Quite a big change. I moved the list logic into the grammar so that it's
easier to tell how to parse different blocks. Also reworked some of the
block vs. phrase stuff - it's a lot cleaner now which helped implement
this. It generates terrible markup at the moment, but at least the
parser is in place.
[SVN r75446]
Since values are now used in most places, a lot of the complexity became
unnecessary. Some of the string stuff in values.cpp is now redundant,
but I'll leave it alone for now.
[SVN r75358]
It's only used there so it fits into main_grammar's local state. It
would probably be a good idea to move more parsing stuff out of actions.
Or maybe the document state should be moved out of actions.
[SVN r75268]
Files are now permanently loaded. Quickbook substrings are stored as
references into the file. Now positions are stored as iterators into
the original file, the line and column is calculated when messages are
output.
This doesn't have much effect on efficiency but it simplifies a few
things.
[SVN r75253]
This encapsulates more in the id and section handling code. Fixes a few
edge cases and separates implementation of the new and old section
handling code. Always normalizes and truncates generated ids in quickbook
1.6.
[SVN r75251]
And shuffle things round a bit to avoid adding too many dependents to
id_generator. For 1.6 I want section_info to use id_generator to nest
ids properly when there are duplicate section ids.
[SVN r75246]
Imports were failing because doc info expected values. The process_state
stuff is too fragile, so instead just process the contents of a
conditional phrase and discard them afterwards. Still suppresses all
elements.
[SVN r71062]
Pretty clumsy, but it's a dodgy thing to do. Maybe the values could
include whether they're blocks or phrases and actions could react
accordingly.
[SVN r71060]
If an included file has a docinfo block, use that docinfo. I'm not
entirely sure if the nested file be treated as if it's completely
standalone - i.e. should macros and templates defined in the parent be
used?
Also, there's some odd behaviour in quickbook if the docinfo block in an
included file has a parse error. It just goes back to the start and
parses as a file without a docinfo block. I think it would be better if
this was a hard error, but what should happen in someone has defined a
template with the same name as a docinfo block? That's actually valid in
existing quickbook, so the dodgy docinfo block might just be a template
call.
[SVN r70963]
This is more consistent than it is at the moment. For the __FILENAME__
macro, use the path relative to the original file or the location last
matched on the include path. This ensures that the macro is consistent
no matter which directory you call quickbook from.
Also enable `__FILENAME__` in debug mode - so I can test this.
[SVN r70230]
Rather than a switch case, the final version will probably use something
like `std::map<value_tag, boost::function<...> >` to dispatch the
actions.
[SVN r69167]
