article.cgi - gophercgis - Collection of gopher CGI/DCGI for geomyidae
 (HTM) hg clone https://bitbucket.org/iamleot/gophercgis
 (DIR) Log
 (DIR) Files
 (DIR) Refs
 (DIR) README
 (DIR) LICENSE
       ---
       article.cgi
       ---
            1 #!/bin/sh
            2 
            3 . ../common/config.sh
            4 . ../common/html.sh
            5 
            6 #
            7 # Extract article content
            8 #
            9 extract_article()
           10 {
           11 
           12         awk '
           13         /<article/,/<\/article>/ {
           14                 print
           15         }
           16         '
           17 }
           18 
           19 
           20 #
           21 # Filter HTML content
           22 #
           23 filter_html()
           24 {
           25 
           26         awk '
           27         /<div class="hentry hentry--item hentry--banner">/,/<\/div>/ {
           28                 next
           29         }
           30         /<div class="item_tag_single">/,/<\/div>/ {
           31                 next
           32         }
           33         /<div class="item_share_inline">/,/<\/div>/ {
           34                 next
           35         }
           36         /<div class="item_tags" itemprop="keywords">/,/<\/div>/ {
           37                 next
           38         }
           39         /<div class="item_subscribe">/,/<\/div>/ {
           40                 next
           41         }
           42         /<div class="content_date">/,/<\/div>/ {
           43                 next
           44         }
           45 
           46         {
           47                 # Remove all img-s
           48                 gsub(/<img [^>]+\/?>/, "")
           49 
           50                 print
           51         }
           52         '
           53 
           54 }
           55 
           56 
           57 url=$2
           58 curl -Lgs -- "${url}" |
           59     extract_article |
           60     filter_html |
           61     html_to_text
           62 
           63 echo ""
           64 echo "URL: <${url}>"