oleges1 / code-completion Goto Github PK

Hello, and thank you for sharing this work.

Got a little question:

the paper (https://arxiv.org/abs/1711.09573) says we need to encode two additional bits of info for each flattened AST node:

• has children
• has a right sibling

the note about "right" is essential, as otherwise it will be impossible (I believe) to restore the tree.

However, the implementation seems to encode the fact that this node has (any) siblings (not necessarily a "right" one):

Am I missing anything or it's something to be be fixed?

Thank you.

In the original paper, it is said that each program is divided into many short segments (len=50) to feed into model respectively and they are interacted by reusing the hidden states and memory states from previous segment as initial states. However, in this implementation, two points do not fit this description:

1. Segments from the same program are scattered in the data list by sorting with their lengths (maybe for reducing padding EOF I guess ?) during data loading and they may be even not in a single batch together during forward. How can they interact with each other without existing in the same batch?
2. The initial values of hs and hc are always set to default value in MixturePointer (ones and None respectively), they are never using previous states as initial values, hence no interactions between segments of the same program.

This is only my own question for this implementation. Thanks for any explainations or replies.