Closed oepen closed 5 years ago
Sorry to be a bit slow here, but how would we know whether a given MRP file should be interpreted as ‘ordered’, and where would we find the relevant information (position of nodes)?
On 16 Jun 2019, at 18:37, Stephan Oepen notifications@github.com<mailto:notifications@github.com> wrote:
./main.py --read mrp data/sample/dm/wsj.mrp --analyze
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/cfmrp/mtool/issues/21?email_source=notifications&email_token=ABT3Q53S3B6AUPE3RP3VOPDP2ZT5RA5CNFSM4HYRRST2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4GZYCG3Q, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ABT3Q53FPUIAER2BUVK7AM3P2ZT5RANCNFSM4HYRRSTQ.
i would look to the top-level ‘flavor’ and/or ‘framework’ properties. for bi-lexical graphs, the ‘nodes’ list is ordered.
Okay, but how do we handle MRP files containing graphs with different flavors? Is it okay to update order-related properties only for those graphs for which flavor == 0?
the idea had not occurred to me that one could mix graphs of different flavors or frameworks in a single file :-). i could imagine we just outlaw that! for example discard any graphs that diverge on these properties from the first entry in the file?