start files for lab2

lab2/grammars/Labels.gf

@@ -0,0 +1,112 @@
+fun Q800_Costa_Rica_Country : Country ; 
+fun Q219060_State_of_Palestine_Country : Country ; 
+fun Q37_Lithuania_Country : Country ; 
+fun Q137816_Taiwan_under_Japanese_rule_Country : Country ; 
+fun Q1028_Morocco_Country : Country ; 
+fun Q796_Iraq_Country : Country ; 
+fun Q184_Belarus_Country : Country ; 
+fun Q225_Bosnia_and_Herzegovina_Country : Country ; 
+fun Q20_Norway_Country : Country ; 
+fun Q211_Latvia_Country : Country ; 
+fun Q117_Ghana_Country : Country ; 
+fun Q39_Switzerland_Country : Country ; 
+fun Q159631_Kingdom_of_Württemberg_Country : Country ; 
+fun Q17_Japan_Country : Country ; 
+fun Q189_Iceland_Country : Country ; 
+fun Q221_North_Macedonia_Country : Country ; 
+fun Q9683_Tang_dynasty_Country : Country ; 
+fun Q79_Egypt_Country : Country ; 
+fun Q408_Australia_Country : Country ; 
+fun Q4628_Faroe_Islands_Country : Country ; 
+fun Q145_United_Kingdom_Country : Country ; 
+fun Q214_Slovakia_Country : Country ; 
+fun Q16_Canada_Country : Country ; 
+fun Q924_Tanzania_Country : Country ; 
+fun Q55502_Kingdom_of_Jerusalem_Country : Country ; 
+fun Q183_Germany_Country : Country ; 
+fun Q754_Trinidad_and_Tobago_Country : Country ; 
+fun Q298_Chile_Country : Country ; 
+fun Q41_Greece_Country : Country ; 
+fun Q30623_Manchukuo_Country : Country ; 
+fun Q774_Guatemala_Country : Country ; 
+fun Q836_Myanmar_Country : Country ; 
+fun Q902_Bangladesh_Country : Country ; 
+fun Q215_Slovenia_Country : Country ; 
+fun Q7313_Yuan_dynasty_Country : Country ; 
+fun Q822_Lebanon_Country : Country ; 
+fun Q12548_Holy_Roman_Empire_Country : Country ; 
+fun Q12407080_early_Islamic_period_in_Palestine_Country : Country ; 
+fun Q717_Venezuela_Country : Country ; 
+fun Q31_Belgium_Country : Country ; 
+fun Q794_Iran_Country : Country ; 
+fun Q43_Turkey_Country : Country ; 
+fun Q948_Tunisia_Country : Country ; 
+fun Q258_South_Africa_Country : Country ; 
+fun Q28_Hungary_Country : Country ; 
+fun Q80061_Nobel_Prize_in_Physiology_or_Medicine_Award : Award ; 
+fun Q142_France_Country : Country ; 
+fun Q805_Yemen_Country : Country ; 
+fun Q881_Vietnam_Country : Country ; 
+fun Q7462_Song_dynasty_Country : Country ; 
+fun Q12544_Byzantine_Empire_Country : Country ; 
+fun Q664_New_Zealand_Country : Country ; 
+fun Q33_Finland_Country : Country ; 
+fun Q282428_Mamluk_Sultanate_Country : Country ; 
+fun Q38104_Nobel_Prize_in_Physics_Award : Award ; 
+fun Q9903_Ming_dynasty_Country : Country ; 
+fun Q739_Colombia_Country : Country ; 
+fun Q13426199_Republic_of_China_Country : Country ; 
+fun Q55_Netherlands_Country : Country ; 
+fun Q159_Russia_Country : Country ; 
+fun Q27_Ireland_Country : Country ; 
+fun Q48685_Kingdom_of_Judah_Country : Country ; 
+fun Q810_Jordan_Country : Country ; 
+fun Q36_Poland_Country : Country ; 
+fun Q1014_Liberia_Country : Country ; 
+fun Q38872_Prussia_Country : Country ; 
+fun 'Q574_Timor-Leste_Country' : Country ; 
+fun Q974_Democratic_Republic_of_the_Congo_Country : Country ; 
+fun Q15843470_Roman_Palestine_Country : Country ; 
+fun Q40_Austria_Country : Country ; 
+fun Q928_Philippines_Country : Country ; 
+fun Q148_People's_Republic_of_China_Country : Country ; 
+fun Q35_Denmark_Country : Country ; 
+fun Q954_Zimbabwe_Country : Country ; 
+fun Q216173_Free_City_of_Danzig_Country : Country ; 
+fun Q227_Azerbaijan_Country : Country ; 
+fun Q252_Indonesia_Country : Country ; 
+fun Q801_Israel_Country : Country ; 
+fun Q155_Brazil_Country : Country ; 
+fun Q29_Spain_Country : Country ; 
+fun Q7075820_Occupied_Enemy_Territory_Administration_Country : Country ; 
+fun Q2685298_Romanian_People's_Republic_Country : Country ; 
+fun Q45_Portugal_Country : Country ; 
+fun Q32_Luxembourg_Country : Country ; 
+fun Q115_Ethiopia_Country : Country ; 
+fun Q193714_Mandatory_Palestine_Country : Country ; 
+fun Q34_Sweden_Country : Country ; 
+fun Q262_Algeria_Country : Country ; 
+fun Q37922_Nobel_Prize_in_Literature_Award : Award ; 
+fun Q843_Pakistan_Country : Country ; 
+fun Q35637_Nobel_Peace_Prize_Award : Award ; 
+fun Q1033_Nigeria_Country : Country ; 
+fun Q38_Italy_Country : Country ; 
+fun Q668_India_Country : Country ; 
+fun Q496922_Hasmonean_dynasty_Country : Country ; 
+fun Q212_Ukraine_Country : Country ; 
+fun Q44585_Nobel_Prize_in_Chemistry_Award : Award ; 
+fun Q760_Saint_Lucia_Country : Country ; 
+fun Q414_Argentina_Country : Country ; 
+fun Q218_Romania_Country : Country ; 
+fun Q213_Czech_Republic_Country : Country ; 
+fun Q219_Bulgaria_Country : Country ; 
+fun Q12560_Ottoman_Empire_Country : Country ; 
+fun Q224_Croatia_Country : Country ; 
+fun Q419_Peru_Country : Country ; 
+fun Q1019_Madagascar_Country : Country ; 
+fun Q30_United_States_Country : Country ; 
+fun Q180114_Ayyubid_dynasty_Country : Country ; 
+fun Q8733_Qing_dynasty_Country : Country ; 
+fun Q96_Mexico_Country : Country ; 
+fun Q884_South_Korea_Country : Country ; 
+fun Q114_Kenya_Country : Country ; 

lab2/grammars/LabelsEng.gf

@@ -0,0 +1,112 @@
+lin Q800_Costa_Rica_Country = mkCountry "Costa Rica" ; 
+lin Q219060_State_of_Palestine_Country = mkCountry "State of Palestine" ; 
+lin Q37_Lithuania_Country = mkCountry "Lithuania" ; 
+lin Q137816_Taiwan_under_Japanese_rule_Country = mkCountry "Taiwan under Japanese rule" ; 
+lin Q1028_Morocco_Country = mkCountry "Morocco" ; 
+lin Q796_Iraq_Country = mkCountry "Iraq" ; 
+lin Q184_Belarus_Country = mkCountry "Belarus" ; 
+lin Q225_Bosnia_and_Herzegovina_Country = mkCountry "Bosnia and Herzegovina" ; 
+lin Q20_Norway_Country = mkCountry "Norway" ; 
+lin Q211_Latvia_Country = mkCountry "Latvia" ; 
+lin Q117_Ghana_Country = mkCountry "Ghana" ; 
+lin Q39_Switzerland_Country = mkCountry "Switzerland" ; 
+lin Q159631_Kingdom_of_Württemberg_Country = mkCountry "Kingdom of Württemberg" ; 
+lin Q17_Japan_Country = mkCountry "Japan" ; 
+lin Q189_Iceland_Country = mkCountry "Iceland" ; 
+lin Q221_North_Macedonia_Country = mkCountry "North Macedonia" ; 
+lin Q9683_Tang_dynasty_Country = mkCountry "Tang dynasty" ; 
+lin Q79_Egypt_Country = mkCountry "Egypt" ; 
+lin Q408_Australia_Country = mkCountry "Australia" ; 
+lin Q4628_Faroe_Islands_Country = mkCountry "Faroe Islands" ; 
+lin Q145_United_Kingdom_Country = mkCountry "United Kingdom" ; 
+lin Q214_Slovakia_Country = mkCountry "Slovakia" ; 
+lin Q16_Canada_Country = mkCountry "Canada" ; 
+lin Q924_Tanzania_Country = mkCountry "Tanzania" ; 
+lin Q55502_Kingdom_of_Jerusalem_Country = mkCountry "Kingdom of Jerusalem" ; 
+lin Q183_Germany_Country = mkCountry "Germany" ; 
+lin Q754_Trinidad_and_Tobago_Country = mkCountry "Trinidad and Tobago" ; 
+lin Q298_Chile_Country = mkCountry "Chile" ; 
+lin Q41_Greece_Country = mkCountry "Greece" ; 
+lin Q30623_Manchukuo_Country = mkCountry "Manchukuo" ; 
+lin Q774_Guatemala_Country = mkCountry "Guatemala" ; 
+lin Q836_Myanmar_Country = mkCountry "Myanmar" ; 
+lin Q902_Bangladesh_Country = mkCountry "Bangladesh" ; 
+lin Q215_Slovenia_Country = mkCountry "Slovenia" ; 
+lin Q7313_Yuan_dynasty_Country = mkCountry "Yuan dynasty" ; 
+lin Q822_Lebanon_Country = mkCountry "Lebanon" ; 
+lin Q12548_Holy_Roman_Empire_Country = mkCountry "Holy Roman Empire" ; 
+lin Q12407080_early_Islamic_period_in_Palestine_Country = mkCountry "early Islamic period in Palestine" ; 
+lin Q717_Venezuela_Country = mkCountry "Venezuela" ; 
+lin Q31_Belgium_Country = mkCountry "Belgium" ; 
+lin Q794_Iran_Country = mkCountry "Iran" ; 
+lin Q43_Turkey_Country = mkCountry "Turkey" ; 
+lin Q948_Tunisia_Country = mkCountry "Tunisia" ; 
+lin Q258_South_Africa_Country = mkCountry "South Africa" ; 
+lin Q28_Hungary_Country = mkCountry "Hungary" ; 
+lin Q80061_Nobel_Prize_in_Physiology_or_Medicine_Award = mkAward "Nobel Prize in Physiology or Medicine" ; 
+lin Q142_France_Country = mkCountry "France" ; 
+lin Q805_Yemen_Country = mkCountry "Yemen" ; 
+lin Q881_Vietnam_Country = mkCountry "Vietnam" ; 
+lin Q7462_Song_dynasty_Country = mkCountry "Song dynasty" ; 
+lin Q12544_Byzantine_Empire_Country = mkCountry "Byzantine Empire" ; 
+lin Q664_New_Zealand_Country = mkCountry "New Zealand" ; 
+lin Q33_Finland_Country = mkCountry "Finland" ; 
+lin Q282428_Mamluk_Sultanate_Country = mkCountry "Mamluk Sultanate" ; 
+lin Q38104_Nobel_Prize_in_Physics_Award = mkAward "Nobel Prize in Physics" ; 
+lin Q9903_Ming_dynasty_Country = mkCountry "Ming dynasty" ; 
+lin Q739_Colombia_Country = mkCountry "Colombia" ; 
+lin Q13426199_Republic_of_China_Country = mkCountry "Republic of China" ; 
+lin Q55_Netherlands_Country = mkCountry "Netherlands" ; 
+lin Q159_Russia_Country = mkCountry "Russia" ; 
+lin Q27_Ireland_Country = mkCountry "Ireland" ; 
+lin Q48685_Kingdom_of_Judah_Country = mkCountry "Kingdom of Judah" ; 
+lin Q810_Jordan_Country = mkCountry "Jordan" ; 
+lin Q36_Poland_Country = mkCountry "Poland" ; 
+lin Q1014_Liberia_Country = mkCountry "Liberia" ; 
+lin Q38872_Prussia_Country = mkCountry "Prussia" ; 
+lin 'Q574_Timor-Leste_Country' = mkCountry "Timor-Leste" ; 
+lin Q974_Democratic_Republic_of_the_Congo_Country = mkCountry "Democratic Republic of the Congo" ; 
+lin Q15843470_Roman_Palestine_Country = mkCountry "Roman Palestine" ; 
+lin Q40_Austria_Country = mkCountry "Austria" ; 
+lin Q928_Philippines_Country = mkCountry "Philippines" ; 
+lin Q148_People's_Republic_of_China_Country = mkCountry "People's Republic of China" ; 
+lin Q35_Denmark_Country = mkCountry "Denmark" ; 
+lin Q954_Zimbabwe_Country = mkCountry "Zimbabwe" ; 
+lin Q216173_Free_City_of_Danzig_Country = mkCountry "Free City of Danzig" ; 
+lin Q227_Azerbaijan_Country = mkCountry "Azerbaijan" ; 
+lin Q252_Indonesia_Country = mkCountry "Indonesia" ; 
+lin Q801_Israel_Country = mkCountry "Israel" ; 
+lin Q155_Brazil_Country = mkCountry "Brazil" ; 
+lin Q29_Spain_Country = mkCountry "Spain" ; 
+lin Q7075820_Occupied_Enemy_Territory_Administration_Country = mkCountry "Occupied Enemy Territory Administration" ; 
+lin Q2685298_Romanian_People's_Republic_Country = mkCountry "Romanian People's Republic" ; 
+lin Q45_Portugal_Country = mkCountry "Portugal" ; 
+lin Q32_Luxembourg_Country = mkCountry "Luxembourg" ; 
+lin Q115_Ethiopia_Country = mkCountry "Ethiopia" ; 
+lin Q193714_Mandatory_Palestine_Country = mkCountry "Mandatory Palestine" ; 
+lin Q34_Sweden_Country = mkCountry "Sweden" ; 
+lin Q262_Algeria_Country = mkCountry "Algeria" ; 
+lin Q37922_Nobel_Prize_in_Literature_Award = mkAward "Nobel Prize in Literature" ; 
+lin Q843_Pakistan_Country = mkCountry "Pakistan" ; 
+lin Q35637_Nobel_Peace_Prize_Award = mkAward "Nobel Peace Prize" ; 
+lin Q1033_Nigeria_Country = mkCountry "Nigeria" ; 
+lin Q38_Italy_Country = mkCountry "Italy" ; 
+lin Q668_India_Country = mkCountry "India" ; 
+lin Q496922_Hasmonean_dynasty_Country = mkCountry "Hasmonean dynasty" ; 
+lin Q212_Ukraine_Country = mkCountry "Ukraine" ; 
+lin Q44585_Nobel_Prize_in_Chemistry_Award = mkAward "Nobel Prize in Chemistry" ; 
+lin Q760_Saint_Lucia_Country = mkCountry "Saint Lucia" ; 
+lin Q414_Argentina_Country = mkCountry "Argentina" ; 
+lin Q218_Romania_Country = mkCountry "Romania" ; 
+lin Q213_Czech_Republic_Country = mkCountry "Czech Republic" ; 
+lin Q219_Bulgaria_Country = mkCountry "Bulgaria" ; 
+lin Q12560_Ottoman_Empire_Country = mkCountry "Ottoman Empire" ; 
+lin Q224_Croatia_Country = mkCountry "Croatia" ; 
+lin Q419_Peru_Country = mkCountry "Peru" ; 
+lin Q1019_Madagascar_Country = mkCountry "Madagascar" ; 
+lin Q30_United_States_Country = mkCountry "United States" ; 
+lin Q180114_Ayyubid_dynasty_Country = mkCountry "Ayyubid dynasty" ; 
+lin Q8733_Qing_dynasty_Country = mkCountry "Qing dynasty" ; 
+lin Q96_Mexico_Country = mkCountry "Mexico" ; 
+lin Q884_South_Korea_Country = mkCountry "South Korea" ; 
+lin Q114_Kenya_Country = mkCountry "Kenya" ; 

lab2/scripts/analyse_nobel.py

@@ -0,0 +1,49 @@
+import json
+
+# query: https://w.wiki/3tEM
+
+DATA_FILE = 'query.json'
+
+with open(DATA_FILE) as file:
+    data = json.load(file)
+
+print(data[0])
+
+awards = {(d['award'], d['awardLabel']) for d in data}
+
+#print(awards)
+#print(len(awards))
+
+countries = {(d['country'], d['countryLabel']) for d in data}
+
+#print(countries)
+#print(len(countries))
+
+#print(data[0].keys())
+
+def pronoun(d):
+    sex = d.get('sexLabel', 'other')
+    if sex == 'female':
+        return 'she'
+    elif sex == 'male':
+        return 'he'
+    else:
+        return 'they'
+
+
+def year(date):
+    return date[:4]
+
+
+def person_descr(d):
+    died = f"{d['personLabel']} died {year(d['deathDate'])}" if 'deathDate' in d else ''
+    return (
+        f"{d['personLabel']} from {d['countryLabel']} was born in {year(d['birthDate'])}. " +
+        f"{pronoun(d)} got {d['awardLabel']} in {year(d['date'])}." +
+        died
+        )
+
+for d in data:
+    print(person_descr(d))
+
+

lab2/scripts/find_labels.py

@@ -0,0 +1,101 @@
+# https://www.wikidata.org/wiki/Special:EntityData/Q18644475.json
+
+import urllib.request
+import json
+import sys
+import ssl
+from gf_utils import *
+
+"""
+To collect labels from query.json (Wikidata query result) and extract grammars:
+python3 find_labels.py init >labels.jsonl
+python3 find_labels.py abstract >Labels.gf
+python3 find_labels.py en >LabelsEng.gf
+"""
+
+WIKIDATA_FILE = '../data/query.json'
+WIKIDATA_PREFIX = 'http://www.wikidata.org/entity/'
+WIKIDATA_URL_PREFIX = 'http://www.wikidata.org/wiki/Special:EntityData/'
+NOBEL_FIELDS = ['award', 'country']
+LABEL_FILE = '../data/labels.jsonl'
+
+USAGE = 'usage: find_labels.py (init | abstract | en | sv | fi | ...)'
+
+if sys.argv[1:]:
+    MODE = sys.argv[1]
+else:
+    print(USAGE)
+    
+
+# qids given in the data file
+def get_wikidata_qids(jsonfile, fields):
+    qids = set()
+    with open(jsonfile) as file:
+        data = json.load(file)
+        for d in data:
+            for f in fields:
+                if f in d:
+                    qids.add((f, d[f][len(WIKIDATA_PREFIX):]))
+    return qids
+
+
+qids = get_wikidata_qids(WIKIDATA_FILE, NOBEL_FIELDS)
+
+if __name__ == '__mainz__':
+    for qid in qids:
+        print(qid)
+
+# get all wikidata for each qid
+# use this only once, because it is slow
+def get_wikidata_json(qids):
+    context = ssl._create_unverified_context()
+    for field, qid in qids:
+        try:
+            with urllib.request.urlopen(WIKIDATA_URL_PREFIX + qid +'.json', context=context) as url:
+                data = json.load(url)
+                yield (field, qid, data)
+        except Exception as error:
+            pass
+
+# extract the labels, redirect to LABEL_FILE (only once)
+def get_wikidata_labels(data, languages=None):
+    for field, qid, dict in data:
+        entities = dict.get('entities', {'foo': {}})
+        entity = list(entities.values())[0]
+        entitylabels = entity.get('labels', {})
+        entitylabels = {val['language']: val['value']
+                        for val in entitylabels.values()
+                        if (languages is None) or
+                           (val['language'] in languages)}
+        entitylabels['field'] = field
+        print(json.dumps({qid: entitylabels}, ensure_ascii=False))
+
+
+# {"Q800": {"tg": "Коста Рика", "sk": "Kostarika", ... "field": <field>}}
+def extract_labels(labeldata, mode):
+    for entry in data:
+        qid, labels = list(entry.items())[0]
+        eng = labels.get('en', 'X')
+        cat = labels['field'].capitalize()
+        fun = mk_fun_from_strs([qid, eng, cat])
+        if mode == 'abstract':
+            print(mk_fun_rule(fun, cat))
+        else:
+            lin = labels.get(mode, labels.get('en', 'X'))
+            oper = 'mk' + cat
+            print(mk_lin_rule(fun, mk_lin(oper, [lin], [])))
+
+
+if MODE == 'init':
+    # do this only once, redirect to labels.jsonl
+    data = get_wikidata_json(list(qids))
+    get_wikidata_labels(data)
+else:
+    # do this once for abs and for every language you want
+    with open(LABEL_FILE) as file:
+        data = [json.loads(line) for line in file]
+        extract_labels(data, MODE)
+
+
+
+        

lab2/scripts/find_labels.py~

@@ -0,0 +1,101 @@
+# https://www.wikidata.org/wiki/Special:EntityData/Q18644475.json
+
+import urllib.request
+import json
+import sys
+import ssl
+from gf_utils import *
+
+"""
+To collect labels from query.json (Wikidata query result) and extract grammars:
+python3 find_labels.py init >labels.jsonl
+python3 find_labels.py abstract >Labels.gf
+python3 find_labels.py en >LabelsEng.gf
+"""
+
+WIKIDATA_FILE = 'query.json'
+WIKIDATA_PREFIX = 'http://www.wikidata.org/entity/'
+WIKIDATA_URL_PREFIX = 'http://www.wikidata.org/wiki/Special:EntityData/'
+NOBEL_FIELDS = ['award', 'country']
+LABEL_FILE = 'labels.jsonl'
+
+USAGE = 'usage: find_labels.py (init | abstract | en | sv | fi | ...)'
+
+if sys.argv[1:]:
+    MODE = sys.argv[1]
+else:
+    print(USAGE)
+    
+
+# qids given in the data file
+def get_wikidata_qids(jsonfile, fields):
+    qids = set()
+    with open(jsonfile) as file:
+        data = json.load(file)
+        for d in data:
+            for f in fields:
+                if f in d:
+                    qids.add((f, d[f][len(WIKIDATA_PREFIX):]))
+    return qids
+
+
+qids = get_wikidata_qids(WIKIDATA_FILE, NOBEL_FIELDS)
+
+if __name__ == '__mainz__':
+    for qid in qids:
+        print(qid)
+
+# get all wikidata for each qid
+# use this only once, because it is slow
+def get_wikidata_json(qids):
+    context = ssl._create_unverified_context()
+    for field, qid in qids:
+        try:
+            with urllib.request.urlopen(WIKIDATA_URL_PREFIX + qid +'.json', context=context) as url:
+                data = json.load(url)
+                yield (field, qid, data)
+        except Exception as error:
+            pass
+
+# extract the labels, redirect to LABEL_FILE (only once)
+def get_wikidata_labels(data, languages=None):
+    for field, qid, dict in data:
+        entities = dict.get('entities', {'foo': {}})
+        entity = list(entities.values())[0]
+        entitylabels = entity.get('labels', {})
+        entitylabels = {val['language']: val['value']
+                        for val in entitylabels.values()
+                        if (languages is None) or
+                           (val['language'] in languages)}
+        entitylabels['field'] = field
+        print(json.dumps({qid: entitylabels}, ensure_ascii=False))
+
+
+# {"Q800": {"tg": "Коста Рика", "sk": "Kostarika", ... "field": <field>}}
+def extract_labels(labeldata, mode):
+    for entry in data:
+        qid, labels = list(entry.items())[0]
+        eng = labels.get('en', 'X')
+        cat = labels['field'].capitalize()
+        fun = mk_fun_from_strs([qid, eng, cat])
+        if mode == 'abstract':
+            print(mk_fun_rule(fun, cat))
+        else:
+            lin = labels.get(mode, labels.get('en', 'X'))
+            oper = 'mk' + cat
+            print(mk_lin_rule(fun, mk_lin(oper, [lin], [])))
+
+
+if MODE == 'init':
+    # do this only once, redirect to labels.jsonl
+    data = get_wikidata_json(list(qids))
+    get_wikidata_labels(data)
+else:
+    # do this once for abs and for every language you want
+    with open(LABEL_FILE) as file:
+        data = [json.loads(line) for line in file]
+        extract_labels(data, MODE)
+
+
+
+        

lab2/scripts/gf_utils.py

@@ -0,0 +1,48 @@
+# import pgf
+
+def mk_fun(s):
+    s = '_'.join(s.split())  # spaces replaced by underscores
+    
+    if (s[0].isalpha() and
+        all(ord(c)<256 and (c.isdigit() or c.isalpha() or c in "_'")
+            for c in s)):  # test if legal GF identifier
+        return s
+    else:
+        return "'" + s.replace("'", "\\'") + "'"  # if not, single quotes make it legal
+
+
+def mk_fun_from_strs(ss):
+    return mk_fun('_'.join(ss))
+
+
+def quote(s):
+    return '"' + s + '"'
+
+
+def app(fun, args):
+    return ' '.join([fun, quote(args[0])] + args[1:])
+
+def empty_variants():
+    return 'variants {}'
+
+def mk_lin(oper, words, params):
+    return ' '.join([oper] + [quote(w) for w in words] + params) 
+
+
+def mk_cat_rule(cat):
+    return ' '.join(['cat', cat, ';\n'])
+
+
+def mk_fun_rule(fun, cat, comment=None):
+    co = '--' + comment if comment else ''
+    return ' '.join(['fun', fun, ':', cat, ';', co])
+
+
+def mk_lin_rule(fun, lin, comment=None):
+    co = '--' + comment if comment else ''
+    return ' '.join(['lin', fun, '=', lin, ';', co])
+
+
+def mk_lincat_rule(cat, lin):
+    return ' '.join(['lincat', cat, '=', lin, ';'])
+