def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
if args.stage:
if args.stage != 'unannotated' and not args.annotator:
sys.exit("--annotator is required unless --stage is unannotated")
elif args.stage == 'unannotated' and args.annotator:
sys.exit("--annotator is forbidden if --stage is unannotated")
output_dir = get_output_dir(args, default_overwrite=True)
corpus = read_corpus(args, verbose=True)
source = args.source
target = _get_target(args, source, corpus)
for k in corpus:
print(k)
doc = corpus[k]
_rename_in_doc(source, target, doc)
save_document(output_dir, k, doc)
pretty_source = anno_id_from_tuple(source)
pretty_target = anno_id_from_tuple(target)
print("Renamed from %s to %s" % (pretty_source, pretty_target),
file=sys.stderr)
announce_output_dir(output_dir)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
corpus = read_corpus(args, verbose=True)
tcache = TimestampCache()
output_dir = get_output_dir(args, default_overwrite=True)
for key in corpus:
print(key)
new_doc = corpus[key]
old_doc = copy.deepcopy(new_doc)
span = _split_dialogue(tcache, new_doc, args.turn)
diffs = _mini_diff(key, args, old_doc, new_doc, span)
print("\n".join(diffs).encode('utf-8'), file=sys.stderr)
save_document(output_dir, key, new_doc)
commit_info = CommitInfo(key=key,
before=old_doc,
after=new_doc,
span=span,
tid=args.turn)
announce_output_dir(output_dir)
if commit_info and not args.no_commit_msg:
print("-----8<------")
print(commit_msg(commit_info))
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
corpus = read_corpus(args, verbose=True)
tcache = TimestampCache()
output_dir = get_output_dir(args, default_overwrite=True)
for key in corpus:
print(key)
new_doc = corpus[key]
old_doc = copy.deepcopy(new_doc)
span = _split_dialogue(tcache, new_doc, args.turn)
diffs = _mini_diff(key, args, old_doc, new_doc, span)
print("\n".join(diffs).encode('utf-8'), file=sys.stderr)
save_document(output_dir, key, new_doc)
commit_info = CommitInfo(key=key,
before=old_doc,
after=new_doc,
span=span,
tid=args.turn)
announce_output_dir(output_dir)
if commit_info and not args.no_commit_msg:
print("-----8<------")
print(commit_msg(commit_info))
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
corpus = read_corpus(args, verbose=True)
output_dir = get_output_dir(args, default_overwrite=True)
for k in corpus:
doc = corpus[k]
if args.diff_friendly:
doc.units = _diff_friendly(doc.units)
doc.relations = _diff_friendly(doc.relations)
doc.schemas = _diff_friendly(doc.schemas)
save_document(output_dir, k, doc)
announce_output_dir(output_dir)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
corpus = read_corpus(args, verbose=True)
output_dir = get_output_dir(args, default_overwrite=True)
for k in corpus:
doc = corpus[k]
if args.diff_friendly:
doc.units = _diff_friendly(doc.units)
doc.relations = _diff_friendly(doc.relations)
doc.schemas = _diff_friendly(doc.schemas)
save_document(output_dir, k, doc)
announce_output_dir(output_dir)
def main():
"create a .seg file for every file in the corpus"
args = mk_argparser().parse_args()
corpus = read_corpus(args)
output_dir = get_output_dir(args)
if args.pipeline:
args.resources = True
args.resource_status = False
args.dialogue_acts = False
args.dialogue_boundaries = False
args.fake_turn_ids = True
config = Config(emit_resources=args.resources,
emit_resource_status=args.resource_status,
emit_dialogue_acts=args.dialogue_acts,
emit_dialogue_boundaries=args.dialogue_boundaries,
fake_turn_ids=args.fake_turn_ids)
for key in corpus:
process_document(config, corpus, key, output_dir)
announce_output_dir(output_dir)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
corpus = read_corpus(args, preselected={'stage': ['discourse', 'units']})
output_dir = get_output_dir(args, default_overwrite=True)
for key in corpus:
doc = corpus[key]
to_delete = []
for sch in doc.schemas:
if not sch.members:
to_delete.append(sch)
for sch in to_delete:
doc.schemas.remove(sch)
save_document(output_dir, key, doc)
announce_output_dir(output_dir)
def main():
"create a .seg file for every file in the corpus"
args = mk_argparser().parse_args()
corpus = read_corpus(args)
output_dir = get_output_dir(args)
if args.pipeline:
args.resources = True
args.resource_status = False
args.dialogue_acts = False
args.dialogue_boundaries = False
args.fake_turn_ids = True
config = Config(emit_resources=args.resources,
emit_resource_status=args.resource_status,
emit_dialogue_acts=args.dialogue_acts,
emit_dialogue_boundaries=args.dialogue_boundaries,
fake_turn_ids=args.fake_turn_ids)
for key in corpus:
process_document(config, corpus, key, output_dir)
announce_output_dir(output_dir)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
corpus = read_corpus(args, preselected={"stage": ["units"]})
output_dir = get_output_dir(args, default_overwrite=True)
for k in corpus:
doc = corpus[k]
for edu in [x for x in doc.units if educe.stac.is_edu(x)]:
etypes = frozenset(educe.stac.split_type(edu))
etypes2 = frozenset(RENAMES.get(t, t) for t in etypes)
if etypes != etypes2:
edu.type = "/".join(sorted(etypes2))
save_document(output_dir, k, doc)
announce_output_dir(output_dir)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
corpus = read_corpus(args,
preselected={"stage": ["units"]})
output_dir = get_output_dir(args, default_overwrite=True)
for k in corpus:
doc = corpus[k]
for edu in [x for x in doc.units if educe.stac.is_edu(x)]:
etypes = frozenset(educe.stac.split_type(edu))
etypes2 = frozenset(RENAMES.get(t, t) for t in etypes)
if etypes != etypes2:
edu.type = "/".join(sorted(etypes2))
save_document(output_dir, k, doc)
announce_output_dir(output_dir)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
corpus = read_corpus(args,
preselected={'stage': ['discourse', 'units']})
output_dir = get_output_dir(args, default_overwrite=True)
for key in corpus:
doc = corpus[key]
to_delete = []
for sch in doc.schemas:
if not sch.members:
to_delete.append(sch)
for sch in to_delete:
doc.schemas.remove(sch)
save_document(output_dir, key, doc)
announce_output_dir(output_dir)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
_screen_args(args)
corpus = read_corpus(args, verbose=True)
output_dir = get_output_dir(args, default_overwrite=True)
old_span = args.span
new_span = Span(old_span.char_start + args.nudge_start,
old_span.char_end + args.nudge_end)
for k in corpus:
old_doc = corpus[k]
new_doc = copy.deepcopy(old_doc)
found = False
for anno in new_doc.units:
if anno.span == old_span:
anno.span = copy.deepcopy(new_span)
found = True
if found:
diffs = _mini_diff(k, (old_doc, old_span), (new_doc, new_span))
print("\n".join(diffs).encode('utf-8'), file=sys.stderr)
else:
print("WARNING: No annotations found for %s in %s" % (old_span, k),
file=sys.stderr)
save_document(output_dir, k, new_doc)
# for commit message generation
span = old_span.merge(new_span)
commit_info = CommitInfo(key=k,
before=old_doc,
after=new_doc,
span=span)
if commit_info and not args.no_commit_msg:
print("-----8<------")
print(commit_msg(commit_info))
announce_output_dir(output_dir)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
_screen_args(args)
corpus = read_corpus(args, verbose=True)
output_dir = get_output_dir(args, default_overwrite=True)
old_span = args.span
new_span = Span(old_span.char_start + args.nudge_start,
old_span.char_end + args.nudge_end)
for k in corpus:
old_doc = corpus[k]
new_doc = copy.deepcopy(old_doc)
found = False
for anno in new_doc.units:
if anno.span == old_span:
anno.span = copy.deepcopy(new_span)
found = True
if found:
diffs = _mini_diff(k, (old_doc, old_span), (new_doc, new_span))
print("\n".join(diffs).encode('utf-8'), file=sys.stderr)
else:
print("WARNING: No annotations found for %s in %s" % (old_span, k),
file=sys.stderr)
save_document(output_dir, k, new_doc)
# for commit message generation
span = old_span.merge(new_span)
commit_info = CommitInfo(key=k,
before=old_doc,
after=new_doc,
span=span)
if commit_info and not args.no_commit_msg:
print("-----8<------")
print(commit_msg(commit_info))
announce_output_dir(output_dir)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
if args.stage:
if args.stage != 'unannotated' and not args.annotator:
sys.exit("--annotator is required unless --stage is unannotated")
elif args.stage == 'unannotated' and args.annotator:
sys.exit("--annotator is forbidden if --stage is unannotated")
output_dir = get_output_dir(args, default_overwrite=True)
corpus = read_corpus(args, verbose=True)
for key in corpus:
print(key)
doc = corpus[key]
_delete_in_doc(args.anno_id, doc)
save_document(output_dir, key, doc)
pretty_id = anno_id_from_tuple(args.anno_id)
print("Deleted %s" % pretty_id, file=sys.stderr)
announce_output_dir(output_dir)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
if args.stage:
if args.stage != 'unannotated' and not args.annotator:
sys.exit("--annotator is required unless --stage is unannotated")
elif args.stage == 'unannotated' and args.annotator:
sys.exit("--annotator is forbidden if --stage is unannotated")
output_dir = get_output_dir(args, default_overwrite=True)
corpus = read_corpus(args, verbose=True)
for key in corpus:
print(key)
doc = corpus[key]
_delete_in_doc(args.anno_id, doc)
save_document(output_dir, key, doc)
pretty_id = anno_id_from_tuple(args.anno_id)
print("Deleted %s" % pretty_id, file=sys.stderr)
announce_output_dir(output_dir)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
if not args.turns and len(args.dialogues) < 2:
sys.exit("Must specify at least two dialogues")
output_dir = get_output_dir(args, default_overwrite=True)
corpus = read_corpus(args, verbose=True)
if args.turns:
try:
sought = _dialogues_in_turns(corpus, args.turns[0], args.turns[1])
if len(sought) < 2:
sys.exit("Must specify at least two dialogues")
print("Merging dialogues: " +
", ".join(anno_id_from_tuple(x) for x in sought),
file=sys.stderr)
except GlozzException as oops:
sys.exit(str(oops))
else:
sought = args.dialogues
if corpus and not args.no_commit_msg:
key0 = list(corpus)[0]
# compute this before we change things
cmsg = commit_msg(args, corpus, key0, sought)
for k in corpus:
doc = corpus[k]
_merge_dialogues_in_document(sought, doc)
save_document(output_dir, k, doc)
announce_output_dir(output_dir)
if corpus and not args.no_commit_msg:
print("-----8<------")
print(cmsg)
def main(args):
"""
Subcommand main.
You shouldn't need to call this yourself if you're using
`config_argparser`
"""
if not args.turns and len(args.dialogues) < 2:
sys.exit("Must specify at least two dialogues")
output_dir = get_output_dir(args, default_overwrite=True)
corpus = read_corpus(args, verbose=True)
if args.turns:
try:
sought = _dialogues_in_turns(corpus, args.turns[0], args.turns[1])
if len(sought) < 2:
sys.exit("Must specify at least two dialogues")
print("Merging dialogues: " +
", ".join(anno_id_from_tuple(x) for x in sought),
file=sys.stderr)
except GlozzException as oops:
sys.exit(str(oops))
else:
sought = args.dialogues
if corpus and not args.no_commit_msg:
key0 = list(corpus)[0]
# compute this before we change things
cmsg = commit_msg(args, corpus, key0, sought)
for k in corpus:
doc = corpus[k]
_merge_dialogues_in_document(sought, doc)
save_document(output_dir, k, doc)
announce_output_dir(output_dir)
if corpus and not args.no_commit_msg:
print("-----8<------")
print(cmsg)
