#!/bin/bash

#####################################################################
#       Script:  autoPicSmth.sh
#       Author:  cheyo
#        Email:  icheyo at Gmail dot com
#         From:  www.icheyo.net
#         Date:  2008-02-22
#
#  Description:
#      This script is used for downloading pictures from the MyPhoto 
#      board in the newsmth.net automatically.
#
#####################################################################

# Usage: autoPicSmth.sh [days] 
# days:  download all pictures of recent /days/ days
# For Example:   ./autoPicSmth.sh 3

WORKING_DIR=working
PIC_OUT_DIR=photo
DAYS_TO_DOWN=3
QUERY_FILE=QueryResult.tmp
THREAD_FILE=ThreadUrl.tmp
FORMAT_FILE=ThreadInfo.tmp
CURR_THREAD_FILE=CurrThread.tmp 
PIC_URL_FILE=PicUrl.tmp
PIC_DOWN_LOG=PicDown.log
PIC_INFO_FILE1=PicInfo1.tmp
PIC_INFO_FILE2=PicInfo2.tmp
PIC_INFO_FILE3=PicInfoFinal.tmp

# ------------------------------------------------------------------ #
# ShowUsage()
# Show the usage of this script
# ------------------------------------------------------------------ #
ShowUsage()
{
    echo "This script is used for automatic downloading pictures from MyPhoto board in the newsmth.net"
    echo "  Usage: autoPicSmth.sh [days]"
    echo "  days:  download all pictures of recent /days/ days. 3 for default."
    echo "  Example:   ./autoPicSmth.sh 3"
}

# check arguments
if [ $# -gt 1 ]; then
    ShowUsage
    exit 1;
elif [ $# -eq 1 ]; then
    DAYS_TO_DOWN=$1
fi

mkdir -p $WORKING_DIR 
cd $WORKING_DIR

# Get the thread search result HTML page to local
SearchUrl="http://bbs4.newsmth.net/bbsbfind.php?q=1&board=MyPhoto&dt=${DAYS_TO_DOWN}&ag=1"
curl "${SearchUrl}" -o ${QUERY_FILE}

# Create a file to store all Thread URLs
egrep "<a href="bbscon.php?bid=" $QUERY_FILE | awk -F[<>"] '{print "http://bbs4.newsmth.net/"$9}' > $THREAD_FILE
ThreadCount=`cat $THREAD_FILE | wc -l`
echo "Total ${ThreadCount} threads are found."

# Create a file to store all BoardId and ThreadId
awk -F[=&] '{print $2,$4}' $THREAD_FILE > $FORMAT_FILE 

# Create a file to sotre all pictures infomation
# Format: BoardId ArticleId FileName FileSize FileId
echo "# BoardId ArticleId FileName FileSize FileId" > $PIC_INFO_FILE1

cat $FORMAT_FILE | while read BoardId ArticleId
do
    ThreadUrl=`echo "http://bbs4.newsmth.net/bbscon.php?bid=$BoardId&id=$ArticleId"`
    curl "$ThreadUrl" -o $CURR_THREAD_FILE
    grep "attach" $CURR_THREAD_FILE | tr ");" ") " | grep "attach" | awk -F[' ,)] -v BoardId=$BoardId -v ArticleId=$ArticleId '{print BoardId,ArticleId,$2,$5,$7}' >> $PIC_INFO_FILE1
done

# Create a file to store all pictures info with file extention name 
# but not full file name.
# Format:  BoardId ArticleId FileExt FileSize FileId
# echo "# BoardId ArticleId FileExt FileSize FileId" > $PIC_INFO_FILE2
awk -F[. ]  '$0~/^[^#]/ {print $1,$2,$4,$5,$6}' $PIC_INFO_FILE1 >> $PIC_INFO_FILE2
# Remove the records which don't contain enough info.
# in normal case, it should be 5 columns in the file.
awk '$5~/^[^$]/ {print $0}' $PIC_INFO_FILE2 > $PIC_INFO_FILE3 

# Create a file to store all picture url
grep ^[^#] $PIC_INFO_FILE3 | while read BoardId ArticleId FileExt FileSize FileId
do
    if [ $FileSize -gt 51200 ]; then
        FileType="p"
    else
        FileType="s"
    fi

    PicUrl=`echo "http://att.newsmth.net/att.php?$FileType.$BoardId.$ArticleId.$FileId.$FileExt"`
    echo "$PicUrl" >> $PIC_URL_FILE 
done

# Remove all duplicted URL from the file
mv ${PIC_URL_FILE} ${PIC_URL_FILE}.tmp
sort -dfu ${PIC_URL_FILE}.tmp > ${PIC_URL_FILE}
rm ${PIC_URL_FILE}.tmp

# Remove the URLs which have been downed before
if [ -f "../${PIC_OUT_DIR}/${PIC_DOWN_LOG}" ]; then
    cp ../$PIC_OUT_DIR/${PIC_DOWN_LOG} .
    awk '{print $3}' ${PIC_DOWN_LOG} > ${PIC_URL_FILE}.history
    sort -dfu ${PIC_URL_FILE}.history > ${PIC_URL_FILE}.tmp
    mv ${PIC_URL_FILE}.tmp ${PIC_URL_FILE}.history
    comm -1 -3 ${PIC_URL_FILE}.history ${PIC_URL_FILE} > ${PIC_URL_FILE}.tmp
    mv ${PIC_URL_FILE}.tmp ${PIC_URL_FILE}
    rm ${PIC_URL_FILE}.history
fi

# Download all pictures from server to local 
PicCount=`wc -l $PIC_URL_FILE | awk '{print $1}'`
PicIndex=1
mkdir -p ../$PIC_OUT_DIR 
echo "Total number of pictures to be downloaded: $PicCount"

cat $PIC_URL_FILE | while read CurrUrl
do
    FileName=`echo "$CurrUrl" | awk -F[?] '{print $2}'`
    echo "[$PicIndex/$PicCount] Start to download $CurrUrl"
    curl "$CurrUrl" -o ../$PIC_OUT_DIR/$FileName
    # Write download log to log file
    CurrTime=`date +"%Y-%m-%d %H:%M:%S"`
    echo "$CurrTime $CurrUrl" >> "../$PIC_OUT_DIR/$PIC_DOWN_LOG"
    echo "[$PicIndex/$PicCount] Download finished."
    echo ""
    PicIndex=`expr $PicIndex + 1`
done

#mv $PIC_URL_FILE ../$PIC_OUT_DIR/PicUrl.list
#mv $PIC_INFO_FILE3 ../$PIC_OUT_DIR/PicInfo.list
cd ..
rm -r $WORKING_DIR

echo "All Pictures Downloading finished."

http://write.blog.csdn.net/postedit