handspeaker/word2vec

word2phrase: AddWord2Vocab strcpy's full length of strings longer than MAX_STRING

GoogleCodeExporter opened this issue · 1 comments

The strcmp should be changed to strncmp.  This may be causing segmentation 
faults in some cases when the input contains words that are too long.

Original issue reported on code.google.com by chad.p...@gmail.com on 17 Jul 2015 at 11:12

Also, in SortVocab() the free is indexed to vocab_size, not the element 
actually being freed.

Original comment by chad.p...@gmail.com on 17 Jul 2015 at 11:26