Now, by default, the script adds a link to the original post only if its text doesn’t fit into the post to be posted; added a “always_link_gancio_post” to make it do it always; added a “test mode”; fixed some flaws; updated help text

gancioff

@@ -105,22 +105,48 @@ post_visibility = unlisted
 # «fedi_hostname»; it can be used for testing purposes or just to keep the
 # posts shorter than they would be otherwise; for example:
 max_post_length = 840
+
+# «always_link_gancio_post» is optional and if unspecified it defaults to
+# «false», which means that {$SNAME} adds to the Mastodon post a link to the
+# original Gancio post only if the latter is too long to fit into the first
+# (i.e. into the Mastodon instance “max post length”, or into the
+# «max_post_length» specified in this configuration file - see above); this
+# way, {$SNAME} reduces the burden on the Gancio instance that is due to the
+# requests that it gets from every Mastodon instance where a Mastodon post
+# with a link to the original Gancio post will end up, in order for each of
+# them to generate a “link preview”; such burden gets reduced in different
+# measures depending on the average length of a post on the Gancio instance
+# and on the “max post length” on the Mastodon instance that {$SNAME} is using
+# to post (or on the «max_post_length» explicitly specified in this file).
+# If set to «true», {$SNAME} will instead always add a link to the original
+# Gancio post.
+always_link_gancio_post = true
 --- End of example configuration file ---
 
 [[[ OPTIONS ]]]
 
 -h / --help
   Show this help text and exit.
+-t / --test
+  Do a test: {$SNAME} will try as always to read the configuration file, fetch
+ the defined Mastodon instance’s info, load the state file and fetch the feed,
+ but it will post only the first event it may find there, with a visibility of
+ «direct», even if according to the state file it has already been posted, and
+ won’t update the state file.
 -p / --do-post <y|n>
- Setting this option to «n» («no») will make {$SNAME} skip posting. Note that
+  When a state file already exists, this option defaults to «y» («yes»), which
- even in this case it will save into the state file the GUIDs of new events
+ means that {$SNAME} will try to post all the new events it may find in the
- it may find in the feed, so it won’t post them even on subsequent runs.
+ feed; if set to «n» («no»), {$SNAME} will not try to post them, but it will
- Also note that, in order to prevent you from flooding your instance’s local
+ save their GUIDs into the state file nonetheless, so they won’t be posted
- timeline on {$SNAME} first run on a given feed (i.e. when the state file
+ again on subsequent runs.
- specified in the configuration file doesn’t exist yet), if events in the feed
+  This is mainly useful on {$SNAME}’s first run on a given feed, i.e. when
- are more than 5 and «posts_visibility» is set to «public» {$SNAME} refuses to
+ the state file specified in the configuration file doesn’t exist yet and thus
- post them unless you explicitly set this option to «n» («no») or «y» («yes»).
+ all the events in the feed will be considered “new”: in this case, {$SNAME}
- When the state file exists, this option defaults to «y» («yes»).
+ refuses to run unless you explicitly set this option to «y» or «n»: this is a
+ way to prevent you from unintentionally flooding your Mastodon instance with
+ all the events in the feed.
+  When “test mode” is active (see the previous option description), setting
+ this option has no effect.
 -v / --verbose
  Show some more messages about what the script is doing.
 --
@@ -148,10 +174,12 @@ $conf=[
 	'state_file_absolute_path'=>['required'=>true, 'default'=>null],
 	'posts_language'=>['required'=>true, 'default'=>null],
 	'posts_visibility'=>['required'=>false, 'default'=>'public'],
-	'max_post_length'=>['required'=>false, 'default'=>null]
+	'max_post_length'=>['required'=>false, 'default'=>null],
+	'always_link_gancio_post'=>['required'=>false, 'default'=>false]
 ];
 
 $opts=[
+	'test'=>false,
 	'do-post'=>null,
 	'verbose'=>false,
 	'update-language-codes'=>false
@@ -165,6 +193,8 @@ for ($i=1; $i<$argc; $i++) {
 		} elseif ($argv[$i]=='-h' || $argv[$i]=='--help') {
 			echo $help;
 			exit(0);
+		} elseif ($argv[$i]=='-t' || $argv[$i]=='--test') {
+			$opts['test']=true;
 		} elseif ($argv[$i]=='-p' || $argv[$i]=='--do-post') {
 			if ($i+1>=$argc) dieYoung("Error: option «{$argv[$i]}» requires an argument; use «-h» or «--help» to display help.\n",1);
 			if ($argv[$i+1]=='y')
@@ -231,6 +261,12 @@ if (!is_null($conf['max_post_length'])) {
 	$conf['max_post_length']+=0;
 	echo "Info: got «{$conf['max_post_length']}» as «max_post_length» from configuration file.\n";
 }
+if (!is_null($conf['always_link_gancio_post']) && preg_match('#^(true|false)$#',$conf['always_link_gancio_post'])!==1) {
+	dieYoung("Error: configuration file: «always_link_gancio_post» must be «true» or «false».\n",1);
+} else {
+	($conf['always_link_gancio_post']=='true') ? $conf['always_link_gancio_post']=true : $conf['always_link_gancio_post']=false;
+}
+if ($opts['test']) $conf['posts_visibility']='direct';
 echo "Info: got good configuration from configuration file.\n";
 
 $tldsregex=gettlds(__DIR__.'/storage/tlds.txt',true);
@@ -255,22 +291,8 @@ if (is_null($conf['max_post_length'])) {
 }
 //print_r($conf);
 
-echo "Info: trying to fetch feed from «{$conf['feed_url']}».\n";
-$feed=curl($conf['feed_url'],null,['Accept: application/xml']);
-if ($feed['content']===false) dieYoung("Error: could not connect to «{$conf['feed_url']}» (error: «{$feed['error']}»).\n",1);
-if ($feed['httpcode']!='200') dieYoung("Error: «{$conf['feed_url']} returned http code «{$res['httpcode']}».\n",1);
-$feed=@simplexml_load_string($feed['content'],null,LIBXML_NOCDATA);
-if ($feed===false) dieYoung("Error: got no valid XML from «{$conf['feed_url']}».\n",1);
-//print_r($feed);
-if (!isset($feed->channel->item) || !is_iterable($feed->channel->item) || !is_countable($feed->channel->item)) dieYoung("Error: feed from «{$conf['feed_url']}» had unexpected format.\n",1);
-$itemsCount=$feed->channel->item->count();
-if ($itemsCount==0) exitYoung("Info: feed from «{$conf['feed_url']}» was empty, bye.\n");
-echo "Info: got good feed from «{$conf['feed_url']}».\n";
-
 $guids=[];
 echo "Info: trying to load GUIDs of already posted events from state file «{$conf['state_file_absolute_path']}».\n";
-if (!file_exists($conf['state_file_absolute_path']) && is_null($opts['do-post']) && $conf['posts_visibility']=='public' && $itemsCount>5) dieYoung("Error: state file «{$conf['state_file_absolute_path']}» doesn’t exist yet, so this is probably a first run on feed «{$conf['feed_url']}», and since statuses visibility is set to «public» and the feed contains more than 5 events you have to decide whether it would be ok to post {$itemsCount} statuses that would show up, probably all in a row, on «{$conf['fedi_hostname']}»’s local timeline, and explicitly declare whether you want {$SNAME} to post them or not by setting option «-p» or «--do-post» to «y» («yes») or «n» («no»); use «-h» or «--help» to display help.\n",1);
-if (is_null($opts['do-post'])) $opts['do-post']=true;
 if (file_exists($conf['state_file_absolute_path'])) {
 	if (!is_file($conf['state_file_absolute_path'])) dieYoung("Error: «{$conf['state_file_absolute_path']}» exists but it’s not a file.\n",1);
 	if (!is_readable($conf['state_file_absolute_path'])) dieYoung("Error: «{$conf['state_file_absolute_path']}» exists but it’s not readable.\n",1);
@@ -295,12 +317,31 @@ if (file_exists($conf['state_file_absolute_path'])) {
 	}
 	fclose($fh);
 	echo 'Info: got '.count($guids)." GUID(s) for already posted event(s) from state file «{$conf['state_file_absolute_path']}»; removed {$i} line(s) older than one year.\n";
+} elseif (is_null($opts['do-post']) && !$opts['test']) {
+	dieYoung("Error: state file «{$conf['state_file_absolute_path']}» doesn’t exist yet, so this is probably a first run on feed «{$conf['feed_url']}»; thus, all the events {$SNAME} may find in the feed will be considered new and, as a precaution against flooding your local timeline, you have to explicitly declare whether you want it to post them all, or not, by explicitly setting option «-p» or «--do-post» to «y» («yes») or «n» («no»); mind that in both cases they will be recorded as posted in the state file, and won’t be posted again on subsequent runs (you can use «-h» or «--help» to display help).\n",1);
 } else {
 	echo "Info: state file «{$conf['state_file_absolute_path']}» was not found.\n";
 }
+if (is_null($opts['do-post']) || $opts['test']) $opts['do-post']=true;
 
-if (($fh=@fopen($conf['state_file_absolute_path'],'a'))===false) dieYoung("Error: could not open «{$conf['state_file_absolute_path']}» in «append» mode.\n",1);
 $newItemsCount=0;
+echo "Info: trying to fetch feed from «{$conf['feed_url']}».\n";
+$feed=curl($conf['feed_url'],null,['Accept: application/xml']);
+if ($feed['content']===false) dieYoung("Error: could not connect to «{$conf['feed_url']}» (error: «{$feed['error']}»).\n",1);
+if ($feed['httpcode']!='200') dieYoung("Error: «{$conf['feed_url']} returned http code «{$res['httpcode']}».\n",1);
+$feed=@simplexml_load_string($feed['content'],null,LIBXML_NOCDATA);
+if ($feed===false) dieYoung("Error: got no valid XML from «{$conf['feed_url']}».\n",1);
+//print_r($feed);
+if (!isset($feed->channel->item) || !is_iterable($feed->channel->item) || !is_countable($feed->channel->item)) dieYoung("Error: feed from «{$conf['feed_url']}» had unexpected format.\n",1);
+$itemsCount=$feed->channel->item->count();
+if ($itemsCount==0) exitYoung("Info: feed from «{$conf['feed_url']}» was empty, bye.\n");
+foreach ($feed->channel->item as $item)
+	if (isset($item->guid) && $item->guid->__toString()!='' && !array_key_exists($item->guid->__toString(),$guids))
+		$newItemsCount++;
+echo "Info: got good feed with {$itemsCount} events ({$newItemsCount} new) from «{$conf['feed_url']}».\n";
+if (!$opts['test'] && $newItemsCount==0) exitYoung("Info: feed from «{$conf['feed_url']}» had no new events, bye.\n");
+
+if (!$opts['test'] && ($fh=@fopen($conf['state_file_absolute_path'],'a'))===false) dieYoung("Error: could not open «{$conf['state_file_absolute_path']}» in «append» mode.\n",1);
 $goodPostsCount=0;
 $index=0;
 foreach ($feed->channel->item as $item) {
@@ -308,8 +349,7 @@ foreach ($feed->channel->item as $item) {
 //	print_r($item);
 	if (!isset($item->guid) || ($guid=$item->guid->__toString())=='') {
 		fwrite(STDERR,"Warning: event #{$index} has no GUID, skipping.\n");
-	} elseif (!array_key_exists($guid,$guids)) {
+	} elseif (!array_key_exists($guid,$guids) || $opts['test']) {
-		$newItemsCount++;
 		$file=null;
 		if (isset($item->enclosure[0]['url']) && isset($item->enclosure[0]['type']) && isset($item->enclosure[0]['length']))
 			$file=['url'=>$item->enclosure[0]['url']->__toString(), 'type'=>$item->enclosure[0]['type']->__toString(), 'length'=>$item->enclosure[0]['length']->__toString()];
@@ -339,12 +379,17 @@ foreach ($feed->channel->item as $item) {
 			$pcats='';
 		}
 		$post="{$ptext}{$plink}{$pcats}";
+		if (postLength($post,$tldsregex['tlds'])<=$conf['max_post_length'] && !$conf['always_link_gancio_post']) {
+			$plink='';
+		} else {
+			$pcats='';
+		}
+		$post="{$ptext}{$plink}{$pcats}";
 		while (postLength($post,$tldsregex['tlds'])>$conf['max_post_length'] && $ptext!='') {
 			$ptext=preg_replace('#\S+\W*$#','',$ptext);
 //			echo "[[[{$ptext}]]]\n";
 			$post="{$ptext}[…]{$plink}{$pcats}";
 		}
-		if (postLength($post,$tldsregex['tlds'])>$conf['max_post_length']) $post=$plink;
 		if (postLength($post,$tldsregex['tlds'])>$conf['max_post_length']) {
 			fwrite(STDERR,"Warning: could not shorten post for event «{$guid}» to make it fit into {$conf['max_post_length']} characters; won’t post.\n");
 		} else {
@@ -428,7 +473,7 @@ foreach ($feed->channel->item as $item) {
 //						print_r($res['content']);
 						$now=time();
 						$guids[$guid]=$now;
-						fwrite($fh,"{$now}\t{$guid}\n");
+						if (!$opts['test']) fwrite($fh,"{$now}\t{$guid}\n");
 						$goodPostsCount++;
 					}
 				}
@@ -436,20 +481,26 @@ foreach ($feed->channel->item as $item) {
 				echo "Info: would have posted status for event «{$guid}».\n";
 				$now=time();
 				$guids[$guid]=$now;
-				fwrite($fh,"{$now}\t{$guid}\n");
+				if (!$opts['test']) fwrite($fh,"{$now}\t{$guid}\n");
 				$goodPostsCount++;
 			}
 		}
 	} else {
 		if ($opts['verbose']) echo "Info: event «{$guid}» has already been posted on ".date('c',$guids[$guid]).", skipping.\n";
 	}
-//	fclose($fh); exit(0);// to test a single post
+	if ($opts['test']) break;// to test a single post
+}
+if (!$opts['test']) fclose($fh);
+if (!$opts['test']) {
+	if ($opts['do-post'])
+		echo "Info: succesfully posted {$goodPostsCount} of {$newItemsCount} new event(s) (of {$itemsCount} total events in the feed).\n";
+	else
+		echo "Info: would have tried to post {$newItemsCount} new event(s) of {$itemsCount} total events in the feed.\n";
+} elseif ($goodPostsCount==1) {
+	echo "Info: successfully posted the first of {$itemsCount} total events in the feed ({$newItemsCount} are new).\n";
+} else {
+	echo "Info: failed to post the first of {$itemsCount} total events in the feed ({$newItemsCount} are new).\n";
 }
-fclose($fh);
-if ($opts['do-post'])
-	echo "Info: feed got {$itemsCount} events; succesfully posted {$goodPostsCount} of {$newItemsCount} new event(s).\n";
-else
-	echo "Info: feed got {$itemsCount} event(s), {$newItemsCount} new.\n";
 
 exit(0);