wif2mwiki.cg

#
# Copyright (C) 2010 Russ Tremain.
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the Artistic License 2.0 or later.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# Artistic License 2.0 for more details.
#
# You should have received a copy of the Artistic License the file
# LICENSE.  If not, see:
#     <http://www.perlfoundation.org/artistic_license_2_0>
#

#
#wif2vq.cg - convert a topic in wiki intermediate form to vqwiki form.
#

%ifndef CG_ROOT CG_ROOT = bld/mediawiki

%include mediawiki.defs

init_globals := << //
{
    #universal token to temporarily delimit our parsable items.
    #set to ascii formfeed (^L).
    UNIVTOK = $HEX_0C:nameof:hexdecode
    #set to ascii BELL (^G):
    UNIVTOK2 = $HEX_07:nameof:hexdecode

    #other possibilities:  \x1b esc, \x7f del

    DS := $
#%eecho DS='$DS' UNIVTOK='$UNIVTOK'

    tab = $tab:clr:tab
    nl = $nl:clr:newline
}
//

mediawiki_edits := << //
#perform special edits on WIF form before we do the convertions.
#in particular, convert html one-line hrefs to named wiki topics.
{
    #example input:  {=VQ_HTML_LQ=}<A HREF="http://www.cnn.com/">CNN</A>{=VQ_HTML_RQ=}
    #example output: {=MW_EXTERNAL_LQ=}http://www.cnn.com/ CNN{=MW_EXTERNAL_RQ=}

    #should be used with ignore-case option:
    href_pattern := <a\s+href\s*=\s*"([^"]*)"\s*>([^<]*)<\/?a>

    CG_MATCH_SPEC = /{=VQ_HTML_LQ=}\s*$href_pattern\s*{=VQ_HTML_RQ=}/i
    %ifnot $INPUT:match %return

    CG_SUBSTITUTE_SPEC = s/({=VQ_HTML_LQ=}|{=VQ_HTML_RQ=})/$UNIVTOK/gs
    %void $INPUT:substitute:assign

    #must match only one href:
    CG_SUBSTITUTE_SPEC2 := s/^\s*<a\s+href\s*=\s*"([^"]*)"\s*>([^<]*)<\/?a>\s*$/$1 $2/igs
    CG_SUBSTITUTE_SPEC  = s/$UNIVTOK([^$UNIVTOK]*)$UNIVTOK/sprintf("{=MW_EXTERNAL_LQ=}%s{=MW_EXTERNAL_RQ=}", &s2_op(${DS}1))/egs
    %void $INPUT:substitute:assign

ECHO_TXT=$INPUT
echo	/mediawiki_edits01.txt
}
//

##################################### MAIN #####################################

%call init_globals
WIFDIR = bld/cgtstroot/wif

%ifndef TOPIC_INFILE  TOPIC_INFILE = bld/wif/MyLeftMenu.txt
%ifndef TOPIC_INFILE  TOPIC_INFILE = $WIFDIR/wikiconvert00002.txt

#convert TOPIC to standard form, as it is written out by vq2toc:
CG_SUBSTITUTE_SPEC = s/\+/ /g
TOPIC_IN = $TOPIC_INFILE:basename:root:substitute:urldecode:urlencode
TOPIC_INFILE = ${TOPIC_INFILE:dirname}/${TOPIC_IN}.txt

###### exit if file not found:
%ifnot $TOPIC_INFILE:r %return -e -s 1 ${CG_INFILE}:  cannot open '$TOPIC_INFILE' - abort.

#read in and evaluate the file in the presence of the vqwiki defs:
%readtemplate INPUT $TOPIC_INFILE

#mark VQ_HTML variables as reflexive, and eval the rest:
VQ_HTML_LQ := {=VQ_HTML_LQ=}
VQ_HTML_RQ := {=VQ_HTML_RQ=}
%evalmacro INPUT INPUT

#now edit the html blocks...
%call mediawiki_edits

#now reset original mediawiki translation defs:
%include mediawiki.defs

#now finish the transform:
ECHO_TXT = $INPUT
%evalmacro * ECHO_TXT ECHO_TXT
echo	/$TOPIC_INFILE:basename