Compare commits
7 Commits
mt-initial
...
lk-colloca
| Author | SHA1 | Date | |
|---|---|---|---|
| a7f85accce | |||
| 3926d1a199 | |||
|
|
b764d59602 | ||
| b17510ffe7 | |||
| 5396868601 | |||
| 8cb63eaa1d | |||
| 6e1f3bd16a |
2
.gitignore
vendored
2
.gitignore
vendored
@@ -10,5 +10,5 @@ build/*
|
||||
# using kdev4, works fairly nicely!
|
||||
.kdev4
|
||||
**/*.kdev4
|
||||
|
||||
remote
|
||||
|
||||
|
||||
@@ -1,5 +1,3 @@
|
||||
.PHONY: all release build transcrypt copy $(RES_LESS_FILE) $(RES_PY_FILE) $(RES_HTML_FILE) $(RES_XML_FILE)
|
||||
|
||||
SRC_FOLDER=$(CURDIR)/../src
|
||||
RES_FOLDER=$(CURDIR)/../res
|
||||
|
||||
@@ -20,6 +18,7 @@ RES_PY_FILE=$(SRC_FOLDER)/$(PY_FILE)
|
||||
RES_HTML_FILE=$(RES_FOLDER)/$(HTML_FILE)
|
||||
RES_XML_FILE=$(RES_FOLDER)/$(XML_FILE)
|
||||
|
||||
.PHONY: all release build transcrypt copy $(RES_LESS_FILE) $(RES_PY_FILE) $(RES_HTML_FILE) $(RES_XML_FILE) $(CONV_FILE)
|
||||
.DEFAULT_GOAL := all
|
||||
|
||||
all: BR_FLAGS := $(BR_FLAGS) --debug
|
||||
|
||||
@@ -3,15 +3,19 @@
|
||||
# fail if any command fails
|
||||
set -e
|
||||
|
||||
OUT_FILE=conversions.csv
|
||||
|
||||
if [[ -z "${API_KEY}" ]]; then
|
||||
echo "API_KEY not defined, should be api key to CJVT's gitea"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
OUT_FILE=conversions.csv
|
||||
TEMP_FILE=$(mktemp)
|
||||
if [ -f $OUT_FILE ]; then
|
||||
echo "file for conversion of SKE collocations: '$OUT_FILE' already exists, skipping."
|
||||
exit 0
|
||||
fi
|
||||
|
||||
rm -r $OUT_FILE
|
||||
TEMP_FILE=$(mktemp)
|
||||
|
||||
curl -s "https://gitea.cjvt.si/api/v1/repos/generic/data_admin/contents/resources/structure_conversions.csv?token=$API_KEY" -o $TEMP_FILE
|
||||
echo "wc: $(wc $TEMP_FILE)"
|
||||
|
||||
@@ -34,12 +34,23 @@ def build_structure_conversions():
|
||||
structure_conversions.append((__new__(RegExp(vfrom, 'u')), vto_name, vto_id))
|
||||
|
||||
|
||||
def convert_structure(structure):
|
||||
def convert_structure(structure, type):
|
||||
if structure_conversions is None:
|
||||
build_structure_conversions()
|
||||
|
||||
for vfrom, vto_name, vto_id in structure_conversions:
|
||||
match = structure.match(vfrom)
|
||||
# fix for ids 65, 66, 67 which instead matched with 64
|
||||
if match and vto_id == '64' and '-s' in type:
|
||||
vto_name = 's0-vp-s0'
|
||||
vto_id = '65'
|
||||
elif match and vto_id == '64' and '-g' in type:
|
||||
vto_name = 'gg-vp-gg'
|
||||
vto_id = '66'
|
||||
elif match and vto_id == '64' and '-r' in type:
|
||||
vto_name = 'r-vp-r'
|
||||
vto_id = '67'
|
||||
|
||||
if match:
|
||||
# we need to remove replace alias here as we want to use javascript's one
|
||||
__pragma__('noalias', 'replace')
|
||||
@@ -49,4 +60,4 @@ def convert_structure(structure):
|
||||
return result, vto_id
|
||||
|
||||
window.console.log("Unknown structure: ", structure)
|
||||
return None
|
||||
return 'N/A', '/'
|
||||
|
||||
@@ -26,6 +26,7 @@ class SkeExample:
|
||||
self.mid = ""
|
||||
self.s_id = ""
|
||||
self.gf2_good = None
|
||||
self.gf2_check = False
|
||||
|
||||
@staticmethod
|
||||
def fromLine(line):
|
||||
@@ -58,7 +59,7 @@ class SkeCollocation:
|
||||
def __init__(self, data):
|
||||
self.word = data.word
|
||||
self.frequency = data.count
|
||||
self.structure_name, self.structure_id = convert_structure(data.gramrel)
|
||||
self.structure_name, self.structure_id = convert_structure(data.gramrel, data.lempos)
|
||||
|
||||
self.other = {"score": data.score, "cm": data.cm}
|
||||
|
||||
@@ -140,7 +141,7 @@ class SkeModal(ClickMessage):
|
||||
elif type(data) is list:
|
||||
window.console.log(data)
|
||||
# check if gf2 examples are loaded or not
|
||||
if data[0].gf2_good is None:
|
||||
if not data[0].gf2_check and type(data[0]) is SkeExample:
|
||||
# we get the data, we have to match it with available data on our gf2 examples API
|
||||
match_gf2_examples(data, page_num, search_term, ske_index)
|
||||
|
||||
@@ -161,12 +162,18 @@ class SkeModalGf2Update(SkeModal):
|
||||
example.gf_good = False
|
||||
data_dict[example.s_id] = example
|
||||
|
||||
bad_response = dict(response_data["bad"])
|
||||
for gf_sid, gf_data in bad_response.items():
|
||||
data_dict[gf_sid].gf2_good = None
|
||||
data_dict[gf_sid].gf2_check = True
|
||||
|
||||
good_response = dict(response_data["good"])
|
||||
for gf_sid, gf_data in good_response.items():
|
||||
data_dict[gf_sid].left = gf_data.left
|
||||
data_dict[gf_sid].mid = gf_data.mid
|
||||
data_dict[gf_sid].right = gf_data.right
|
||||
data_dict[gf_sid].gf2_good = True
|
||||
data_dict[gf_sid].gf2_check = True
|
||||
|
||||
# changed data_dict, now we can redraw!
|
||||
# just let it do its thing in update_model
|
||||
@@ -217,10 +224,12 @@ class SkeInsert(DataChgClickMessage):
|
||||
console.log("You really should not be here, my lady")
|
||||
continue
|
||||
|
||||
model.reset()
|
||||
|
||||
def _as_corpus_example(self, example):
|
||||
new_example = Example()
|
||||
new_example.inner = CorpusExample()
|
||||
new_example.inner.other_attributes["exampleId"] = example.s_id
|
||||
new_example.inner.other_attributes["example_id"] = example.s_id
|
||||
new_example.inner.cluster = ExampleClusters.first_empty_cluster()
|
||||
|
||||
lex_left = ComponentLexeme()
|
||||
@@ -229,7 +238,7 @@ class SkeInsert(DataChgClickMessage):
|
||||
|
||||
lex_mid = ComponentLexeme()
|
||||
lex_mid.text = example["mid"]
|
||||
lex_mid.role = "headword"
|
||||
lex_mid.role = "collocation"
|
||||
|
||||
lex_right = ComponentLexeme()
|
||||
lex_right.text = example["right"]
|
||||
@@ -243,6 +252,7 @@ class SkeInsert(DataChgClickMessage):
|
||||
new_collocation.inner = MultiwordExample()
|
||||
|
||||
new_collocation.inner.other_attributes["structure_id"] = example.structure_id
|
||||
new_collocation.inner.other_attributes["structureName"] = example.structure_name
|
||||
new_collocation.inner.other_attributes["logDice"] = example.other["score"]
|
||||
new_collocation.inner.other_attributes["frequency"] = example.frequency
|
||||
new_collocation.inner.type = "collocation"
|
||||
@@ -253,7 +263,7 @@ class SkeInsert(DataChgClickMessage):
|
||||
|
||||
lex_mid = ComponentLexeme()
|
||||
lex_mid.text = example.word
|
||||
lex_mid.role = "headword"
|
||||
lex_mid.role = "collocation"
|
||||
|
||||
lex_right = ComponentLexeme()
|
||||
lex_right.text = ""
|
||||
|
||||
Reference in New Issue
Block a user