TexturePacker大图还原成小图工具带源码

    TexturePacker是一个把好多小图打成大图的软件,生成的是大图以及小图在大图位置的.plist描述文件,但是不支持把大图还原成小图。网上偷的图一般都是大图和plist,想得到小图比较麻烦,于是乎用python写了个TexturePacker反向工具,把大图导成小图。

  1.python要用到的库

    python的图片处理要用到PIL(Python Image Library),mac10.10下安装PIL会报 fatal error: 'X11/Xlib.h' file not found的错,解决方法在此。而且在装PIL前要先装zlib/libpng/jpeg,安装方法自行百度。

    plist解析用了xml.dom为python自带的库,不用装。

  2.TexturePacker导出的plist结构分析

    plist文件如下所示。 

 1 <?xml version="1.0" encoding="UTF-8"?>
 2 <!DOCTYPE plist PUBLIC "-//Apple Computer//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
 3 <plist version="1.0">
 4     <dict>
 5         <key>frames</key>
 6         <dict>
 7             <key>grossini_dance_03.png</key>
 8             <dict>
 9                 <key>frame</key>
10                 <string>{{46,324},{63,109}}</string>
11                 <key>offset</key>
12                 <string>{-6,-1}</string>
13                 <key>rotated</key>
14                 <false/>
15                 <key>sourceColorRect</key>
16                 <string>{{5,7},{63,109}}</string>
17                 <key>sourceSize</key>
18                 <string>{85,121}</string>
19             </dict>
20         </dict>
21         <key>metadata</key>
22         <dict>
23             <key>format</key>
24             <integer>2</integer>
25             <key>realTextureFileName</key>
26             <string>bbb.png</string>
27             <key>size</key>
28             <string>{512,512}</string>
29             <key>smartupdate</key>
30             <string>$TexturePacker:SmartUpdate:ea1bbb1419cd4c346debb54e1a7d5de2:1/1$</string>
31             <key>textureFileName</key>
32             <string>bbb.png</string>
33         </dict>
34     </dict>
35 </plist> 

    frames对应的value是所有小图的信息。key为小图的名字,dict为小图的信息。

    frame为图片小图位置及大小(这个大小是经过Trim的大小,TP会把png的无像素白边剔除,来减小图片的大小,也就是说大图中小图的大小不一定等与小图真正的大小)。如上plist{{46,324},{63,109}},46和324为小图在大图中x,y坐标,{63,109}为经过裁剪的图片大小,图2其实是png格式黄色背景是空的,为了看着方便,加两个黄色背景。{63,109}是红框内的大小。

图1图2图3

    rotated是是否旋转,这个光头没有旋转,大图中的play按钮有旋转。

    sourceSize为小图的大小。{85, 121}为小图整个大小,没有经过Trim。

    sourceColorRect为经过Trim的图在小图中的起始坐标及大小。{{5,7},{63, 109}}中{5,7}为图2中红框左上角的坐标,{63,109}为大图中小图的大小。

    offset为中心坐标偏移。图3中小红点处为原大小图中心点p1,小红点左边的交叉点为经过Trim的图片中心点p2,以p1为原点,p2的坐标就是这个offset{-6,-1},x向负轴偏移6像素,y向负轴偏移1像素,这里比较奇怪,y轴好像是向上为正了。

   不知这个有offset有神马用,下面的代码没有用offset,用sourceColorRect、sourceSize、frame、rotated就可以确定出小图的样子。

  3.代码分析

    代码中用到了PIL,PIL的参考手册在这,程序中用到了open、new、crop、crop、rotate几个api。

    代码分三个文件,主文件TexturePacker.py,运行此来生成小图。PlistToDict.py来解析plist文件,生成map,key为小图名,value为小图信息。BraceParser.py为{{2,3},{4,5}}生成列表[[2,3][4,5]]。

    知道了plist的含义,稍微会用PIL,代码应该很好理解,代码如下,玩具代码,莫要嘲笑。

    TexturePacker.py,类TextureParser第一个参数是plist及png文件位置,第二个为文件名字。

 1 import PIL.Image as Image
 2 import BraceParser
 3 import PlistToDict
 4 
 5 class TextureParser(object):
 6     def __init__(self, path, name):
 7         self.__resPath = path;
 8         self._name = name;
 9         self.__plistDict = PlistToDict.PlistToDict(path + "/" + name + ".plist").createDict();
10 
11     #return map[picName : {originPoint : {x:, y:}, size : {width:, height:}}]
12     def __getSmallPicInfos(self):
13         picInfo = {};
14         for key, value in self.__plistDict["frames"].items():
15             size = BraceParser.BraceParser(value["sourceSize"]).createList();
16             origin = BraceParser.BraceParser(value["frame"]).createList();
17             sourceSize = BraceParser.BraceParser(value["sourceColorRect"]).createList();
18             picInfo[key] = {"size" : size,
19                             "origin" : [origin[0][0], origin[0][1]],
20                             "colorOrigin" : [sourceSize[0][0], sourceSize[0][1]],
21                             "colorSize" : [sourceSize[1][0], sourceSize[1][1]],
22                             "isRotated" : value["rotated"]
23                             };
24         return picInfo;
25 
26     def smallPicsCreate(self, pathToStore = None):
27         image = Image.open(self.__resPath + "/" + self._name + ".png");
28         picInfos = self.__getSmallPicInfos();
29         for k, v in picInfos.items():
30             if v["isRotated"] == True:
31                 v["size"][0], v["size"][1] = v["size"][1], v["size"][0];
32                 v["colorSize"][0], v["colorSize"][1] = v["colorSize"][1], v["colorSize"][0];
33             newImage = Image.new("RGBA", (int(v["size"][0]),int(v["size"][1])));
34             box = (int(v["origin"][0]), int(v["origin"][1]),
35                    int(v["origin"][0] + v["colorSize"][0]), int(v["origin"][1] + v["colorSize"][1]));
36             region = image.crop(box);
37             newImage.paste(region, (int(v["colorOrigin"][0]), int(v["colorOrigin"][1])));
38             if v["isRotated"] == True:
39                 newImage = newImage.rotate(90);
40             newImage.save(self.__resPath + k);
41 
42 if __name__ == "__main__":
43     textureUnPacker = TextureParser("/Users/adv/Desktop/", "bbb");
44     textureUnPacker.smallPicsCreate();
45     print("success!")

    PlistToDict.py,用的是dom解析plist。dom怎么用自行百度。

 1 from xml.dom import minidom
 2 
 3 class PlistToDict(object):
 4     def __init__(self, plistPath):
 5         dom = minidom.parse(plistPath);
 6         self.__root = dom.documentElement;
 7 
 8     # get root dict
 9     def __getFirstDictDoc(self):
10         children = self.__root.childNodes;
11         for v in children:
12             if v.nodeType == v.ELEMENT_NODE and v.nodeName == "dict":
13                 return v;
14         return None;
15 
16     # get value by key in doc's children
17     def __getValueDocByKey(self, doc, key):
18         children = doc.childNodes;
19         for v in children:
20             if v.nodeType == v.ELEMENT_NODE and v.nodeName == "key" and v.firstChild.nodeValue == key:
21                 node = v.nextSibling;
22                 while node.nodeType != node.ELEMENT_NODE:
23                     node = node.nextSibling;
24                     if node == None:
25                         return None;
26                 return node;
27         return None;
28 
29     def __firstElementNodeName(self, doc):
30         for v in doc.childNodes:
31             if v.nodeType == v.ELEMENT_NODE:
32                 return v.nodeName;
33 
34     def __docToDict(self, dom, dic):
35         keys = self.__getAllKeyValuesInDoc(dom);
36         for key in keys:
37             valueNode = self.__getValueDocByKey(dom, key);
38             if valueNode.nodeName == "dict":
39                 dic[key] = {}
40                 self.__docToDict(valueNode, dic[key]);
41             elif valueNode.nodeName == "false":
42                 dic[key] = False;
43             elif valueNode.nodeName == "true":
44                 dic[key] = True;
45             else:
46                 dic[key] = valueNode.firstChild.nodeValue;
47 
48     def __getAllKeyValuesInDoc(self, doc):
49         ret = [];
50         for v in doc.childNodes:
51             if v.nodeName == "key":
52                 ret.append(v.firstChild.nodeValue);
53         return ret;
54 
55     def createDict(self):
56         rootDict = self.__getFirstDictDoc();
57         ret = {};
58         self.__docToDict(rootDict, ret);
59         return ret;

    BraceParser.py,用来解析括号。

 1 class BraceParser(object):
 2     def __init__(self, str):
 3         self.__strToParse = str.replace(" ", "");
 4 
 5     def __firstStrIsLeftBrace(self, str):
 6         return True if str[0] == "{" else False;
 7 
 8     def __subOutBrace(self, str):
 9         return str[1:-1];
10 
11     def __findAllSeqCommaPos(self, str):
12         bracketNum = 0;
13         ret = [];
14         for i, v in enumerate(str):
15             if v == "{":
16                 bracketNum += 1;
17             elif v == "}":
18                 bracketNum -= 1;
19             elif v == ",":
20                 if bracketNum == 0:
21                     ret.append(i);
22         return ret;
23 
24     # {111,324},{100,100} return ["{111,324}", "{100,100}"]
25     def __getAllBraceStrs(self, str):
26         listStr = [];
27         posList = self.__findAllSeqCommaPos(str);
28         lastPos = -1;
29         for v in posList:
30             listStr.append(str[lastPos + 1: v]);
31             lastPos = v;
32         listStr.append(str[lastPos + 1: ]);
33         return listStr;
34 
35     def __getValue(self, str):
36         listStr = str.split(",");
37         return listStr[0], listStr[1];
38 
39 
40     def __listCreate(self, str, listIns):
41         if self.__firstStrIsLeftBrace(str) == True:
42             braceStrs = self.__getAllBraceStrs(str);
43             for v in braceStrs:
44                 subList = [];
45                 listIns.append(subList);
46                 self.__listCreate(self.__subOutBrace(v), subList);
47         else:
48             x, y = self.__getValue(str);
49             listIns.append(float(x));
50             listIns.append(float(y));
51 
52     def createList(self):
53         listIns = [];
54         str = self.__subOutBrace(self.__strToParse);
55         self.__listCreate(str, listIns);
56         return listIns;

    最后,我想问博客园怎么上传附件?

posted @ 2015-07-19 21:28  BigFengFeng  阅读(4506)  评论(0编辑  收藏  举报