MPI_File_read错误 [英] MPI_File_read Error

问题描述

我正在尝试使用MPI_File_read从输入文件中读取.但是它没有正确读取值.这是我的输入文件:

I am trying to use MPI_File_read to read from an input file. But it is not correctly reading the values. Here is my input file:

11 3 4 5 2

这是我用来尝试阅读的代码:

And here is the code i am using to try to read it:

char *filename = "input/8";
int n;

if (MPI_File_open(comm_2d, filename, MPI_MODE_RDONLY, MPI_INFO_NULL, &f) != MPI_SUCCESS) {
    fprintf(stderr, "Cannot open file %s\n", filename);
    MPI_Abort(comm_2d, FILE_NOT_FOUND);
    MPI_Finalize();
    return 1;
}
MPI_File_seek(f, 0, MPI_SEEK_SET);
MPI_File_read(f, &n, 1, MPI_INT, &status);

cout << "n :" << n << endl;
MPI_File_close(&f);

根据我的理解，这应该从input.txt中读取第一个int"11".但是，如果给我一个很大的数字.

From my understanding this should read the first int "11" from the input.txt. But if gives me a very large number instead.

我在做什么错了?

推荐答案

MPI_File_Read从文件中读取MPI_INT的二进制表示形式.如果给它一个ASCII(文本)文件，它可能会将前4个字符解释为一个32位整数，并按照您的描述返回非常大的数字".

MPI_File_Read reads a binary representation of an MPI_INT from your file. If you give it an ASCII (text-) -file it'll probably interpret the first 4 characters as one 32-bit integer and return a "very large number" just as you describe.

在这种情况下，您可以例如将整个文件读取到缓冲区中，然后使用sscanf()或任何其他可用的文本解析功能.

In that case, you could for example read the whole file into a buffer and then use sscanf() or any other text parsing function you have available.

这篇关于MPI_File_read错误的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！