[pyblosxom2wxr] / pyblosxom2wxr.sh

#!/bin/bash
#
# pyblosxom2wxr.sh
# http://snarfed.org/pyblosxom2wxr
# Ryan Barrett <pyblosxom2wxr@ryanb.org>
# Version 0.2. This script is public domain.
#
# This script converts PyBlosxom posts and comments into a WXR (WordPress
# eXtensible RSS) XML file that can be imported into a WordPress blog.
#
# Example usage:
#
# $ ./pyblosxom2wxr.sh post1.txt post2.txt > posts.xml
#
# pyblosxom2wxr has been tested with PyBlosxom 1.4 and WordPress 2.9 and 3.0. It
# should work with other versions too, but your mileage may vary.
#
# TODO: comment ordering

# exit on error
set -e

# check args
if [[ $# = "0" || $1 = "--help" ]]; then
  echo 'Usage: pyblosxom2wxr.sh FILES...'
  exit 1
fi

# category
all_categories=""
all_tags=""

# comment id sequence number
commentid=1

# output header
now=`date --rfc-3339=seconds`
cat << EOF
<?xml version="1.0" encoding="UTF-8"?> 

<!-- generator="pyblosxom2wxr/1.0" created="${now}" -->
<rss version="2.0"
  xmlns:excerpt="http://wordpress.org/export/1.0/excerpt/"
  xmlns:content="http://purl.org/rss/1.0/modules/content/"
  xmlns:wfw="http://wellformedweb.org/CommentAPI/"
  xmlns:dc="http://purl.org/dc/elements/1.1/"
  xmlns:wp="http://wordpress.org/export/1.0/">

<channel>
  <title></title>
  <link></link>
  <description></description>
  <pubDate></pubDate>
  <generator>http://snarfed.org/pyblosxom2wxr?v=1.0</generator>
  <language>en</language>
  <wp:wxr_version>1.0</wp:wxr_version>
  <wp:base_site_url></wp:base_site_url>
  <wp:base_blog_url></wp:base_blog_url>
EOF


# convert blog posts
for file in "$@"; do
  creator="mako"
  fullname=`basename "$file" .txt`
  dir=`dirname "$file"`
  commentdir="$(dirname "$file")/comments"
  title=`head -n 1 "$file"`
  type="post"

  name=${fullname}

  pubDate=$(date -uR -r $file)
  date=$(date -r "$file" +'%F %T')
  dateGmt=$(date -u -r "$file" +'%F %T')

  if grep -q ']]>' "$file"; then
    echo "WARNING: $file contains the string ]]>, which makes its CDATA " \
         "section invalid. WordPress handles this ok, but still, heads up." 1>&2
  fi

  cat << EOF
<item>
  <title>${title}</title>
  <pubDate>${pubDate}</pubDate>
  <guid isPermaLink="true">/${fullname}</guid>
  <description></description>
  <content:encoded><![CDATA[$(tail -n +4 "$file" | markdown_py)]]></content:encoded>
  <wp:post_date>${date}</wp:post_date>
  <wp:post_date_gmt>${dateGmt}</wp:post_date_gmt>
  <wp:comment_status>open</wp:comment_status>
  <wp:ping_status>open</wp:ping_status>
  <wp:post_name>${name}</wp:post_name>
  <wp:status>publish</wp:status>
  <wp:post_parent>0</wp:post_parent>
  <wp:menu_order>0</wp:menu_order>
  <wp:post_type>${type}</wp:post_type>
  <wp:post_password></wp:post_password>
  <wp:is_sticky>0</wp:is_sticky>
  <dc:creator>${creator}</dc:creator>
EOF
 
  # split the tags
  raw_tags=$(grep '#tags' "$file" |perl -pe 's/.tags // '|tr ',' "\n")
  for tag in $raw_tags; do
    echo "<category domain=\"post_tag\" nicename=\"${tag}\">${tag}</category>"
    #echo "<category domain=\"category\" nicename=\"$category\">$category</category>"
  done;
  all_tags="$all_tags$raw_tags"

  # category="uncategorized"
  # <category domain="category" nicename="$category">$category</category>

  # other possible elements:
#  <link>/${fullname}</link>
#  <wp:post_id></wp:post_id>
#  <excerpt:encoded></excerpt:encoded>


  for cmtfile in ${commentdir}/"$fullname"-{all,[0-9]*}.cmt; do
    if [[ -e "$cmtfile" ]]; then
      set +e  # because the perl script below uses a non-zero exit code
      tail -q -n +2 "$cmtfile" | \
        sed -r '
          s/^<item>$/<wp:comment>\n<wp:comment_id>X<\/wp:comment_id>/;
          s/^<\/item>$/<wp:comment_approved>1<\/wp:comment_approved>\n<\/wp:comment>/;
          s/<(\/)?author>/<\1wp:comment_author>/g;
          s/<(\/)?link>/<\1wp:comment_author_url>/g;
          s/<(\/)?ipaddress>/<\1wp:comment_author_IP>/g;
          s/<(\/)?description>/<\1wp:comment_content>/g;
          s/^<(ajax|cmt_date|email|openid_url|parent|post|secretToken|source|title|w3cdate)>.+$//;
          s/^<\/?items>$//;
          /^$/d' | \
        perl -pe 'use HTML::Entities; decode_entities($_)' | \
        perl -pe 'use POSIX qw(strftime);
                  s/^<pubDate>(.+)<\/pubDate>$/"<wp:comment_date>" . (strftime "%Y-%m-%d %H:%M:%S", localtime($1)) . "<\/wp:comment_date>"/e;' | \
        perl -e '
          my $id = '${commentid}';
          while (<STDIN>) {
             s/^(<wp:comment_id>)X(<\/wp:comment_id>)$/$1 . $id++ . $2/e;
             print $_;
          }
          exit $id - '${commentid}';'
      # TODO: this is a hack since exit codes are only 8 bits unsigned.
      # this will break on posts with >255 comments.
      let commentid+=$?
      set -e
    fi
  done

  cat << EOF
</item>

EOF
done

index=2
#for category in $(echo "$all_categories"|sort|uniq); do
#    cat << EOF
#    <wp:category><wp:term_id>${index}</wp:term_id><wp:category_nicename>${category}</wp:category_nicename><wp:category_parent></wp:category_parent><wp:cat_name><![CDATA[${category}]]></wp:cat_name></wp:category>
#EOF
#    index=$(expr $index + 1)
#done

for tag in $(echo "$all_tags"|sort|uniq); do
    cat << EOF
    <wp:tag><wp:term_id>${index}</wp:term_id><wp:tag_slug>${tag}</wp:tag_slug><wp:tag_name>${tag}</wp:tag_name></wp:tag>
EOF
    index=$(expr $index + 1)
done


# output footer
cat << EOF
</channel>
</rss>
EOF