updated parser Extract dates()

- updated parser Extract dates() based on user feedback
- added test cases for update
- updated README.md to link to StellA
- increased version number to 1.2.1

Author: sedthh

README.md

@@ -14,7 +14,8 @@ Instead of being an all purpose NLP tool, **Lara** was created to fit the [quirk
 
 ## About Lara
 
-Here is a short list of things you can easily do with **Lara** in Hungarian. For full documentation and further examples, **CHECK OUT [THE WIKI](https://github.com/sedthh/lara-hungarian-nlp/wiki)**.
+Here is a short list of things you can easily do with **Lara** in Hungarian. For full documentation and further examples, **CHECK OUT [THE WIKI](https://github.com/sedthh/lara-hungarian-nlp/wiki)**. A complete case study on [how to make ChatBots and Virtual Assistants](https://chatbotsmagazine.com/a-complete-case-study-for-developing-smart-assistants-79316be80e89) in foreign languages is also available. 
+
 
 #### Find intents
 

lara/__init__.py

@@ -3,7 +3,7 @@
 # Lara - Lingusitic Aim Recognizer API
 
 __all__				= 'nlp','parser','stemmer','entities'
-__version__ 		= '1.2.0'
+__version__ 		= '1.2.1'
 __version_info__	= tuple(int(num) for num in __version__.split('.'))
 
 import lara.nlp

lara/parser.py

@@ -212,10 +212,9 @@ def _generate(self, item):
 			item['pattern']			= r''+item['stem']+item['affix']
 			item['typo_pattern']	= r''+item['typo_stem']+item['typo_affix']
 		else:
-			
-			item['pattern']		= r'(?:'+re.escape(item['stem'])+r'{1,2}'+item['affix']+r')'
+			item['pattern']			= r'(?:'+re.escape(item['stem'])+r'{1,2}'+item['affix']+r')'
 			scramble				= self._scramble(item['typo_stem'], (item['wordclass'] == 'adjective'))
-			item['typo_pattern']= r'(?:'+scramble+item['typo_affix']+')'
+			item['typo_pattern']	= r'(?:'+scramble+item['typo_affix']+')'
 			if not item['ignorecase']:
 				item['pattern']			= r'(?s)'+item['pattern']
 				item['typo_pattern']	= r'(?s)'+item['typo_pattern']
@@ -232,7 +231,7 @@ def _generate(self, item):
 
 		item['pattern']			= item['prefix']+item['pattern']	
 		item['typo_pattern']	= item['typo_prefix']+item['typo_pattern']
-
+				
 		return item
 
 	# generate scrambled keywords
@@ -612,10 +611,13 @@ def phone_numbers(self,normalize=True,convert=True):
 		return results
 
 	# extract list of common Hungarian date formats from text without further processing them
-	def dates(self,normalize=True,convert=True):
+	def dates(self,normalize=True,convert=True,current=False):
 		results	= []
 		if self.text:
-			now = datetime.datetime.now()
+			if current:
+				now		= datetime.datetime.strptime(current,"%Y-%m-%d")
+			else:
+				now 	= datetime.datetime.now()
 			matches	= _re.findall(r'((\d{2})?(\d{2}([\\\/\.\-]\s?|\s))([eé]v\s?)?(\d{1,2}([\\\/\.\-]\s?|\s)(h[oó](nap)?\s?)?)?(\d{1,2}))\W*([aáeéio][ikn]|nap)?\b', re.IGNORECASE, self.text)
 			for item in matches:
 				match	= re.sub('([eé]v|h[oó]|nap)', '', item[0])
@@ -639,7 +641,7 @@ def dates(self,normalize=True,convert=True):
 							results.append(str(now.year)+'-'+parts[0].zfill(2)+'-'+parts[1].zfill(2))
 					else:
 						results.append(item[0])
-			matches	= _re.findall(r'((\d{2}(\d{2})?\W{1,2})?((jan|feb|m[aá]r|[aá]pr|m[aá]j|j[uú][nl]|aug|sz?ep|okt|nov|dec)\w{0,10}\W{1,2}|[ivx]{1,4}\W{0,2})(h[aoó][nv]?\w{0,7}\W{1,2})?(\d{1,2})?\W?\w*)\b', re.IGNORECASE, self.ntext if convert else self.text)
+			matches	= _re.findall(r'\b((\d{2}(\d{2})?\W{1,2})?((jan|feb|m[aá]r|[aá]pr|m[aá]j|j[uú][nl]|aug|sz?ep|okt|nov|dec)\w{0,10}\W{1,2}|[ivx]{1,4}\W{0,2})(h[aoó][nv]?\w{0,7}\W{1,2})?(\d{1,2})?\W?\w*)\b', re.IGNORECASE, self.ntext if convert else self.text)
 			for item in matches:
 				match	= item[0].lower()
 				year	= ''
@@ -713,11 +715,22 @@ def dates(self,normalize=True,convert=True):
 							month	= str(11-len(roman)).zfill(2)
 					else:
 						month	= str(len(roman)).zfill(2)
-				if month and month!='00':
+				if month and month!='00' and len(day)<=2:
 					if normalize:
 						results.append(year+'-'+month+'-'+day)
 					else:
 						results.append(item[0])
+			if not results:
+				matches	= _re.findall(r'\b(?<!\-)([0123]?\d)[\.\-aáeéint]+(?![kloópr])', re.IGNORECASE, self.ntext if convert else self.text)
+				for item in matches:
+					if int(item)<=31:
+						if normalize:
+							year	= str(now.year)
+							month	= str(now.month).zfill(2)
+							day		= item.zfill(2)
+							results.append(year+'-'+month+'-'+day)
+						else:
+							results.append(item)	
 		return results
 
 	# extract times like 12:00 or délután 4

tests/test_parser.py

@@ -636,23 +636,31 @@ def test_parser_extract_convert_numbers(info):
 		{	
 			"text"		: "Hívj fel ezen a számon 2018 IV. huszadikán mondjuk délután nyolc perccel háromnegyed kettő előtt!",
 			"function"	: "dates",
-			"args"		: [True],
+			"args"		: [True,True,'2018-04-01'],
 			"result"	: ["2018-04-20"]
 		}
 	),
 	(
 		{
 			"text"		: "18/01/09 vagy 18-01-09 vagy 2018. 01. 09. vagy 2018. 01. 09-én vagy 2018 VII 20. és így 2018 január 20-án",
 			"function"	: "dates",
-			"args"		: [False],
+			"args"		: [False,True,'2018-04-01'],
 			"result"		: ["18/01/09","18-01-09","2018. 01. 09","2018. 01. 09","2018 VII 20","2018 január 20-án"]
 		}
 	),
+	(
+		{
+			"text"		: "találkozzunk 20-án valamikor vagy 21-én?",
+			"function"	: "dates",
+			"args"		: [True,True,'2018-04-01'],
+			"result"		: ["2018-04-20","2018-04-21"]
+		}
+	),
 	(
 		{
 			"text"		: "18/01/09 vagy 18-01-09 vagy 2019. 01. 09. vagy 2018. 01. 09-én vagy 2018 VII 20. és így 2018 január 20-án",
 			"function"	: "dates",
-			"args"		: [True],
+			"args"		: [True,True,'2018-04-01'],
 			"result"		: ["2018-01-09","2018-01-09","2019-01-09","2018-01-09","2018-07-20","2018-01-20"]
 		}
 	),