From: Ian Jackson <ijackson@chiark.greenend.org.uk>
Date: Tue, 1 Sep 2009 15:31:17 +0000 (+0100)
Subject: Cope with some features of the Cobalt WP chart
X-Git-Tag: 3.4~39
X-Git-Url: http://www.chiark.greenend.org.uk/ucgi/~yarrgweb/git?p=ypp-sc-tools.web-live.git;a=commitdiff_plain;h=243aacd694064310880840502e5fd4dba729dd0a;ds=sidebyside

Cope with some features of the Cobalt WP chart
---

diff --git a/yarrg/yppedia-chart-parser b/yarrg/yppedia-chart-parser
index a093a0c..b35d816 100755
--- a/yarrg/yppedia-chart-parser
+++ b/yarrg/yppedia-chart-parser
@@ -118,7 +118,7 @@ sub yppedia_chart_parse () {
 	s/\<--.*--\>//g;
 	s/^\s*//; chomp; s/\s+$//; s/\s+/ /g;
 	s/\<\/?(?:b|em)\>//g;
-	s/\{\{Chart\ style\|[^{}]*\}\}//g;
+	s/\{\{chart\ style\|[^{}]*\}\}//gi;
 	next unless m/\{\{/; # only interested in chart template stuff
 
 	my ($x,$y, $arch,$island,$solid,$dirn);
@@ -126,9 +126,9 @@ sub yppedia_chart_parse () {
     
 	if (($x,$y,$arch) =
 	    m/^\{\{ chart\ label \|(\d+)\|(\d+)\| .*
-	            (?: \<big\>)? \'+
+	            (?: \<(?: big|center )\>)* \'+
 	            \[\[ [^][\']* \| ([^][\'|]+)\ archipelago \]\]
-	            \'+ (?: \<\/big\>)? \}\}$/xi) {
+	            \'+ (?: \<\/(?: big|center )\>)* \}\}$/xi) {
 	    printf DEBUG "%2d,%-2d arch %s\n", $x,$y,$arch;
 	    push @wiarchlabels, [ $x,$y,$arch ];
 	} elsif (m/^\{\{ chart\ label \|\d+\|\d+\|
diff --git a/yarrg/yppedia-ocean-scraper b/yarrg/yppedia-ocean-scraper
index 30d0c4a..ad35c4d 100755
--- a/yarrg/yppedia-ocean-scraper
+++ b/yarrg/yppedia-ocean-scraper
@@ -89,6 +89,7 @@ def parse_chart():
 	debug('s',s)
 	s = regexp.sub(r'\&lt\;', '<', s)
 	s = regexp.sub(r'\&gt\;', '>', s)
+	s = regexp.sub(r'\&quot\;', '"', s)
 	s = regexp.sub(r'\&amp\;', '&', s)
 	debug('s',s)
 	return s