names) ...,
>> > <contact fullname="Mirja Kühlewind"/>,
>> > ...
>> > <contact asciiFullname="Kazuho Oku" fullname="?"/>,
>> > ...
>> > </t>
>> > 
>> > I don't have Kazuho's real name just yet, so I just added "?" for the moment.
>> > 
>> > What I have observed is that the output includes many additional line
>> > breaks. If I get this right, this is producing a new paragraph break
>> > for the name included in <contact>. That's not expected.
>> 
>> No, and I haven't seen this here in my tests.  Could you send your XML?
>> 
>> > It also appears to remove the "?" I'm using as a placeholder in this
>> > example. I can put some Unicode in fullname (for instance, "øøø"
>> > works fine), but the Japanese text I copied from somewhere failed.
>> > 
>> > fullname="投稿者" causes xml2rfc to choke:
>> > 
>> >   File "/home/martin/bin/xml2rfc", line 11, in <module>
>> >     load_entry_point('xml2rfc==2.37.0', 'console_scripts', 'xml2rfc')()
>> >   File "/home/martin/.local/lib/python3.8/site-packages/xml2rfc/run.py", line 590, in main
>> >     writer.write(filename)
>> >   File "/home/martin/.local/lib/python3.8/site-packages/xml2rfc/writers/text.py", line 274, in write
>> >     lines = self.paginate(lines)
>> >   File "/home/martin/.local/lib/python3.8/site-packages/xml2rfc/writers/text.py", line 412, in paginate
>> >     paginated[i].elem.page = page
>> > AttributeError: 'lxml.etree._Element' object has no attribute 'page'
>> > 
>> > I thought that that specific error might be due to placement, but
>> > moving the text around didn't help.
>> 
>> Interesting.  I haven't seen this in my tests either.  I used Georgian
>> Mkhedruli though, not Japanese Kanji.  I'll try that tomorrow, fix, and do a
>> bugfix release.  I see this is related to pagination, so again your XML file
>> might help debug this.
>> 
>> 
>> 	Henrik
>> 
>> 
>> Attachments:
>> * signature.asc
>