gfm.parse() hangs for a specific input

[Emre Sülün]

A minimal example:

from marko.ext.gfm import gfm
input = "\r\n\r\n['nvcc', '-shared', '-O3', '-use_fast_math', '-Xlinker', '/DEBUG', '-D HAVE_ROUND', '-m64', '-Xcompiler', '-DCUDA_NDARRA\r\nY_CUH=c72d035fdf91890f3b36710688069b2e,-DNPY_NO_DEPRECATED_API=NPY_1_7_API_VERSION,/Zi,/MD', '-IC:\\\\Users\\\\Quantum Liu\\\\\r\nAnaconda2\\\\lib\\\\site-packages\\\\theano\\\\sandbox\\\\cuda', '-IC:\\\\Users\\\\Quantum Liu\\\\Anaconda2\\\\lib\\\\site-packages\\\\numpy\\\\\r\ncore\\\\include', '-IC:\\\\Users\\\\Quantum Liu\\\\Anaconda2\\\\include', '-IC:\\\\Users\\\\Quantum Liu\\\\Anaconda2\\\\lib\\\\site-packages\r\n\\\\theano\\\\gof', '-o', 'C:\\\\Users\\\\Quantum Liu\\\\AppData\\\\Local\\\\Theano\\\\compiledir_Windows-10-10.0.14393-Intel64_Family_6\r\n_Model_60_Stepping_3_GenuineIntel-2.7.12-64\\\\cuda_ndarray\\\\cuda_ndarray.pyd', 'mod.cu', '-LC:\\\\Users\\\\Quantum Liu\\\\Anaco\r\nnda2\\\\libs', '-LC:\\\\Users\\\\Quantum Liu\\\\Anaconda2', '-lcublas', '-lpython27', '-lcudart']\r\nERROR (theano.sandbox.cuda): Failed to compile cuda_ndarray.cu: ('nvcc return status', 2, 'for cmd', 'nvcc -shared -O3 -\r\nuse_fast_math -Xlinker /DEBUG -D HAVE_ROUND -m64 -Xcompiler -DCUDA_NDARRAY_CUH=c72d035fdf91890f3b36710688069b2e,-DNPY_NO\r\n_DEPRECATED_API=NPY_1_7_API_VERSION,/Zi,/MD -IC:\\\\Users\\\\Quantum Liu\\\\Anaconda2\\\\lib\\\\site-packages\\\\theano\\\\sandbox\\\\cu\r\nda -IC:\\\\Users\\\\Quantum Liu\\\\Anaconda2\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include -IC:\\\\Users\\\\Quantum Liu\\\\Anaconda2\\\\in\r\nclude -IC:\\\\Users\\\\Quantum Liu\\\\Anaconda2\\\\lib\\\\site-packages\\\\theano\\\\gof -o C:\\\\Users\\\\Quantum Liu\\\\AppData\\\\Local\\\\Th\r\neano\\\\compiledir_Windows-10-10.0.14393-Intel64_Family_6_Model_60_Stepping_3_GenuineIntel-2.7.12-64\\\\cuda_ndarray\\\\cuda_n\r\ndarray.pyd mod.cu -LC:\\\\Users\\\\Quantum Liu\\\\Anaconda2\\\\libs -LC:\\\\Users\\\\Quantum Liu\\\\Anaconda2 -lcublas -lpython27 -lcu\r\ndart')\r\nWARNING (theano.sandbox.cuda): CUDA is installed, but device gpu is not available  (error: cuda unavailable)"
result = gfm.parse(input)

The last line just hangs, doesn't throw an exception.