Amir Plivatsky

I have just did it in a slightly different way. I unified `E_list` and `Exp`, i.e. added the `next` field of `E_list` to `Exp` and eliminated `E_list`. (At this step...

My final proposals for these fields are using full names: `operand_first` and `operand_next`. (I will just use short loop variable name in order for `for` lines not to be too...

After the E_list removal is incorporated into LG, I can send a PR to fix `LGDictExpContainer.cc too` (I looked there and the fix is trivial). However, I think it is...

Copied from PR #969. My proposal for fixing `LGDictExpContainer.cc` using added API is at the end. > Be aware that this is currently a part of the external API; there's...

>``` >Exp* lg_exp_next(const Exp*); >``` Is it clear then that it returns the next expression in the current sublist, as opposed to some next expression like the original `next`?

I think that now breadth-first is fine, since I converted the breadth-recursions to pure iterations (without any stack usage or even extra temporary heap memory usage) because expression chains may...

Here is why I think my proposed names are better: I wrote: > ``` > Exp *lg_exp_operand_first(Exp *); > Exp *lg_exp_operand_next(Exp *); > ``` You wrote: > ``` > Exp*...

I just wondered what to return if `lg_exp_operand_first() ` is **wrongly** called with a connector (or even NULL). If it just returns `e->operand_first`, that a crash would happen (immediately if...

Not as far as I know. (For runtime it needs the `graphviz` package, so I guess packagers would want to add it to the recommended packages list to be installed....

Implementations details (still unfinished): 1. Change `void wordgraph_show((Sentence sent, const char *modestr)` to `bool sentence_display_wordgraph(Sentence sent, const char *modestr)` and define it as an API. The intention is that it...