X-Git-Url: https://git.plutz.net/?a=blobdiff_plain;f=handlers%2F40_search.sh;h=3661f954b7e07c6665f419484bb00c2edb9e7a90;hb=20c6ce6a3de621383d1b6d985757366590722005;hp=57117e3834a1bbe0c816f92e785bc38c188bc186;hpb=3260c24fcce2db7d0b0eb98234c5ba385c6f55a9;p=shellwiki diff --git a/handlers/40_search.sh b/handlers/40_search.sh index 57117e3..3661f95 100644 --- a/handlers/40_search.sh +++ b/handlers/40_search.sh @@ -17,114 +17,5 @@ [ "$SEARCH_INDEX" != true ] && return 1 [ "${PATH_INFO%\[search\]}" = "$PATH_INFO" ] && return 1 -. "$_EXEC/cgilite/storage.sh" -. "$_EXEC/cgilite/db23.sh" - -I="$_DATA/index" -tags="$( GET q | awk ' - BEGIN { # Field separator FS should include punctuation, including Unicode Block U+2000 - U+206F - if ( length("¡") == 1 ) # Utf-8 aware AWK - FS = "([] \\t\\n\\r!\"'\''()*+,./:;<=>?\\\\^_`{|}~[-]|%[0-9A-Fa-f]{2}|'"$(printf '[\342\200\200-\342\201\257]')"')+"; - else # UTF-8 Hack - FS = "([] \\t\\n\\r!\"'\''()*+,./:;<=>?\\\\^_`{|}~[-]|%[0-9A-Fa-f]{2}|'"$(printf '\342\200[\200-\277]|\342\201[\201-\257]')"')+"; - fi - } - { for (n = 1; n <= NF; n++) if ($n ~ /#[[:alnum:]_]+/) { - sub(/^#/,"",$n) - printf "%s ", toupper($n); - } - } -')" - -ntags="$( GET q | awk ' - BEGIN { # Field separator FS should include punctuation, including Unicode Block U+2000 - U+206F - if ( length("¡") == 1 ) # Utf-8 aware AWK - FS = "([] \\t\\n\\r\"#'\''()*+,./:;<=>?\\\\^_`{|}~[-]|%[0-9A-Fa-f]{2}|'"$(printf '[\342\200\200-\342\201\257]')"')+"; - else # UTF-8 Hack - FS = "([] \\t\\n\\r\"#'\''()*+,./:;<=>?\\\\^_`{|}~[-]|%[0-9A-Fa-f]{2}|'"$(printf '\342\200[\200-\277]|\342\201[\201-\257]')"')+"; - fi - } - { for (n = 1; n <= NF; n++) if ($n ~ /![[:alnum:]_]+/) { - sub(/^!/,"",$n) - printf "%s ", toupper($n); - } - } -')" - -words="$( GET q | awk ' - BEGIN { # Field separator FS should include punctuation, including Unicode Block U+2000 - U+206F - if ( length("¡") == 1 ) # Utf-8 aware AWK - FS = "([] \\t\\n\\r\"#'\''()*+,./:;<=>?\\\\^_`{|}~[-]|%[0-9A-Fa-f]{2}|'"$(printf '[\342\200\200-\342\201\257]')"')+"; - else # UTF-8 Hack - FS = "([] \\t\\n\\r\"#'\''()*+,./:;<=>?\\\\^_`{|}~[-]|%[0-9A-Fa-f]{2}|'"$(printf '\342\200[\200-\277]|\342\201[\201-\257]')"')+"; - fi - } - { for (n = 1; n <= NF; n++) if ($n !~ /![[:alnum:]_]+/) { - sub(/!/," ",$n) - printf "%s ", tolower($n); - } - } -')" - -searchteaser() { - local file="$1" words db3_data - local w l nc nl hits mhits cont mcont - shift 1; words="$*" - - for w in ${words}; do - grep -hiwnF "$w" "$file" - done \ - | sort -t: -k1 -n \ - | { nc=-1 hits=0 mhits=0 - while read -r l; do - nl="$nc" nc="${l%%:*}" - if [ $nc -eq $nl ]; then - hits=$((hits + 1)) - elif [ $nc -eq $((nl + 1 )) ]; then - hits=$((hits + 1)) - cont="${cont}${BR}${l#*:}" - elif [ $hits -gt $mhits ]; then - mhits="$hits" mcont="$cont" - hits=1 cont="${l#*:}" - else - hits=1 cont="${l#*:}" - fi - done - - [ $hits -gt $mhits ] \ - && STRING "$cont" \ - || STRING "$mcont" - } -} - -for w in ${words}; do - [ ! -f "$I/$w" ] && continue - - while read date doc freq num total; do - P="$_DATA/pages$(UNSTRING "$doc")" - d="$(stat -c %Y -- "$P/#index.flag" 2>&-)" - [ "$d" -le "$date" -a -f "$P/#page.md" ] 2>&- || continue - - printf '%s %f\n' "$doc" "$freq" - done <"$I/$w" -done \ -| awk ' - { cnt[$1]++; weight[$1] = weight[$1] ? weight[$1] + $2 : $2; } - END { m = 0; for (d in cnt) m = ( m < cnt[d] ) ? cnt[d] : m; - for (d in cnt) if ( cnt[d] == m ) printf "%f %s\n", weight[d], d; - } -' \ -| sort -nr \ -| while read freq doc; do - page="$(UNSTRING "$doc")" - [ "${page%*/\[*\]/*}" != "$page" ] && continue - if [ "$LANGUAGE_DEFAULT" ]; then - [ -d "${_DATA}/pages/${page}/:${LANGUAGE}/" ] && continue - [ "${page%/:*/}" = "${page%/:${LANGUAGE}/}" ] || continue - fi - acl_read "$page" || continue - has_tags "$page" $tags || continue - has_tag "$page" $ntags && continue - printf '%s %s\n' "$doc" "$(searchteaser "$(mdfile "$page")" $words)" -done \ -| theme_search "${words% }" +theme_page "/[wiki]/search/" "$(page_title "/[wiki]/search/")" +return 0