X-Git-Url: http://git.indexdata.com/?p=tclrobot.git;a=blobdiff_plain;f=robot.tcl;h=968e64e14bd24eac406e5a345aad3fc20182dbb0;hp=fa3c595319cb531f2fddc0d5b1e0f3d79a459d52;hb=5a6e4380b9b39674fb63839a5aeaa23cb68537cb;hpb=73e20940379b3da1488bba449758f9a48833e65a diff --git a/robot.tcl b/robot.tcl index fa3c595..968e64e 100755 --- a/robot.tcl +++ b/robot.tcl @@ -1,5 +1,5 @@ #!/usr/bin/tclsh -# $Id: robot.tcl,v 1.17 2001/06/07 08:10:10 adam Exp $ +# $Id: robot.tcl,v 1.36 2003/06/10 11:55:18 adam Exp $ # proc RobotFileNext1 {area lead} { # puts "RobotFileNext1 area=$area lead=$lead" @@ -49,28 +49,34 @@ proc RobotReadRecord {inf fromurlx distancex} { set fromurl [string trim [gets $inf]] } -proc RobotFileNext {area} { - global robotSeq global idleTime ns +proc RobotFileNext {task area} { + global robotSeq + global idletime ns + global status - # puts "RobotFileNext robotSeq=$robotSeq" - if {$robotSeq < 0} { + # puts "RobotFileNext robotSeq=$robotSeq($task)" + if {$robotSeq($task) < 0} { return {} } - if {$robotSeq == 0} { - if {[catch {set ns [glob ${area}/*]}]} { - return {} + if {$robotSeq($task) == 0} { + if {[catch {set ns($task) [glob $task/$area/*]}]} { + return done } } - set off [string length $area] + # puts "ns=$ns($task)" + set off [string length $task/$area] incr off - set n [lindex $ns $robotSeq] + set n [lindex $ns($task) $robotSeq($task)] + # puts "n=$n" if {![string length $n]} { - set robotSeq -1 + set robotSeq($task) -1 flush stdout - puts "------------ N E X T R O U N D --------" + set statusfile [open $task/status w] + puts $statusfile "$status($task,unvisited) $status($task,bad) $status($task,visited)" + close $statusfile return wait } - incr robotSeq + incr robotSeq($task) if {[file isfile $n/frobots.txt]} { puts "ok returning http://[string range $n $off end]/robots.txt" return http://[string range $n $off end]/robots.txt @@ -81,24 +87,31 @@ proc RobotFileNext {area} { } } puts "no more work at end of RobotFileNext n=$n" - puts "ns=$ns" + puts "ns=$ns($task)" return {} } -proc RobotFileExist {area host path} { - # puts "RobotFileExist begin area=$area host=$host path=$path" +proc RobotFileExist {task area host path} { + global debuglevel + + if {$debuglevel > 3} { + puts "RobotFileExist begin area=$area host=$host path=$path" + } set lpath [split $path /] set l [llength $lpath] incr l -1 set t [lindex $lpath $l] incr l -1 - set npath $area/$host[join [lrange $lpath 0 $l] /d]/f$t - # puts "RobotFileExist end npath=$npath" + set npath $task/$area/$host[join [lrange $lpath 0 $l] /d]/f$t + if {$debuglevel > 3} { + puts "RobotFileExist end npath=$npath" + } return [file exists $npath] } -proc RobotFileUnlink {area host path} { +proc RobotFileUnlink {task area host path} { + global status # puts "RobotFileUnlink begin" # puts "area=$area host=$host path=$path" set lpath [split $path /] @@ -106,13 +119,15 @@ proc RobotFileUnlink {area host path} { incr l -1 set t [lindex $lpath $l] incr l -1 - set npath $area/$host[join [lrange $lpath 0 $l] /d]/f$t + set npath $task/$area/$host[join [lrange $lpath 0 $l] /d]/f$t # puts "npath=$npath" set comp [split $npath /] + if {[catch {exec rm [join $comp /]}]} return + set l [llength $comp] incr l -1 - if {[catch {exec rm [join $comp /]}]} return incr l -1 + incr status($task,$area) -1 for {set i $l} {$i > 0} {incr i -1} { set path [join [lrange $comp 0 $i] /] if {![catch {glob $path/*}]} return @@ -127,139 +142,318 @@ proc RobotFileClose {out} { } } -proc RobotFileOpen {area host path {mode w}} { +proc RobotFileOpen {task area host path {mode w}} { set orgPwd [pwd] global workdir + global status + global debuglevel if {![info exists workdir]} { return stdout } - #puts "RobotFileOpen orgPwd=$orgPwd area=$area host=$host path=$path mode=$mode" + if {$debuglevel > 3} { + puts "RobotFileOpen orgPwd=$orgPwd area=$area host=$host path=$path mode=$mode" + } if {[string compare $orgPwd $workdir]} { puts "ooops. RobotFileOpen failed" puts "workdir = $workdir" puts "pwd = $orgPwd" exit 1 } - set comp [split $area/$host$path /] + + set comp [split $task/$area/$host /] set len [llength $comp] incr len -1 - for {set i 0} {$i < $len} {incr i} { - if {$i > 1} { - set d "d[lindex $comp $i]" - } else { - set d [lindex $comp $i] - } - if {[catch {cd ./$d}]} { + + # puts "1 comp=$comp" + + for {set i 0} {$i <= $len} {incr i} { + set d [lindex $comp $i] + if {[catch {cd $d}]} { exec mkdir $d cd ./$d - if {![string compare $area unvisited] && $i == 1 && $mode == "w"} { - set out [open frobots.txt w] - puts "creating robots.txt in $d" - close $out + if {![string compare $area unvisited] && $i == $len && $mode == "w"} { + if {[string compare $path /robots.txt]} { + set out [open frobots.txt w] + puts "creating robots.txt in $d" + close $out + incr status($task,unvisited) + } } } } + + set comp [split $path /] + set len [llength $comp] + incr len -1 + + # puts "2 path=$path comp=$comp" + + for {set i 0} {$i < $len} {incr i} { + set d "d[lindex $comp $i]" + if {[string length $d] > 1} { + if {[catch {cd $d}]} { + exec mkdir $d + cd ./$d + } + } + } set d [lindex $comp $len] if {[string length $d]} { - if {[file isdirectory $d]} { - set out [open $d/f $mode] - } else { - set out [open f$d $mode] - } + set out [open f$d $mode] } else { set out [open f $mode] } + if {$mode == "w"} { + incr status($task,$area) + } cd $orgPwd return $out } -proc RobotRR {} { - global robotSeq robotsRunning +proc RobotStartJob {fname t} { + global control + + + set f [open $fname r] + set xml [read $f] + puts "Reading $fname" + regexp {([^<]*)} $xml x status + close $f + if {$status == "done"} { + puts "already done" + return + } + puts "status = $status" + if {![task $t]} { + return + } + htmlSwitch $xml \ + url { + url $body + } filter { + set type $parm(type) + set action $parm(action) + if {$type == "domain"} { + $action url http://$body/* + } + if {$type == "url"} { + $action url $body + } + if {$type == "mime"} { + $action mime $body + } + } distance { + set control($t,distance) $body + } status { + set control($t,filestatus) $body + } + if {$status == "pending"} { + regsub {[^<]*} $xml {running} xml2 + set f [open $fname w] + puts -nonewline $f $xml2 + close $f + } +} + +proc RobotDoneJob {t} { + global daemon_dir + + if {![info exists daemon_dir]} { + return + } + + set fname $t.tkl + + set f [open $fname r] + set xml [read $f] + puts "Reading $fname" + regexp {([^<]*)} $xml x status + puts "------" + puts "status = $status" + close $f + + regsub {[^<]*} $xml {done} xml2 + set f [open $fname w] + puts -nonewline $f $xml2 + close $f +} + +proc RobotScanDir {} { + global daemon_dir + + if {![info exists daemon_dir]} { + return + } + foreach d $daemon_dir { + if {[catch {set files [glob $d/*.tkl]}]} { + return + } + foreach fname $files { + if {[file isfile $fname] && [file readable $fname]} { + set t [file rootname $fname] + RobotStartJob $fname $t + } + } + } +} + +proc RobotRR {task} { + global robotSeq robotsRunning tasks robotsMax status + + puts "RobotRR -- running=$robotsRunning max=$robotsMax---------------" incr robotsRunning -1 + + # only one task gets through... + if {[string compare [lindex $tasks 0] $task]} { + return + } + puts "RobotRR. task = $task" while {$robotsRunning} { vwait robotsRunning } - set robotSeq 0 - RobotStart + puts "Scan" + if {[catch {RobotScanDir} msg]} { + puts "RobotScanDir failed" + puts $msg + } + foreach t $tasks { + set statusfile [open $t/status w] + puts $statusfile "$status($t,unvisited) $status($t,bad) $status($t,visited)" + close $statusfile + set robotSeq($t) 0 + RobotStart $t + } +} + +proc RobotDaemonSig {} { + global daemon_cnt + + incr daemon_cnt +} + +proc RobotDaemonLoop {} { + global daemon_cnt tasks robotsRunning status + + set daemon_cnt 0 + while 1 { + puts $daemon_cnt + + RobotScanDir + + if {[info exists tasks]} { + puts "daemon loop tasks $tasks" + foreach t $tasks { + set robotSeq($t) 0 + RobotStart $t + } + while {$robotsRunning} { + vwait robotsRunning + } + } + after 30000 RobotDaemonSig + vwait daemon_cnt + } } -proc RobotRestart {url sock} { +proc RobotRestart {task url sock} { global URL robotsRunning close $sock after cancel $URL($sock,cancel) - foreach v [array names URL $url,*] { + foreach v [array names URL $task,$url,*] { unset URL($v) } incr robotsRunning -1 - RobotStart + RobotStart $task } -proc RobotStart {} { +proc RobotStart {task} { global URL - global robotsRunning robotsMax idleTime + global robotsRunning robotsMax idletime status tasks - # puts "RobotStart" + # puts "RobotStart $task running=$robotsRunning" while {1} { - set url [RobotFileNext unvisited] + set url [RobotFileNext $task unvisited] + if {[string compare $url done] == 0} { + puts "In RobotStart task $task done" + + catch {unset ntasks} + foreach t $tasks { + if {[string compare $t $task]} { + lappend ntasks $t + } else { + puts "task $t done" + } + } + if {![info exists ntasks]} { + unset tasks + puts "all done" + } else { + set tasks $ntasks + } + RobotDoneJob $task + return + } if {![string length $url]} { return } - incr robotsRunning + incr robotsRunning if {[string compare $url wait] == 0} { - after $idleTime RobotRR - return + after $idletime [list RobotRR $task] + return } - set r [RobotGetUrl $url {}] + set r [RobotGetUrl $task $url {}] if {!$r} { if {$robotsRunning >= $robotsMax} return } else { incr robotsRunning -1 - if {![RobotFileExist bad $URL($url,hostport) $URL($url,path)]} { - set outf [RobotFileOpen bad $URL($url,hostport) $URL($url,path)] + if {![RobotFileExist $task bad $URL($task,$url,hostport) $URL($task,$url,path)]} { + set outf [RobotFileOpen $task bad $URL($task,$url,hostport) $URL($task,$url,path)] RobotFileClose $outf } - RobotFileUnlink unvisited $URL($url,hostport) $URL($url,path) + RobotFileUnlink $task unvisited $URL($task,$url,hostport) $URL($task,$url,path) } } } -proc headSave {url out} { +proc headSave {task url out} { global URL - if {[info exists URL($url,head,last-modified)]} { - puts $out "$URL($url,head,last-modified)" + if {[info exists URL($task,$url,head,last-modified)]} { + puts $out "$URL($task,$url,head,last-modified)" } puts $out {} - if {[info exists URL($url,head,date)]} { - puts $out " $URL($url,head,date)" + if {[info exists URL($task,$url,head,date)]} { + puts $out " $URL($task,$url,head,date)" } - if {[info exists URL($url,head,content-length)]} { - puts $out " $URL($url,head,content-length)" + if {[info exists URL($task,$url,head,content-length)]} { + puts $out " $URL($task,$url,head,content-length)" } - if {[info exists URL($url,head,server)]} { - puts $out " $URL($url,head,server)" + if {[info exists URL($task,$url,head,server)]} { + puts $out " $URL($task,$url,head,server)" } puts $out {} puts $out {} puts $out " $url" - if {[info exists URL($url,head,content-type)]} { - puts $out " $URL($url,head,content-type)" + if {[info exists URL($task,$url,head,content-type)]} { + puts $out " $URL($task,$url,head,content-type)" } puts $out {} } -proc RobotHref {url hrefx hostx pathx} { - global URL domains +proc RobotHref {task url hrefx hostx pathx} { + global URL control debuglevel upvar $hrefx href upvar $hostx host upvar $pathx path - puts "Ref url = $url href=$href" + if {$debuglevel > 1} { + puts "Ref input url = $url href=$href" + } if {[string first { } $href] >= 0} { return 0 @@ -267,9 +461,11 @@ proc RobotHref {url hrefx hostx pathx} { if {[string length $href] > 256} { return 0 } - if {[string first {?} $url] >= 0 && [string first {?} $href] >= 0} { - return 0 - } + +# Skip pages that have ? in them +# if {[string first {?} $url] >= 0 && [string first {?} $href] >= 0} { +# return 0 +# } # get method (if any) if {![regexp {^([^/:]+):(.*)} $href x method hpath]} { set hpath $href @@ -284,13 +480,13 @@ proc RobotHref {url hrefx hostx pathx} { if {![string length $surl]} { set surl / } - if {[info exist domains]} { + if {[info exist control($task,domains)]} { set ok 0 - foreach domain $domains { + foreach domain $control($task,domains) { if {[string match $domain $host]} { set ok 1 break - } + } } if {!$ok} { return 0 @@ -298,14 +494,18 @@ proc RobotHref {url hrefx hostx pathx} { } } else { regexp {^([^\#]*)} $hpath x surl - set host $URL($url,hostport) + set host $URL($task,$url,hostport) } if {![string length $surl]} { return 0 } if {[string first / $surl]} { # relative path - regexp {^([^\#?]*)} $URL($url,path) x dpart + set curpath $URL($task,$url,path) + if {[info exists URL($task,$url,bpath)]} { + set curpath $URL($task,$url,bpath) + } + regexp {^([^\#?]*)} $curpath x dpart set l [string last / $dpart] if {[expr $l >= 0]} { set surl [string range $dpart 0 $l]$surl @@ -319,9 +519,10 @@ proc RobotHref {url hrefx hostx pathx} { foreach c $surllist { switch -- $c { .. { - if {$pathl > 0} { - incr pathl -1 + if {$pathl > 1} { + incr pathl -2 set path [lrange $path 0 $pathl] + incr pathl } } . { @@ -333,63 +534,68 @@ proc RobotHref {url hrefx hostx pathx} { } } } - if {$pathl} { - set path [join $path /] - } else { - set path "" + if {$debuglevel > 4} { + puts "pathl=$pathl output path=$path" + } + set path [join $path /] + if {![string length $path]} { + set path / } regsub -all {~} $path {%7E} path set href "$method://$host$path" - puts "Ref href = $href" - return 1 + + if {$debuglevel > 1} { + puts "Ref result = $href" + } + return [checkrule $task url $href] } -proc RobotError {url code} { +proc RobotError {task url code} { global URL - puts "Bad URL $url, $code" + puts "Bad URL $url (code $code)" set fromurl {} set distance -1 - if {[RobotFileExist unvisited $URL($url,hostport) $URL($url,path)]} { - set inf [RobotFileOpen unvisited $URL($url,hostport) $URL($url,path) r] + if {[RobotFileExist $task unvisited $URL($task,$url,hostport) $URL($task,$url,path)]} { + set inf [RobotFileOpen $task unvisited $URL($task,$url,hostport) $URL($task,$url,path) r] RobotReadRecord $inf fromurl distance RobotFileClose $inf } - RobotFileUnlink unvisited $URL($url,hostport) $URL($url,path) - if {![RobotFileExist bad $URL($url,hostport) $URL($url,path)]} { - set outf [RobotFileOpen bad $URL($url,hostport) $URL($url,path)] + RobotFileUnlink $task unvisited $URL($task,$url,hostport) $URL($task,$url,path) + if {![RobotFileExist $task bad $URL($task,$url,hostport) $URL($task,$url,path)]} { + set outf [RobotFileOpen $task bad $URL($task,$url,hostport) $URL($task,$url,path)] RobotWriteRecord $outf $fromurl $distance RobotFileClose $outf } } -proc RobotRedirect {url tourl code} { +proc RobotRedirect {task url tourl code} { global URL puts "Redirecting from $url to $tourl" set distance {} set fromurl {} - if {[RobotFileExist unvisited $URL($url,hostport) $URL($url,path)]} { - set inf [RobotFileOpen unvisited $URL($url,hostport) $URL($url,path) r] + if {[RobotFileExist $task unvisited $URL($task,$url,hostport) $URL($task,$url,path)]} { + set inf [RobotFileOpen $task unvisited $URL($task,$url,hostport) $URL($task,$url,path) r] RobotReadRecord $inf fromurl distance RobotFileClose $inf } - if {![RobotFileExist bad $URL($url,hostport) $URL($url,path)]} { - set outf [RobotFileOpen bad $URL($url,hostport) $URL($url,path)] + if {![RobotFileExist $task bad $URL($task,$url,hostport) $URL($task,$url,path)]} { + set outf [RobotFileOpen $task bad $URL($task,$url,hostport) $URL($task,$url,path)] RobotWriteRecord $outf $fromurl $distance RobotFileClose $outf } - if {[RobotHref $url tourl host path]} { - if {![RobotFileExist visited $host $path]} { - if {![RobotFileExist unvisited $host $path]} { - set outf [RobotFileOpen unvisited $host $path] + if {[RobotHref $task $url tourl host path]} { + if {![RobotFileExist $task visited $host $path]} { + if {![RobotFileExist $task unvisited $host $path]} { + set outf [RobotFileOpen $task unvisited $host $path] RobotWriteRecord $outf $fromurl $distance RobotFileClose $outf } } else { set olddistance {} - set inf [RobotFileOpen visited $host $path r] + set inf [RobotFileOpen $task visited $host $path r] RobotReadRecord $inf oldurl olddistance RobotFileClose $inf if {[string length $olddistance] == 0} { @@ -400,157 +606,171 @@ proc RobotRedirect {url tourl code} { } puts "distance=$distance olddistance=$olddistance" if {[expr $distance < $olddistance]} { - set outf [RobotFileOpen unvisited $host $path] + set outf [RobotFileOpen $task unvisited $host $path] RobotWriteRecord $outf $tourl $distance RobotFileClose $outf } } } - if {[catch {RobotFileUnlink unvisited $URL($url,hostport) $URL($url,path)}]} { + if {[catch {RobotFileUnlink $task unvisited $URL($task,$url,hostport) $URL($task,$url,path)}]} { puts "unlink failed" exit 1 } } -proc RobotTextHtml {url out} { - global URL maxDistance +proc link {task url out href body distance} { + global URL control + if {[expr $distance > $control($task,distance)]} return + + if {![RobotHref $task $url href host path]} return + + puts $out "" + puts $out "$href" + puts $out "$body" + puts $out "" + + if {![RobotFileExist $task visited $host $path]} { + set olddistance 1000 + if {![RobotFileExist $task bad $host $path]} { + if {[RobotFileExist $task unvisited $host $path]} { + set inf [RobotFileOpen $task unvisited $host $path r] + RobotReadRecord $inf oldurl olddistance + RobotFileClose $inf + } + } else { + set olddistance 0 + } + if {[string length $olddistance] == 0} { + set olddistance 1000 + } + if {[expr $distance < $olddistance]} { + set outf [RobotFileOpen $task unvisited $host $path] + RobotWriteRecord $outf $url $distance + RobotFileClose $outf + } + } elseif {[string compare $href $url]} { + set inf [RobotFileOpen $task visited $host $path r] + RobotReadRecord $inf xurl olddistance + close $inf + if {[string length $olddistance] == 0} { + set olddistance 1000 + } + if {[expr $distance < $olddistance]} { + puts "OK remarking url=$url href=$href" + puts "olddistance = $olddistance" + puts "newdistance = $distance" + set outf [RobotFileOpen $task unvisited $host $path] + RobotWriteRecord $outf $url $distance + RobotFileClose $outf + } + } +} + +proc RobotTextHtml {task url out} { + global URL control + + # set title so we can emit it for the body + set title {} + # if true, nothing will be indexed + set noindex 0 + # if true, nothing will be followed + set nofollow 0 set distance 0 - if {$maxDistance < 1000 && [info exists URL($url,dist)]} { - set distance [expr $URL($url,dist) + 1] + set fdistance 0 + if {$control($task,distance) < 1000 && [info exists URL($task,$url,dist)]} { + set fdistance $URL($task,$url,dist) + set distance [expr $fdistance + 1] } - htmlSwitch $URL($url,buf) \ + htmlSwitch $URL($task,$url,buf) \ title { - puts $out "$body" + set title $body } -nonest meta { + # collect metadata and save NAME= CONTENT=.. + set metaname {} + set metacontent {} puts -nonewline $out "" + # go through robots directives (af any) + if {![string compare $metaname robots]} { + set direcs [split [string tolower $metacontent] ,] + if {[lsearch $direcs noindex] >= 0} { + set noindex 1 + } + if {[lsearch $direcs nofollow] >= 0} { + set nofollow 1 + } } - puts $out {>} } body { - regsub -all -nocase {))*} $body {} abody - regsub -all {<[^\>]+>} $abody {} nbody - puts $out "" - puts $out $nbody - puts $out "" - } -nonest a { + # don't print title of document content if noindex is used + if {!$noindex} { + puts $out "$title" + regsub -all {} $body { } abody + regsub -all -nocase {} $abody {} bbody + regsub -all {<[^\>]+>} $bbody {} nbody + puts $out "" + puts $out $nbody + puts $out "" + } + } -nonest base { + # if {![info exists parm(href)]} { - puts "no href" continue } - if {[expr $distance <= $maxDistance]} { - set href [string trim $parm(href)] - if {![RobotHref $url href host path]} continue - - puts $out "" - puts $out "$href" - puts $out "$body" - puts $out "" - - if {![RobotFileExist visited $host $path]} { - set olddistance 1000 - if {![RobotFileExist bad $host $path]} { - if {[RobotFileExist unvisited $host $path]} { - set inf [RobotFileOpen unvisited $host $path r] - RobotReadRecord $inf oldurl olddistance - RobotFileClose $inf - } - } else { - set olddistance 0 - } - if {[string length $olddistance] == 0} { - set olddistance 1000 - } - if {[expr $distance < $olddistance]} { - set outf [RobotFileOpen unvisited $host $path] - RobotWriteRecord $outf $url $distance - RobotFileClose $outf - } - } elseif {[string compare $href $url]} { - set inf [RobotFileOpen visited $host $path r] - RobotReadRecord $inf xurl olddistance - close $inf - if {[string length $olddistance] == 0} { - set olddistance 1000 - } - if {[expr $distance < $olddistance]} { - puts "OK remarking url=$url href=$href" - puts "olddistance = $olddistance" - puts "newdistance = $distance" - set outf [RobotFileOpen unvisited $host $path] - RobotWriteRecord $outf $url $distance - RobotFileClose $outf - } - } - } + set href [string trim $parm(href)] + if {![RobotHref $task $url href host path]} continue + set URL($task,$url,bpath) $path + } a { + # .. + # we're not using nonest - otherwise body isn't set + if {$nofollow} continue + if {![info exists parm(href)]} { + continue + } + link $task $url $out [string trim $parm(href)] $body $distance } -nonest area { + if {$nofollow} continue if {![info exists parm(href)]} { - puts "no href" continue } - if {[expr $distance <= $maxDistance]} { - set href [string trim $parm(href)] - if {![RobotHref $url href host path]} continue - - puts $out "" - puts $out "$href" - puts $out "" - puts $out "" - - if {![RobotFileExist visited $host $path]} { - set olddistance 1000 - if {![RobotFileExist bad $host $path]} { - if {[RobotFileExist unvisited $host $path]} { - set inf [RobotFileOpen unvisited $host $path r] - RobotReadRecord $inf oldurl olddistance - RobotFileClose $inf - } - } else { - set olddistance 0 - } - if {[string length $olddistance] == 0} { - set olddistance 1000 - } - if {[expr $distance < $olddistance]} { - set outf [RobotFileOpen unvisited $host $path] - RobotWriteRecord $outf $url $distance - RobotFileClose $outf - } - } elseif {[string compare $href $url]} { - set inf [RobotFileOpen visited $host $path r] - RobotReadRecord $inf xurl olddistance - close $inf - if {[string length $olddistance] == 0} { - set olddistance 1000 - } - if {[expr $distance < $olddistance]} { - puts "OK remarking url=$url href=$href" - puts "olddistance = $olddistance" - puts "newdistance = $distance" - set outf [RobotFileOpen unvisited $host $path] - RobotWriteRecord $outf $url $distance - RobotFileClose $outf - } - } - } + link $task $url $out [string trim $parm(href)] $body $distance + } -nonest frame { + if {![info exists parm(src)]} { + continue + } + link $task $url $out [string trim $parm(src)] $body $fdistance } } -proc RobotsTxt {url} { +proc RobotsTxt {task url} { global agent URL - RobotsTxt0 URL(URL($url,hostport),robots) $URL($url,buf) + RobotsTxt0 $task URL(URL($task,$url,hostport),robots) $URL($task,$url,buf) } -proc RobotsTxt0 {v buf} { +proc RobotsTxt0 {task v buf} { global URL agent set section 0 foreach l [split $buf \n] { - if {[regexp {([-A-Za-z]+):[ \t]*([^\#\t ]+)} $l match cmd arg]} { + if {[regexp {([-A-Za-z]+):[ ]*([^\# ]+)} $l match cmd arg]} { + set arg [string trim $arg] puts "cmd=$cmd arg=$arg" switch -- [string tolower $cmd] { user-agent { @@ -575,172 +795,189 @@ proc RobotsTxt0 {v buf} { } } -proc RobotTextPlain {url out} { +proc RobotTextPlain {task url out} { global URL puts $out "" - puts $out $URL($url,buf) + regsub -all {<} $URL($task,$url,buf) {\<} content + puts $out $content puts $out "" - if {![string compare $URL($url,path) /robots.txt]} { - RobotsTxt $url + if {![string compare $URL($task,$url,path) /robots.txt]} { + RobotsTxt $task $url } } -proc Robot200 {url} { - global URL domains - - set out [RobotFileOpen raw $URL($url,hostport) $URL($url,path)] - puts -nonewline $out $URL($url,buf) - RobotFileClose $out +proc RobotWriteMetadata {task url out} { + global URL - set out [RobotFileOpen visited $URL($url,hostport) $URL($url,path)] puts $out "" set distance 1000 - if {[RobotFileExist unvisited $URL($url,hostport) $URL($url,path)]} { - set inf [RobotFileOpen unvisited $URL($url,hostport) $URL($url,path) r] + if {[RobotFileExist $task unvisited $URL($task,$url,hostport) $URL($task,$url,path)]} { + set inf [RobotFileOpen $task unvisited $URL($task,$url,hostport) $URL($task,$url,path) r] RobotReadRecord $inf fromurl distance RobotFileClose $inf } - set URL($url,dist) $distance + set URL($task,$url,dist) $distance puts $out "" puts $out " $distance" puts $out "" - headSave $url $out + headSave $task $url $out puts "Parsing $url distance=$distance" - switch $URL($url,head,content-type) { - text/html { - if {[string length $distance]} { - RobotTextHtml $url $out - } - } - text/plain { - RobotTextPlain $url $out $outr - } - application/pdf { - set pdff [open test.pdf w] - puts -nonewline $pdff $URL($url,buf) - close $pdff - } + switch $URL($task,$url,head,content-type) { + text/html { + if {[string length $distance]} { + RobotTextHtml $task $url $out + } + } + text/plain { + RobotTextPlain $task $url $out + } } puts $out "" +} + +proc Robot200 {task url} { + global URL + + set out [RobotFileOpen $task raw $URL($task,$url,hostport) $URL($task,$url,path)] + puts -nonewline $out $URL($task,$url,buf) + RobotFileClose $out + + set out [RobotFileOpen $task visited $URL($task,$url,hostport) $URL($task,$url,path)] + RobotWriteMetadata $task $url $out RobotFileClose $out - # puts "Parsing done" - RobotFileUnlink unvisited $URL($url,hostport) $URL($url,path) + + RobotFileUnlink $task unvisited $URL($task,$url,hostport) $URL($task,$url,path) } -proc RobotReadContent {url sock binary} { +proc RobotReadContent {task url sock binary} { global URL - puts "RobotReadContent $url" set buffer [read $sock 16384] set readCount [string length $buffer] if {$readCount <= 0} { - Robot200 $url - RobotRestart $url $sock + Robot200 $task $url + RobotRestart $task $url $sock } elseif {!$binary && [string first \0 $buffer] >= 0} { - Robot200 $url - RobotRestart $url $sock + Robot200 $task $url + RobotRestart $task $url $sock } else { # puts "Got $readCount bytes" - set URL($url,buf) $URL($url,buf)$buffer + set URL($task,$url,buf) $URL($task,$url,buf)$buffer } } -proc RobotReadHeader {url sock} { - global URL +proc RobotReadHeader {task url sock} { + global URL debuglevel - puts "RobotReadHeader $url" + if {$debuglevel > 1} { + puts "HTTP head $url" + } if {[catch {set buffer [read $sock 2148]}]} { - RobotError $url 404 - RobotRestart $url $sock + RobotError $task $url 404 + RobotRestart $task $url $sock + return } set readCount [string length $buffer] if {$readCount <= 0} { - RobotError $url 404 - RobotRestart $url $sock + RobotError $task $url 404 + RobotRestart $task $url $sock } else { # puts "Got $readCount bytes" - set URL($url,buf) $URL($url,buf)$buffer + set URL($task,$url,buf) $URL($task,$url,buf)$buffer - set n [string first \r\n\r\n $URL($url,buf)] + set n [string first \r\n\r\n $URL($task,$url,buf)] if {$n > 1} { set code 0 set version {} - set headbuf [string range $URL($url,buf) 0 $n] + set headbuf [string range $URL($task,$url,buf) 0 $n] incr n 4 - set URL($url,buf) [string range $URL($url,buf) $n end] + set URL($task,$url,buf) [string range $URL($task,$url,buf) $n end] regexp {^HTTP/([0-9.]+)[ ]+([0-9]+)} $headbuf x version code set lines [split $headbuf \n] foreach line $lines { if {[regexp {^([^:]+):[ ]+([^;]*)} $line x name value]} { - set URL($url,head,[string tolower $name]) [string trim $value] + set URL($task,$url,head,[string tolower $name]) [string trim $value] } } - puts "code = $code" - set URL($url,state) skip + puts "HTTP CODE $code" + set URL($task,$url,state) skip switch $code { 301 { - RobotRedirect $url $URL($url,head,location) 301 - RobotRestart $url $sock + RobotRedirect $task $url $URL($task,$url,head,location) 301 + RobotRestart $task $url $sock } 302 { - RobotRedirect $url $URL($url,head,location) 302 - RobotRestart $url $sock + RobotRedirect $task $url $URL($task,$url,head,location) 302 + RobotRestart $task $url $sock } 200 { - if {![info exists URL($url,head,content-type)]} { - set URL($url,head,content-type) {} + if {![info exists URL($task,$url,head,content-type)]} { + set URL($task,$url,head,content-type) {} } - set binary 0 - switch $URL($url,head,content-type) { - application/pdf { - set binary 1 + set binary 1 + switch -glob -- $URL($task,$url,head,content-type) { + text/* { + set binary 0 } } - fileevent $sock readable [list RobotReadContent $url $sock $binary] + if {![regexp {/robots.txt$} $url]} { + if {![checkrule $task mime $URL($task,$url,head,content-type)]} { + RobotError $task $url mimedeny + RobotRestart $task $url $sock + return + } + } + fileevent $sock readable [list RobotReadContent $task $url $sock $binary] } default { - RobotError $url $code - RobotRestart $url $sock + RobotError $task $url $code + RobotRestart $task $url $sock } } } } } -proc RobotSockCancel {url sock} { +proc RobotSockCancel {task url sock} { puts "RobotSockCancel sock=$sock url=$url" - RobotError $url 401 - RobotRestart $url $sock + RobotError $task $url 401 + RobotRestart $task $url $sock } -proc RobotConnect {url sock} { - global URL agent +proc RobotConnect {task url sock} { + global URL agent acceptLanguage fconfigure $sock -translation {lf crlf} -blocking 0 - fileevent $sock readable [list RobotReadHeader $url $sock] - puts $sock "GET $URL($url,path) HTTP/1.0" - puts $sock "Host: $URL($url,host)" + fileevent $sock readable [list RobotReadHeader $task $url $sock] + puts $sock "GET $URL($task,$url,path) HTTP/1.0" + puts $sock "Host: $URL($task,$url,host)" puts $sock "User-Agent: $agent" + if {[string length $acceptLanguage]} { + puts $sock "Accept-Language: $acceptLanguage" + } puts $sock "" - flush $sock - set URL($sock,cancel) [after 30000 [list RobotSockCancel $url $sock]] + set URL($sock,cancel) [after 30000 [list RobotSockCancel $task $url $sock]] + if {[catch {flush $sock}]} { + RobotError $task $url 404 + RobotRestart $task $url $sock + } } proc RobotNop {} { } -proc RobotGetUrl {url phost} { +proc RobotGetUrl {task url phost} { global URL robotsRunning flush stdout - puts "RobotGetUrl --------- robotsRunning=$robotsRunning url=$url" + puts "Retrieve running=$robotsRunning url=$url task=$task" if {![regexp {([^:]+)://([^/]+)(.*)} $url x method hostport path]} { return -1 } @@ -748,25 +985,25 @@ proc RobotGetUrl {url phost} { set port 80 set host $hostport } - set URL($url,method) $method - set URL($url,host) $host - set URL($url,hostport) $hostport - set URL($url,path) $path - set URL($url,state) head - set URL($url,buf) {} + set URL($task,$url,method) $method + set URL($task,$url,host) $host + set URL($task,$url,hostport) $hostport + set URL($task,$url,path) $path + set URL($task,$url,state) head + set URL($task,$url,buf) {} if {[string compare $path /robots.txt]} { set ok 1 if {![info exists URL($hostport,robots)]} { puts "READING robots.txt for host $hostport" - if {[RobotFileExist visited $hostport /robots.txt]} { - set inf [RobotFileOpen visited $hostport /robots.txt r] + if {[RobotFileExist $task visited $hostport /robots.txt]} { + set inf [RobotFileOpen $task visited $hostport /robots.txt r] set buf [read $inf 32768] close $inf } else { set buf "User-agent: *\nAllow: /\n" } - RobotsTxt0 URL($hostport,robots) $buf + RobotsTxt0 $task URL($hostport,robots) $buf } if {[info exists URL($hostport,robots)]} { foreach l $URL($hostport,robots) { @@ -784,7 +1021,7 @@ proc RobotGetUrl {url phost} { if [catch {set sock [socket -async $host $port]}] { return -1 } - RobotConnect $url $sock + RobotConnect $task $url $sock return 0 } @@ -796,7 +1033,7 @@ if {![llength [info commands htmlSwitch]]} { } } -set agent "zmbot/0.0" +set agent "zmbot/0.2" if {![catch {set os [exec uname -s -r]}]} { set agent "$agent ($os)" } @@ -810,33 +1047,159 @@ proc bgerror {m} { } set robotsRunning 0 -set robotSeq 0 set workdir [pwd] -set idleTime 60000 +set idletime 30000 +set acceptLanguage {} +set debuglevel 0 + +# Rules: allow, deny, url + +proc checkrule {task type this} { + global control + global debuglevel + + if {$debuglevel > 3} { + puts "CHECKRULE $type $this" + } + if {[info exist control($task,alrules)]} { + foreach l $control($task,alrules) { + if {$debuglevel > 3} { + puts "consider $l" + } + # consider type + if {[lindex $l 1] != $type} continue + # consider mask (! negates) + set masks [lindex $l 2] + set ok 0 + foreach mask $masks { + if {$debuglevel > 4} { + puts "consider single mask $mask" + } + if {[string index $mask 0] == "!"} { + set mask [string range $mask 1 end] + if {[string match $mask $this]} continue + } else { + if {![string match $mask $this]} continue + } + set ok 1 + } + if {$debuglevel > 4} { + puts "ok = $ok" + } + if {!$ok} continue + # OK, we have a match + if {[lindex $l 0] == "allow"} { + if {$debuglevel > 3} { + puts "CHECKRULE MATCH OK" + } + return 1 + } else { + if {$debuglevel > 3} { + puts "CHECKFULE MATCH FAIL" + } + return 0 + } + } + } + if {$debuglevel > 3} { + puts "CHECKRULE MATCH OK" + } + return 1 +} -set i 0 -set l [llength $argv] -# For testing only -if {0} { - set url "http://www.sportsfiskeren.dk/sportsfiskeren/corner/index.htm" - set href "../../data/../../data2/newsovs.asp?Mode=5" +proc url {href} { + global debuglevel task - set URL($url,path) /sportsfiskeren/corner/index.htm - set URL($url,hostport) www.sportsfiskeren.dk - RobotHref $url href host path - exit 0 + if {[RobotHref $task http://www.indexdata.dk/ href host path]} { + if {![RobotFileExist $task visited $host $path]} { + set outf [RobotFileOpen $task unvisited $host $path] + RobotWriteRecord $outf href 0 + RobotFileClose $outf + } + } +} + +proc deny {type stuff} { + global control task + + lappend control($task,alrules) [list deny $type $stuff] +} + +proc allow {type stuff} { + global control task + + lappend control($task,alrules) [list allow $type $stuff] +} + +proc debug {level} { + global debuglevel + + set debuglevel $level +} + +proc task {t} { + global tasks task status robotSeq control + + set task $t + + if {[info exists tasks]} { + if {[lsearch -exact $tasks $t] >= 0} { + return 0 + } + } + + lappend tasks $t + set status($t,unvisited) 0 + set status($t,visited) 0 + set status($t,bad) 0 + set status($t,raw) 0 + set status($t,active) 1 + set robotSeq($t) 0 + set control($t,distance) 10 + return 1 } +# Little utility that ensures that at least one task is present (main). +proc chktask {} { + global tasks + if {![info exist tasks]} { + task main + } +} + + +# Parse options + +set i 0 +set l [llength $argv] + if {$l < 2} { - puts {tclrobot: usage [-j jobs] [-i idle] [-c count] [-d domain] [url ..]} + puts {tclrobot: usage:} + puts {tclrobot [-j jobs] [-i idle] [-c count] [-d domain] [-D dir] [-r rules] [url ..]} puts " Example: -c 3 -d '*.dk' http://www.indexdata.dk/" + exit 1 } + while {$i < $l} { set arg [lindex $argv $i] switch -glob -- $arg { + -t* { + set t [string range $arg 2 end] + if {![string length $t]} { + set t [lindex $argv [incr i]] + } + task $t + } + -D* { + set dir [string range $arg 2 end] + if {![string length $dir]} { + set dir [lindex $argv [incr i]] + } + lappend daemon_dir $dir + } -j* { set robotsMax [string range $arg 2 end] if {![string length $robotsMax]} { @@ -844,29 +1207,51 @@ while {$i < $l} { } } -c* { - set maxDistance [string range $arg 2 end] - if {![string length $maxDistance]} { - set maxDistance [lindex $argv [incr i]] + chktask + set control($task,distance) [string range $arg 2 end] + if {![string length $control($task,distance)]} { + set control($task,distance) [lindex $argv [incr i]] } } -d* { + chktask set dom [string range $arg 2 end] if {![string length $dom]} { set dom [lindex $argv [incr i]] } - lappend domains $dom + lappend control($task,domains) $dom } -i* { - set idleTime [string range $arg 2 end] - if {![string length $idleTime]} { - set idleTime [lindex $argv [incr i]] + set idletime [string range $arg 2 end] + if {![string length $idletime]} { + set idletime [lindex $argv [incr i]] } } + -l* { + chktask + set acceptLanguage [string range $arg 2 end] + if {![string length $acceptLanguage]} { + set acceptLanguage [lindex $argv [incr i]] + } + } + -r* { + chktask + set rfile [string range $arg 2 end] + if {![string length $rfile]} { + set rfile [lindex $argv [incr i]] + } + catch {unset maxdistance} + source $rfile + if {[info exists maxdistance]} { + set control($task,distance) $maxdistance + } + } default { + chktask set href $arg - if {[RobotHref http://www.indexdata.dk/ href host path]} { - if {![RobotFileExist visited $host $path]} { - set outf [RobotFileOpen unvisited $host $path] + if {[RobotHref $task http://www.indexdata.dk/ href host path]} { + if {![RobotFileExist $task visited $host $path]} { + set outf [RobotFileOpen $task unvisited $host $path] RobotWriteRecord $outf href 0 RobotFileClose $outf } @@ -876,22 +1261,36 @@ while {$i < $l} { incr i } -if {![info exist domains]} { - set domains {*} -} -if {![info exist maxDistance]} { - set maxDistance 50 -} if {![info exist robotsMax]} { set robotsMax 5 } -puts "domains=$domains" -puts "max distance=$maxDistance" -puts "max jobs=$robotsMax" - -RobotStart - -while {$robotsRunning} { - vwait robotsRunning +if {[info exist daemon_dir]} { + RobotDaemonLoop +} else { + foreach t $tasks { + puts "task $t" + puts "max distance=$control($t,distance)" + if {[info exists control($t,domains)]} { + puts "domains=$control($t,domains)" + } + } + puts "max jobs=$robotsMax" + + foreach t $tasks { + RobotStart $t + } + + while {$robotsRunning} { + vwait robotsRunning + } + + if {[info exists tasks]} { + foreach t $tasks { + set statusfile [open $t/status w] + puts $statusfile "$status($t,unvisited) $status($t,bad) $status($t,visited)" + close $statusfile + } + } } +