The file "image.txt" contains the information of qualified images(with emotion). Each entry is for a image, which is designed to written using a pattern:imgId,ownerId,title,imgDate,numofDesc,numofTag,tag0,tag1..,numOfComm+"\n"+desc + "\n" + comm1 + "\n" +comm2 + "\n" + ...+ "\n". And the "," is the separator.

for example, there are several lines for a image:

This seven lines means: "113788" is imgId, "13" is ownerId, "Industrial SF" is title, "2004-07-26 03:49:11" is the published date, "1" means there is one line for description, "10" means there are 10 tags, and "San Francisco,industrial,port,water,crane,black and white,unfound,bw,The Ramp,foundinsf" are 10 tags, "5" is the number of following comments,

"This picture was taken near the The Ramp. I converted it into a blackand white picture and added some extra contrast. This is one of thethings I miss from analog photography: Seeing Black and White" is the description,

"10,2005-05-26 08:00:00,Islais Creek?" is a comment, and "10" is the id of commentator, "2005-05-26 08:00:00" is the published date of comment, and "Islais Creek?" is the content. The four following lines are the same;