using xpath to scrape web

Makefile

@@ -1,12 +1,22 @@
 HAML=haml
+XMLLINT=xmllint
 
 all: index.html doprava.html
 
-refresh:
-	curl 'http://vvv.chmi.cz/hydro/detail_stanice/307225.html' | sed -n '153,302 p' > data/table.html
+get:
+	curl 'http://vvv.chmi.cz/hydro/detail_stanice/307225.html' > data/307225.html
 	wget 'http://vvv.chmi.cz/hydro/graph/big/307225_H.png' -O data/stav.png
 	wget 'http://vvv.chmi.cz/hydro/graph/big/307225_Q.png' -O data/prutok.png
 
+scrap: get
+	xmllint --html --encode utf8 data/307225.html --output data/307225.html
+	xmllint --html --xpath '//div[@class="box"]/div[@class="cont"]/p[2]/text()' data/307225.html > data/timestamp.html
+	xmllint --html --xpath '//div[@class="box"]/div[@class="cont"]/table[2]//table[5]/tr[position()>1]' data/307225.html > data/table.html
+#	xmllint --html --xpath '//div[@class="box"]/div[@class="cont"]/table[2]//table[3]//tr' data/307225.html | iconv -f cp1250 -t utf8 > data/legend.html
+#	xmllint --html --xpath '//div[@class="box"]/div[@class="cont"]/table[2]//table[1]//tr' data/307225.html | iconv -f cp1250 -t utf8 > data/info.html
+
+refresh: get scrap
+
 %.html: %.haml
 	$(HAML) $< $@
 

index.haml

@@ -13,7 +13,8 @@
       %li <a href='doprava.html'>Doprava</a>
 
     .content
-
+      %p
+        =File.read('data/timestamp.html')
       %table.twocolumn
         %tr
           %td(style="padding: 16px;")