[1.49] Release candidate PDF doc build available
I've put a release candidate of the PDF documentation build for 1.49 here: https://sourceforge.net/projects/boost/files/boost-docs/1.49.0/ Let me know if there are any SNAFU's, Enjoy, John.
2012/2/21 John Maddock <boost.regex@virgin.net>:
Boost.Lexical_cast library has QuickBook documentation since 1.48. Can it now have a pdf doc version? What should be done?
I've updated the zip to include it, let me know if there are any issues with it.
Still can not find it at https://sourceforge.net/projects/boost/files/boost-docs/1.49.0/. Did you already uploaded the archive? Was there any issues building the lexical_casts pdf (I`ve newer tested that)? P.S.: lexical_cast docs are at libs/conversion/doc Archive has a conversion.pdf, but that is a NumericConversion documentation. Best regards, Antony Polukhin
I've updated the zip to include it, let me know if there are any issues with it.
Still can not find it at https://sourceforge.net/projects/boost/files/boost-docs/1.49.0/. Did you already uploaded the archive? Was there any issues building the lexical_casts pdf (I`ve newer tested that)?
No issues, I've updated the zip file again, asio and a couple of other libraries have been fixed as well. John.
Hi John, On 21-2-2012 16:46, John Maddock wrote:
I've put a release candidate of the PDF documentation build for 1.49 here: https://sourceforge.net/projects/boost/files/boost-docs/1.49.0/
Let me know if there are any SNAFU's,
I first (accidentally) downloaded 1.47 (announced as latest version - no problem). That includes Boost.Geometry. After that I downloaded 1.49. But that does not contain Boost.Geometry anymore. Is it possible to still include it? Do I have to take action for that? I also noted three 0-byte files (asio, intrusive and intrusive.auto_index) Thanks, Barend
After that I downloaded 1.49. But that does not contain Boost.Geometry anymore.
There was an error in building the geometry docs: fop.render.pdf ..\..\bin.v2\libs\geometry\doc\msvc-10.0\debug\threading-multi\geometry.pdf M:/download/programs/xep/xep.bat ..\..\bin.v2\libs\geometry\doc\msvc-10.0\debug\threading-multi\geometry.fo ..\..\bin.v2\libs\geometry\doc\msvc-10.0\debug\threading-multi\geometry.pdf (document [system-id file:/M:/data/boost/branches/release/doc/pdf/../../bin.v2/libs/geometry/doc/msvc-10.0/debug/threading-multi/geometry.fo] (validate [error] Empty 'fo:table-body'. [validation total: 1 error] Parse error: Invalid XSL FO source 'file:/M:/data/boost/branches/release/doc/pdf/../../bin.v2/libs/geometry/doc/msvc-10.0/debug/threading-multi/geometry.fo': 1 error found during validation No idea what the isue is, but there's some invalid FO XML somewhere.
I also noted three 0-byte files (asio, intrusive and intrusive.auto_index)
Oh :-( Will investigate, thanks, John.
Hi John, On 21-2-2012 20:36, John Maddock wrote:
After that I downloaded 1.49. But that does not contain Boost.Geometry anymore.
There was an error in building the geometry docs:
fop.render.pdf ..\..\bin.v2\libs\geometry\doc\msvc-10.0\debug\threading-multi\geometry.pdf
M:/download/programs/xep/xep.bat ..\..\bin.v2\libs\geometry\doc\msvc-10.0\debug\threading-multi\geometry.fo ..\..\bin.v2\libs\geometry\doc\msvc-10.0\debug\threading-multi\geometry.pdf
(document [system-id file:/M:/data/boost/branches/release/doc/pdf/../../bin.v2/libs/geometry/doc/msvc-10.0/debug/threading-multi/geometry.fo] (validate [error] Empty 'fo:table-body'. [validation total: 1 error] Parse error: Invalid XSL FO source 'file:/M:/data/boost/branches/release/doc/pdf/../../bin.v2/libs/geometry/doc/msvc-10.0/debug/threading-multi/geometry.fo': 1 error found during validation
No idea what the isue is, but there's some invalid FO XML somewhere.
OK that will be the same issue as Daniel had. I thought to refer to Boost.Build as "../../../b2" now because of Boost Trunk/Release incompatibilies w.r.t. project.jam. But that was not so wise... Daniel said he would manage, will I revert it (make_qbk.py) or can we solve it otherwise? Regards, Barend
OK that will be the same issue as Daniel had. I thought to refer to Boost.Build as "../../../b2" now because of Boost Trunk/Release incompatibilies w.r.t. project.jam.
But that was not so wise...
Daniel said he would manage, will I revert it (make_qbk.py) or can we solve it otherwise?
No idea, I'm just running the scripts, and I haven't been able to figure out what the issue is. If you can let me have an alternative version of make_qbk.py I can see if it solves the problem though. Cheers, John.
Hi John, On 22-2-2012 11:15, John Maddock wrote:
OK that will be the same issue as Daniel had. I thought to refer to Boost.Build as "../../../b2" now because of Boost Trunk/Release incompatibilies w.r.t. project.jam.
But that was not so wise...
Daniel said he would manage, will I revert it (make_qbk.py) or can we solve it otherwise?
No idea, I'm just running the scripts, and I haven't been able to figure out what the issue is. If you can let me have an alternative version of make_qbk.py I can see if it solves the problem though.
This was the only structural change. So herewith the alternative version. This is in fact the Trunk version but contains the (original) reference to bjam. Thanks, Barend
No idea, I'm just running the scripts, and I haven't been able to figure out what the issue is. If you can let me have an alternative version of make_qbk.py I can see if it solves the problem though.
This was the only structural change.
So herewith the alternative version. This is in fact the Trunk version but contains the (original) reference to bjam.
Nope same error: fop.render.pdf ..\..\..\bin.v2\libs\geometry\doc\msvc-10.0\debug\threading-multi \geometry.pdf (document [system-id file:/M:/data/boost/branches/release/libs/geometry/doc/../. ./../bin.v2/libs/geometry/doc/msvc-10.0/debug/threading-multi/geometry.fo] (validate [error] Empty 'fo:table-body'. [validation total: 1 error] Parse error: Invalid XSL FO source 'file:/M:/data/boost/branches/release/libs/geometry/doc/../../../bin.v2/libs/geometry/doc/msvc-10.0/debug/threading-multi/geometry.fo': 1 error found during validation HTML builds OK though. Somewhere there's a table with no content being generated, and the FO processor doesn't like it, don't know where that's coming from though... John.
So herewith the alternative version. This is in fact the Trunk version but contains the (original) reference to bjam.
OK, I've found the issue with the geometry docs: in your indexes you have a table entry for "1-dimensional (adapted)" which itself contains another table.... but that table is completely empty. Down the line this produces the FO: <fo:table id="id733492" space-before.optimum="1em" space-before.minimum="0.8em" space-before.maximum="1.2em" table-layout="fixed" width="100%"> <fo:table-column column-number="1" column-width="proportional-column-width(1)"/> <fo:table-body start-indent="0pt" end-indent="0pt"/> </fo:table> Which is invalid XML and gets rejected by XEP. Am I correct in thinking there should really be something in that table anyway? HTH, John.
Hi John, Thanks for your research. On 23-2-2012 11:31, John Maddock wrote:
So herewith the alternative version. This is in fact the Trunk version but contains the (original) reference to bjam.
OK, I've found the issue with the geometry docs: in your indexes you have a table entry for "1-dimensional (adapted)" which itself contains another table.... but that table is completely empty. Down the line this produces the FO:
<fo:table id="id733492" space-before.optimum="1em" space-before.minimum="0.8em" space-before.maximum="1.2em" table-layout="fixed" width="100%"> <fo:table-column column-number="1" column-width="proportional-column-width(1)"/> <fo:table-body start-indent="0pt" end-indent="0pt"/> </fo:table>
Which is invalid XML and gets rejected by XEP.
Am I correct in thinking there should really be something in that table anyway?
Yes and no, there are currently no explicitly 1-dimensional adapted features. There are (adaptors for Boost.Ranges) but they are for both polygon (2-dim) and linestring (1-rim) and their documentation is still lacking. So I removed the simplelist in the Boost Book XML (which is used for the quick-reference). I merged this (this file only) to the release branch, so hope you can try it again with this update. I checked but this was the only empty simplelist (sub-table) in this file. The strange thing is that this (the empty table) was also the case for 1.47 (where the PDF was generated successfully). Anyway, hope that this will solve it. Regards, Barend
Am I correct in thinking there should really be something in that table anyway?
Yes and no, there are currently no explicitly 1-dimensional adapted features. There are (adaptors for Boost.Ranges) but they are for both polygon (2-dim) and linestring (1-rim) and their documentation is still lacking.
So I removed the simplelist in the Boost Book XML (which is used for the quick-reference). I merged this (this file only) to the release branch, so hope you can try it again with this update.
That gets it - works now, zip file updated to include geometry.pdf.
I checked but this was the only empty simplelist (sub-table) in this file.
The strange thing is that this (the empty table) was also the case for 1.47 (where the PDF was generated successfully). Anyway, hope that this will solve it.
Ah, I did upgrade XEP at one point, so maybe it's got stricter :-( Cheers, John.
participants (3)
-
Antony Polukhin -
Barend Gehrels -
John Maddock