[sword-devel] .conf files encoding/tags
Troy A. Griffitts
scribe at crosswire.org
Sat Oct 6 12:36:00 MST 2007
Thanks for all the work DM. A few quick comments (I'm trying to package
up the promised 1.5.10RC1 for weekend test).
continuation \ is handled in FileMgr getLine, if I remember correct. So
the filters don't have to deal with it. They are quite acceptable in
.conf files. Not sure who discouraged against it, and I'm not sure
referencing the history of sword-devel is necessarily helpful as it's
likely someone at sometime has recommended or discouraged just about
XML .conf files would provide validation. This would be a plus.
Current properties-like impl works, is simple to parse, and mandates
simple key=value pairs, which is how configuration logic is coded in SWORD.
We can write a validator for .conf files.
My vote is to keep the properties-like impl.
DM Smith wrote:
> On Oct 3, 2007, at 7:46 PM, Chris Little wrote:
>> .conf files are entirely plain text except for the About field,
>> which is
>> RTF. RTF is only used in the About field and only RTF (or no
>> markup) may
>> be used in the About field.
>> The use of RTF here is basically a legacy issue carried from BibleCS.
>> It shouldn't be a big deal for you because I believe we only use 4
>> different tags:
>> \qc (center the following)
>> \par (paragraph break)
>> \pard (paragraph break + reset formatting)
>> \uXXXXX? (non-CP1252 characters expressed as UTF-16)
> I checked the rtfhtml filter and it only handles.
> \pard, \par, \qc
> \qc is treated as a <center> and it is noted that the parser is "in
> \pard is treated as a </center> if in center and nothing else. It is
> not treated as a paragraph break.
> \par is a paragraph start. No </p> is ever output.
> \ followed by a space is ignored and \ at the end of the line is also.
> All other \ are output as plain text, including the \.
>> Eeli Kaikkonen wrote:
>>> I browsed through the beta area module .conf files. It's great to see
>>> so many new ones with new features.
>>> One rant I have. Why on earth is stupid braindead rtf or other
>>> strange formatting used in .conf files? See this example from TurNTB:
>>> About=New Turkish Bible translation, jointly translated and
>>> published by
>>> K\u00305?tab\u00305? Mukaddes \u00350?irketler
>>> and Yeni Ya\u00351?am Yay\u00305?nlar\u00305?
>>> (www.yyyayinlari.com). We
>>> are grateful for the permission by Yeni Ya\u00351?am
>>> Yay\u00305?nlar\u00305? to distribute this translation.
>>> How are the frontends supposed to display this correctly? This is
>>> 2007 and this kind of project should use utf8 in conf files also.
>>> is quite easy to replace but why not use <br> instead, that kind
>>> of html
>>> tagging is used in many places even without real html browsers.
>>> Rtf is
>>> M$ proprietary format.
>> So you propose that we abandon a markup system already in place and
>> convert everything to a different arbitrarily selected markup
>> format? As
>> a result, every user will have to update every module, or the about
>> will be mis-rendered. Existing strategies for rendering RTF will
>> have to
>> be re-written to handle <insert arbitrarily chosen markup
>> language>. And
>> all to solve a problem that doesn't exist.
>> Suggesting that we should not use RTF because it comes from MS is
>> RTF is quite well documented and widely used. It's cross-platform and
>> wasn't even developed by MS. A full description may be found at
>> Converting the RTF we use to HTML requires about three lines of Perl,
>> which you can port to the language of your choice:
>> $about =~ s/\u(\d+)\?/pack("U", $1)/eg; # assumes no surrogate pairs
>> $about =~ s/\qc ?(.*?)(\pard|$)/<center>$1<\/center>$2/g;
>> $about =~ s/\pard? ?/<br\/>/g;
>> sword-devel mailing list: sword-devel at crosswire.org
>> Instructions to unsubscribe/change your settings at above page
> sword-devel mailing list: sword-devel at crosswire.org
> Instructions to unsubscribe/change your settings at above page
More information about the sword-devel