initial import

git-svn-id: https://pandoc.googlecode.com/svn/trunk@2 788f1e2b-df1e-0410-8736-df70ead52e1b

1 files changed, 39 insertions, 0 deletions

diff --git a/html2markdown b/html2markdown
new file mode 100644
index 000000000..3f9a4857e
--- /dev/null
+++ b/html2markdown
@@ -0,0 +1,39 @@
+#!/bin/sh -e
+# converts html to markdown
+# uses an available program to fetch URL and tidy to normalize it first
+
+[ -n "$(which pandoc)" ] || {
+    echo >&2 "You need 'pandoc' to use this program!"
+    exit 1
+}
+[ -n "$(which tidy)" ] || {
+    echo >&2 "You need 'tidy' to use this program!"
+    exit 1
+}
+
+if [ -z "$1" ] || [ -f $1 ]; then 
+    tidy -utf8 $1 2>/dev/null | pandoc -r html -w markdown -s 
+else
+    # Treat given argument as an URL.  Locate a
+    # sensible text based browser (note the order).
+    for p in wget lynx w3m curl links w3c; do
+        if which $p >/dev/null; then
+            DUMPER=$p
+            break
+        fi
+    done
+    # Setup proper options.
+    case "$DUMPER" in
+        wget)  OPT="-O-" ;;
+        lynx)  OPT="-source" ;;
+        w3m)   OPT="-dump_source" ;;
+        curl)  OPT="" ;;
+        links) OPT="-source" ;;
+        w3c)   OPT="-n -get" ;;
+        "")    echo -n >&2 "Needs a program to fetch the URL "
+               echo -n >&2 "(e.g. wget, w3m, lynx, w3m or curl)."
+               exit 1 ;;
+    esac
+    # Fetch and feed to pandoc.
+    $DUMPER $OPT $1 2>/dev/null | tidy -utf8 2>/dev/null | pandoc -r html -w markdown -s
+fi