#!/bin/bash TIMEFORMAT=" took %lR using %P%% CPU" time { pushd /opt/merged rm -fr Foswiki/* cp -r /opt/merged_EMPTY/Foswiki . rm -fr PmWiki/* cp -r /opt/merged_EMPTY/PmWiki . rm -fr unsorted mkdir -p unsorted rm -fr users mkdir -p users filter=" -name _default -prune -o \ -name _empty -prune -o \ -name System -prune -o \ -name Trash -prune -o \ -name TWiki -prune -o \ " URL="https://fos.wiki.devuan.org" time find /opt/Foswiki/data ${filter} \ -name "*.txt" -type f,l -printf "%P\n" | while read line do base=`echo "${line}" | cut -d '/' -f 1` file=`echo "${line}" | cut -d '/' -f 2- | rev | cut -b 5- | rev` if [[ ! $file =~ (AdminGroup|AdminUser|AdminUserLeftBar|EditorGroup|GroupTemplate|GroupViewTemplate|NobodyGroup|PatternSkinUserViewTemplate|ProjectContributor|RegistrationAgent|SitePreferences|UnprocessedRegistrations|UnprocessedRegistrationsLog|UserHomepageHeader|UserList|UserListByDateJoined|UserListByLocation|UserList|UserListHeader|WebAtom|WebChanges|WebCreateNewTopic|WebHome|WebIndex|WebLeftBar|WebLeftBarExample|WebNotify|WebPreferences|WebRss|WebSearch|WebSearchAdvanced|WebTopicList|WikiGroups|WikiUsers)$ ]]; then time=`date --rfc-3339=seconds -ur /opt/Foswiki/data/${base}/${file}.txt | cut -d '+' -f 1` mkdir -p Foswiki/$base mkdir -p Foswiki/${base}/`dirname ${file}` echo -e "ogWiki=Foswiki\nogURL=${URL}\nogBase=${base}\nogFile=${file}\ntimestamp=${time}\n" > Foswiki/${base}/${file}.md.md echo "downloading ${URL}/${base}/${file}?cover=print" # Doesn't help with redownloads, coz natch a dynamic site isn't cached. But I can at least comment out the curl command during testing to save time. curl --silent --no-progress-meter ${URL}/${base}/${file}?cover=print -o Foswiki/${base}/${file}.HTM # Attempt to separate user profiles from user content. Doesn't work when people turn their profiles into content. if [[ "${base}" == "Main" ]]; then dest="unsorted" mkdir -p `dirname users/${file}` sed -i -E Foswiki/${base}/${file}.HTM -e "s%UserForm%%w users/${file}_fos.SED" if [ -s users/${file}_fos.SED ]; then dest="users" else rm users/${file}_fos.SED rm -d `dirname users/${file}` >/dev/null 2>&1 fi mkdir -p `dirname ${dest}/${file}` touch Foswiki/${base}/${file}.md ln -sfr Foswiki/${base}/${file}.md ${dest}/${file}_fos.md ln -sfr Foswiki/${base}/${file}.md.md ${dest}/${file}_fos.md.md rm Foswiki/${base}/${file}.md fi fi done filter=" -not -name "*~" -a \ -not -name ".flock" -a \ -not -name ".htaccess" -a \ -not -name ".lastmod" -a \ -not -name ".pageindex" -a \ " URL="https://wiki.devuan.org" time find /opt/pmwiki/wiki.d ${filter} \ -name "*.*" -type f,l -printf "%P\n" | while read line do base=`echo "${line}" | cut -d '.' -f 1` if [[ "${base}" != "Site" ]]; then file=`echo "${line}" | cut -d '.' -f 2` time=`date --rfc-3339=seconds -ur /opt/pmwiki/wiki.d/${base}.${file} | cut -d '+' -f 1` mkdir -p PmWiki/$base echo -e "ogWiki=PmWiki\nogURL=${URL}\nogBase=${base}\nogFile=${file}\ntimestamp=${time}\n" > PmWiki/${base}/${file}.md.md # echo "downloading ${URL}/?n=${base}.${file}?action=markdown" # curl --no-progress-meter ${URL}/?n=${base}.${file}?action=markdown -o PmWiki/${base}/${file}.MARKDOWN echo "downloading ${URL}/?n=${base}.${file}?action=print" curl --no-progress-meter ${URL}/?n=${base}.${file}?action=print -o PmWiki/${base}/${file}.HTM # Seems there's no way to tell user profiles apart from user content. Unless I can find a list of users somewhere. Don't think there is one. if [[ "${base}" == "Profiles" ]]; then touch PmWiki/${base}/${file}.md ln -sfr PmWiki/${base}/${file}.md unsorted/${file}_pm.md ln -sfr PmWiki/${base}/${file}.md unsorted/${file}_pm.md.md rm PmWiki/${base}/${file}.md fi # TODO - groups are PmWiki/Onefang and PmWiki/Tiki # pandoc -f markdown -t commonmark_x --self-contained PmWiki//${base}/${file}.MD >PmWiki/${base}/${file}.md # pandoc -f html -t commonmark_x --self-contained PmWiki//${base}/${file}.HTM >PmWiki/${base}/${file}.md fi done time notYetAnotherWiki.lua popd }