2

我正在尝试制作一个简单的算法，将 .obj 文件转换为我的程序可以使用的数据。

这是输入字符串数据示例：“v 1.000000 -1.000000 -1.000000”“f 1 2 3 4”

file = new BufferedReader(new InputStreamReader(am.open("test.obj"))); //Loads the .obj
        String out = " ";
        try{
            while((out=file.readLine()) != null){ //Grabs a line from the file and checks so that it aint null
                int index = 1;
                int nextIndex = 1;
                if(out.startsWith("v")){Log.w("OBJs", out); //If line starts with "v", go ahead
                    while(index != out.lastIndexOf(" ")){ //checks so the index is not the last index of an empty space " "
                        nextIndex = out.indexOf(" ", index +1); //Looks for the next occurance of an empty space from +1 of current empty space
                        vertices.add(Float.valueOf(out.substring(index, nextIndex))); //Grabs the string in between the two spaces
                        index = nextIndex;// 
                        }
                    vertices.add(Float.valueOf(out.substring(index, out.length())));//Grabs the last variable from the file as the above while is now false
                    }

                int index2 = 1;
                int nextIndex2 = 1;
                if(out.startsWith("f")){Log.w("OBJs", out);// THis works exactly as above, gives error and i dunno why
                    while(index2 != out.lastIndexOf(" ")){
                        nextIndex2 = out.indexOf(" ", index2 +1);
                        edges.add(Integer.valueOf(out.substring(index2, nextIndex2)));
                        Log.w("OBJs", out.substring(index2, nextIndex2)); 
                        index2 = nextIndex2;

                    }
                    edges.add(Integer.valueOf(out.substring(index2, out.length())));
                }

            }
        }

这导致：

FATAL EXCEPTION: main
java.lang.RuntimeException: Unable to start activity ComponentInfo{my.stuffs.blodvite/my.stuffs.blodvite.MainActivity}: java.lang.NumberFormatException: Invalid int: " 1"

而这只发生在第二个循环部分，我试图收集面部信息。收集顶点完美无缺，我对两者都使用完全相同的算法，所以我不明白问题是什么。还有一件有趣的事情是，当我上次关闭 eclipse 时，我认为它确实有效，不知道发生了什么，或者我是否记错了。哦，正如您从代码中看到的那样，它将读取的行打印到 Log cat。这就是结果。

01-26 23:42:13.916: W/OBJs(18442): v 1.000000 -1.000000 -1.000000
01-26 23:42:13.916: W/OBJs(18442): v 1.000000 -1.000000 1.000000
01-26 23:42:13.916: W/OBJs(18442): v -1.000000 -1.000000 1.000000
01-26 23:42:13.916: W/OBJs(18442): v -1.000000 -1.000000 -1.000000
01-26 23:42:13.916: W/OBJs(18442): v 1.000000 1.000000 -0.999999
01-26 23:42:13.916: W/OBJs(18442): v 0.999999 1.000000 1.000001
01-26 23:42:13.916: W/OBJs(18442): v -1.000000 1.000000 1.000000
01-26 23:42:13.916: W/OBJs(18442): v -1.000000 1.000000 -1.000000
01-26 23:42:13.916: W/OBJs(18442): f 1 2 3 4

所以它清楚地表明顶点收集工作没有错误，但是面收集在第一次迭代时停止。

4

1 回答 1

1

This happens because you have a leading space in " 1". As already pointed out by @Raghav Sood, you can call trim on your Strings before trying to parse to int to get it to work. You can also simply take the substring starting at index2+1 (i.e. skipping the first space).

I am curious why you didn't simply use StringTokenizer for this.

于 2013-01-28T19:36:00.853 回答