From 6267fc195bb078554850eae18147105781198424 Mon Sep 17 00:00:00 2001
From: emphasize <swen_g@t-online.de>
Date: Thu, 28 Jan 2021 13:47:53 +0100
Subject: [PATCH 01/13] fix extract_datetime offset-aware/naive bug

---
 lingua_franca/lang/parse_de.py | 95 +++++++++-------------------------
 1 file changed, 25 insertions(+), 70 deletions(-)

diff --git a/lingua_franca/lang/parse_de.py b/lingua_franca/lang/parse_de.py
index 5f7f7b32..cc193e02 100644
--- a/lingua_franca/lang/parse_de.py
+++ b/lingua_franca/lang/parse_de.py
@@ -20,64 +20,7 @@
     extract_numbers_generic, Normalizer
 from lingua_franca.lang.common_data_de import _DE_NUMBERS
 from lingua_franca.lang.format_de import pronounce_number_de
-
-de_numbers = {
-    'null': 0,
-    'ein': 1,
-    'eins': 1,
-    'eine': 1,
-    'einer': 1,
-    'einem': 1,
-    'einen': 1,
-    'eines': 1,
-    'zwei': 2,
-    'drei': 3,
-    'vier': 4,
-    'fünf': 5,
-    'sechs': 6,
-    'sieben': 7,
-    'acht': 8,
-    'neun': 9,
-    'zehn': 10,
-    'elf': 11,
-    'zwölf': 12,
-    'dreizehn': 13,
-    'vierzehn': 14,
-    'fünfzehn': 15,
-    'sechzehn': 16,
-    'siebzehn': 17,
-    'achtzehn': 18,
-    'neunzehn': 19,
-    'zwanzig': 20,
-    'einundzwanzig': 21,
-    'zweiundzwanzig': 22,
-    'dreiundzwanzig': 23,
-    'vierundzwanzig': 24,
-    'fünfundzwanzig': 25,
-    'sechsundzwanzig': 26,
-    'siebenundzwanzig': 27,
-    'achtundzwanzig': 28,
-    'neunundzwanzig': 29,
-    'dreißig': 30,
-    'einunddreißig': 31,
-    'vierzig': 40,
-    'fünfzig': 50,
-    'sechzig': 60,
-    'siebzig': 70,
-    'achtzig': 80,
-    'neunzig': 90,
-    'hundert': 100,
-    'zweihundert': 200,
-    'dreihundert': 300,
-    'vierhundert': 400,
-    'fünfhundert': 500,
-    'sechshundert': 600,
-    'siebenhundert': 700,
-    'achthundert': 800,
-    'neunhundert': 900,
-    'tausend': 1000,
-    'million': 1000000
-}
+from mycroft.util.time import default_timezone
 
 # TODO: short_scale and ordinals don't do anything here.
 # The parameters are present in the function signature for API compatibility
@@ -405,7 +348,8 @@ def date_found():
                 dayOffset -= 7
                 used += 1
                 start -= 1
-                # parse 15 of July, June 20th, Feb 18, 19 of February
+
+        # parse 15 Mai, Mai der 20ste, Dez 18
         elif word in months or word in monthsShort and not fromFlag:
             try:
                 m = months.index(word)
@@ -413,10 +357,13 @@ def date_found():
                 m = monthsShort.index(word)
             used += 1
             datestr = months[m]
+            #commonly spoken : 15(.=gets replaced) Mai <Year>/<time>
             if wordPrev and (wordPrev[0].isdigit() or
-                             (wordPrev == "of" and wordPrevPrev[0].isdigit())):
-                if wordPrev == "of" and wordPrevPrev[0].isdigit():
-                    datestr += " " + words[idx - 2]
+                             (((wordNext == "der") or (wordNext == "den")) and
+                             wordNextNext[0].isdigit())):
+                #Mai der fünfte(5)
+                if ((wordNext == "der") or (wordNext == "den")) and wordNextNext[0].isdigit():
+                    datestr += " " + words[idx + 2]
                     used += 1
                     start -= 1
                 else:
@@ -424,12 +371,20 @@ def date_found():
                 start -= 1
                 used += 1
                 if wordNext and wordNext[0].isdigit():
-                    datestr += " " + wordNext
-                    used += 1
-                    hasYear = True
-                else:
-                    hasYear = False
-
+                    #normally the time comes in as ##:## and therefor
+                    #would break int(wordnext)
+                    if ':' in wordNext:
+                        tmp_word = wordNext.split(':')
+                        wordNext = tmp_word[0]
+                    #determine if wordnext is year data; eg 3 Januar 10:10 uhr
+                    #10:10 / 10 would be seen as such, leaving us with no time data
+                    if int(wordNext) > 2000:
+                        datestr += " " + wordNext
+                        used += 1
+                        hasYear = True
+                    else:
+                        hasYear = False
+            # Mai <Year>
             elif wordNext and wordNext[0].isdigit():
                 datestr += " " + wordNext
                 used += 1
@@ -439,9 +394,9 @@ def date_found():
                     hasYear = True
                 else:
                     hasYear = False
+
         # parse 5 days from tomorrow, 10 weeks from next thursday,
         # 2 months from July
-
         if (
                 word == "von" or word == "nach" or word == "ab") and wordNext \
                 in validFollowups:
@@ -832,7 +787,7 @@ def date_found():
         for idx, en_month in enumerate(en_monthsShort):
             datestr = datestr.replace(monthsShort[idx], en_month)
 
-        temp = datetime.strptime(datestr, "%B %d")
+        temp = datetime.strptime(datestr, "%B %d").replace(tzinfo=default_timezone())
         if not hasYear:
             temp = temp.replace(year=extractedDate.year)
             if extractedDate < temp:

From bafc7710686fc6166c96124bddd693dba16a2742 Mon Sep 17 00:00:00 2001
From: emphasize <swen_g@t-online.de>
Date: Thu, 28 Jan 2021 14:26:01 +0100
Subject: [PATCH 02/13] lower wordnext check to > 60

---
 lingua_franca/lang/parse_de.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lingua_franca/lang/parse_de.py b/lingua_franca/lang/parse_de.py
index cc193e02..37762e68 100644
--- a/lingua_franca/lang/parse_de.py
+++ b/lingua_franca/lang/parse_de.py
@@ -378,7 +378,7 @@ def date_found():
                         wordNext = tmp_word[0]
                     #determine if wordnext is year data; eg 3 Januar 10:10 uhr
                     #10:10 / 10 would be seen as such, leaving us with no time data
-                    if int(wordNext) > 2000:
+                    if int(wordNext) > 60:
                         datestr += " " + wordNext
                         used += 1
                         hasYear = True

From 18712f68cbe295c3738adc9820a3276a82b19c93 Mon Sep 17 00:00:00 2001
From: emphasize <swen_g@t-online.de>
Date: Thu, 28 Jan 2021 15:54:34 +0100
Subject: [PATCH 03/13] extract tzinfo from extractedDate instead of
 default_timezone()

---
 lingua_franca/lang/parse_de.py | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/lingua_franca/lang/parse_de.py b/lingua_franca/lang/parse_de.py
index 37762e68..0245ed5f 100644
--- a/lingua_franca/lang/parse_de.py
+++ b/lingua_franca/lang/parse_de.py
@@ -20,7 +20,6 @@
     extract_numbers_generic, Normalizer
 from lingua_franca.lang.common_data_de import _DE_NUMBERS
 from lingua_franca.lang.format_de import pronounce_number_de
-from mycroft.util.time import default_timezone
 
 # TODO: short_scale and ordinals don't do anything here.
 # The parameters are present in the function signature for API compatibility
@@ -787,7 +786,7 @@ def date_found():
         for idx, en_month in enumerate(en_monthsShort):
             datestr = datestr.replace(monthsShort[idx], en_month)
 
-        temp = datetime.strptime(datestr, "%B %d").replace(tzinfo=default_timezone())
+        temp = datetime.strptime(datestr, "%B %d").replace(tzinfo=extractedDate.tzinfo)
         if not hasYear:
             temp = temp.replace(year=extractedDate.year)
             if extractedDate < temp:

From a3cbd593323bcc3bb51e0ffb85c9f3d44b041844 Mon Sep 17 00:00:00 2001
From: emphasize <swen_g@t-online.de>
Date: Mon, 8 Mar 2021 00:19:36 +0100
Subject: [PATCH 04/13] datetime parsing: No restriction to year, added
 ValueError exception

To allow future year abbrevations ('20) parsing.
Essential ValueException to be able to parse the year if any is passed (not possible atm)
---
 lingua_franca/lang/parse_de.py | 33 +++++++++++++++------------------
 1 file changed, 15 insertions(+), 18 deletions(-)

diff --git a/lingua_franca/lang/parse_de.py b/lingua_franca/lang/parse_de.py
index 0245ed5f..aaac5fd2 100644
--- a/lingua_franca/lang/parse_de.py
+++ b/lingua_franca/lang/parse_de.py
@@ -369,25 +369,17 @@ def date_found():
                     datestr += " " + wordPrev
                 start -= 1
                 used += 1
-                if wordNext and wordNext[0].isdigit():
-                    #normally the time comes in as ##:## and therefor
-                    #would break int(wordnext)
-                    if ':' in wordNext:
-                        tmp_word = wordNext.split(':')
-                        wordNext = tmp_word[0]
-                    #determine if wordnext is year data; eg 3 Januar 10:10 uhr
-                    #10:10 / 10 would be seen as such, leaving us with no time data
-                    if int(wordNext) > 60:
-                        datestr += " " + wordNext
-                        used += 1
-                        hasYear = True
-                    else:
-                        hasYear = False
-            # Mai <Year>
+                if wordnext and wordnext[0].isdigit() and not ":" in wordnext:
+                    datestr += " " + wordNext
+                    used += 1
+                    hasYear = True                         
+                else:
+                    hasYear = False
+            # Mai <day> <Year>
             elif wordNext and wordNext[0].isdigit():
                 datestr += " " + wordNext
                 used += 1
-                if wordNextNext and wordNextNext[0].isdigit():
+                if wordNextNext and wordNextNext[0].isdigit() and not ":" in wordnext:
                     datestr += " " + wordNextNext
                     used += 1
                     hasYear = True
@@ -785,8 +777,13 @@ def date_found():
             datestr = datestr.replace(months[idx], en_month)
         for idx, en_month in enumerate(en_monthsShort):
             datestr = datestr.replace(monthsShort[idx], en_month)
-
-        temp = datetime.strptime(datestr, "%B %d").replace(tzinfo=extractedDate.tzinfo)
+        
+        try:
+            temp = datetime.strptime(datestr, "%B %d").replace(tzinfo=extractedDate.tzinfo)
+        except ValueError:
+            # Try again, allowing the year
+            temp = datetime.strptime(datestr, "%B %d %Y").replace(tzinfo=extractedDate.tzinfo)
+            
         if not hasYear:
             temp = temp.replace(year=extractedDate.year)
             if extractedDate < temp:

From 8a6d1d0c1017ca8b651c7395e63f5e84ff100e54 Mon Sep 17 00:00:00 2001
From: emphasize <swen_g@t-online.de>
Date: Mon, 8 Mar 2021 00:25:14 +0100
Subject: [PATCH 05/13] datetime parsing: No restriction to year, added
 ValueError exception

without typos
---
 lingua_franca/lang/parse_de.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/lingua_franca/lang/parse_de.py b/lingua_franca/lang/parse_de.py
index aaac5fd2..72427b0d 100644
--- a/lingua_franca/lang/parse_de.py
+++ b/lingua_franca/lang/parse_de.py
@@ -369,7 +369,7 @@ def date_found():
                     datestr += " " + wordPrev
                 start -= 1
                 used += 1
-                if wordnext and wordnext[0].isdigit() and not ":" in wordnext:
+                if wordNext and wordNext[0].isdigit() and not ":" in wordNext:
                     datestr += " " + wordNext
                     used += 1
                     hasYear = True                         
@@ -379,7 +379,7 @@ def date_found():
             elif wordNext and wordNext[0].isdigit():
                 datestr += " " + wordNext
                 used += 1
-                if wordNextNext and wordNextNext[0].isdigit() and not ":" in wordnext:
+                if wordNextNext and wordNextNext[0].isdigit() and not ":" in wordNextNext:
                     datestr += " " + wordNextNext
                     used += 1
                     hasYear = True

From 25119a5e72cc3a44c64dabe16d0e44399abfb42a Mon Sep 17 00:00:00 2001
From: empasize <swen_g@t-online.de>
Date: Mon, 29 Mar 2021 12:53:40 +0200
Subject: [PATCH 06/13] normalizer -> json

---
 lingua_franca/lang/parse_de.py              | 77 ++++++++-------------
 lingua_franca/res/text/de-de/normalize.json | 74 ++++++++++++++++++++
 2 files changed, 102 insertions(+), 49 deletions(-)
 create mode 100644 lingua_franca/res/text/de-de/normalize.json

diff --git a/lingua_franca/lang/parse_de.py b/lingua_franca/lang/parse_de.py
index 5f7f7b32..3a356094 100644
--- a/lingua_franca/lang/parse_de.py
+++ b/lingua_franca/lang/parse_de.py
@@ -14,12 +14,14 @@
 # limitations under the License.
 #
 import re
+import json
 from datetime import datetime, timedelta
 from dateutil.relativedelta import relativedelta
 from lingua_franca.lang.parse_common import is_numeric, look_for_fractions, \
     extract_numbers_generic, Normalizer
 from lingua_franca.lang.common_data_de import _DE_NUMBERS
 from lingua_franca.lang.format_de import pronounce_number_de
+from lingua_franca.internal import resolve_resource_file
 
 de_numbers = {
     'null': 0,
@@ -141,7 +143,7 @@ def repl(match):
     return (duration, text)
 
 
-def extract_number_de(text, short_scale=True, ordinals=False):
+def extractnumber_de(text, short_scale=True, ordinals=False):
     """
     This function prepares the given text for parsing by making
     numbers consistent, getting rid of contractions, etc.
@@ -485,7 +487,6 @@ def date_found():
             daySpecified = True
 
     # parse time
-    timeStr = ""
     hrOffset = 0
     minOffset = 0
     secOffset = 0
@@ -816,10 +817,8 @@ def date_found():
     # perform date manipulation
 
     extractedDate = dateNow
-    extractedDate = extractedDate.replace(microsecond=0,
-                                          second=0,
-                                          minute=0,
-                                          hour=0)
+    extractedDate = extractedDate.replace(microsecond=0)
+
     if datestr != "":
         en_months = ['january', 'february', 'march', 'april', 'may', 'june',
                      'july', 'august', 'september', 'october', 'november',
@@ -833,6 +832,7 @@ def date_found():
             datestr = datestr.replace(monthsShort[idx], en_month)
 
         temp = datetime.strptime(datestr, "%B %d")
+        extractedDate = extractedDate.replace(hour=0, minute=0, second=0)
         if not hasYear:
             temp = temp.replace(year=extractedDate.year)
             if extractedDate < temp:
@@ -853,11 +853,10 @@ def date_found():
                 month=int(temp.strftime("%m")),
                 day=int(temp.strftime("%d")))
 
-    if timeStr != "":
-        temp = datetime(timeStr)
-        extractedDate = extractedDate.replace(hour=temp.strftime("%H"),
-                                              minute=temp.strftime("%M"),
-                                              second=temp.strftime("%S"))
+    else:
+        # ignore the current HH:MM:SS if relative using days or greater
+        if hrOffset == 0 and minOffset == 0 and secOffset == 0:
+            extractedDate = extractedDate.replace(hour=0, minute=0, second=0)
 
     if yearOffset != 0:
         extractedDate = extractedDate + relativedelta(years=yearOffset)
@@ -866,17 +865,21 @@ def date_found():
     if dayOffset != 0:
         extractedDate = extractedDate + relativedelta(days=dayOffset)
 
-    if hrAbs is None and minAbs is None and default_time:
-        hrAbs = default_time.hour
-        minAbs = default_time.minute
-
     if hrAbs != -1 and minAbs != -1:
+        # If no time was supplied in the string set the time to default
+        # time if it's available
+        if hrAbs is None and minAbs is None and default_time is not None:
+            hrAbs, minAbs = default_time.hour, default_time.minute
+        else:
+            hrAbs = hrAbs or 0
+            minAbs = minAbs or 0
 
-        extractedDate = extractedDate + relativedelta(hours=hrAbs or 0,
-                                                      minutes=minAbs or 0)
-        if (hrAbs or minAbs) and datestr == "":
-            if not daySpecified and dateNow > extractedDate:
+        extractedDate = extractedDate + relativedelta(hours=hrAbs,
+                                                      minutes=minAbs)
+        if (hrAbs != 0 or minAbs != 0) and datestr == "":
+            if not daySpecified and anchorDate > extractedDate:
                 extractedDate = extractedDate + relativedelta(days=1)
+
     if hrOffset != 0:
         extractedDate = extractedDate + relativedelta(hours=hrOffset)
     if minOffset != 0:
@@ -970,33 +973,6 @@ def is_ordinal_de(input_str):
 
     return False
 
-
-def normalize_de(text, remove_articles=True):
-    """ German string normalization """
-    # TODO return GermanNormalizer().normalize(text, remove_articles)
-    words = text.split()  # this also removed extra spaces
-    normalized = ""
-    for word in words:
-        if remove_articles and word in ["der", "die", "das", "des", "den",
-                                        "dem"]:
-            continue
-
-        # Expand common contractions, e.g. "isn't" -> "is not"
-        contraction = ["net", "nett"]
-        if word in contraction:
-            expansion = ["nicht", "nicht"]
-            word = expansion[contraction.index(word)]
-
-        # Convert numbers into digits, e.g. "two" -> "2"
-
-        if word in _DE_NUMBERS:
-            word = str(_DE_NUMBERS[word])
-
-        normalized += " " + word
-
-    return normalized[1:]  # strip the initial space
-
-
 def extract_numbers_de(text, short_scale=True, ordinals=False):
     """
         Takes in a string and extracts a list of numbers.
@@ -1011,9 +987,12 @@ def extract_numbers_de(text, short_scale=True, ordinals=False):
     Returns:
         list: list of extracted numbers as floats
     """
-    return extract_numbers_generic(text, pronounce_number_de, extract_number_de,
+    return extract_numbers_generic(text, pronounce_number_de, extractnumber_de,
                                    short_scale=short_scale, ordinals=ordinals)
 
-
 class GermanNormalizer(Normalizer):
-    """ TODO implement language specific normalizer"""
+    with open(resolve_resource_file("text/de-de/normalize.json")) as f:
+        _default_config = json.load(f)
+
+def normalize_de(text, remove_articles):
+    return GermanNormalizer().normalize(text, remove_articles)
diff --git a/lingua_franca/res/text/de-de/normalize.json b/lingua_franca/res/text/de-de/normalize.json
new file mode 100644
index 00000000..21c30cbc
--- /dev/null
+++ b/lingua_franca/res/text/de-de/normalize.json
@@ -0,0 +1,74 @@
+{
+  "lowercase": false,
+  "numbers_to_digits": true,
+  "expand_contractions": true,
+  "remove_symbols": false,
+  "remove_accents": false,
+  "remove_articles": false,
+  "remove_stopwords": false,
+  "contractions": {
+    "am": "an dem",
+    "ans": "an das",
+    "aufs": "auf das",
+    "beim": "bei dem",
+    "durchs": "durch das",
+    "fürs": "für das",
+    "hinterm": "hinter dem",
+    "ins": "in das",
+    "im": "in dem",
+    "übers": "über das",
+    "unters": "unter das",
+    "unterm": "unter dem",
+    "vom": "von dem",
+    "vors": "vor das",
+    "vorm": "vor dem",
+    "zum": "zu dem"
+  },
+  "word_replacements": {
+    "m", "meter",
+    "km", "kilometer"
+  },
+  "number_replacements": {
+    "null": "0",
+    "one": "1",
+    "zwei": "2",
+    "drei": "3",
+    "vier": "4",
+    "fünf": "5",
+    "sechs": "6",
+    "sieben": "7",
+    "acht": "8",
+    "neun": "9",
+    "zehn": "10",
+    "elf": "11",
+    "zwölf": "12",
+    "dreizehn": "13",
+    "vierzehn": "14",
+    "fünfzehn": "15",
+    "sechzehn": "16",
+    "siebzehn": "17",
+    "achtzehn": "18",
+    "neunzehn": "19",
+    "zwanzig": "20",
+    "einundzwanzig": "21",
+    "zweiundzwanzig": "22",
+    "dreiundzwanzig": "23",
+    "vierundzwanzig": "24",
+    "dreißig": "30",
+    "vierzig": "40",
+    "fünfzig": "50",
+    "sechzig": "60",
+    "siebzig": "70",
+    "achtzig": "80",
+    "neunzig": "90"
+  },
+  "stopwords": [],
+  "articles": [
+    "der",
+    "die",
+    "das",
+    "dem",
+    "den",
+    "des"
+  ]
+}

From 85c6ae159c47ea19cb0f3741cb6f8a91c0b175d1 Mon Sep 17 00:00:00 2001
From: empasize <swen_g@t-online.de>
Date: Mon, 29 Mar 2021 13:22:11 +0200
Subject: [PATCH 07/13] revert changes to temp to reflect #180

---
 lingua_franca/lang/parse_de.py | 9 ++-------
 1 file changed, 2 insertions(+), 7 deletions(-)

diff --git a/lingua_franca/lang/parse_de.py b/lingua_franca/lang/parse_de.py
index 72427b0d..f6c194e4 100644
--- a/lingua_franca/lang/parse_de.py
+++ b/lingua_franca/lang/parse_de.py
@@ -777,13 +777,8 @@ def date_found():
             datestr = datestr.replace(months[idx], en_month)
         for idx, en_month in enumerate(en_monthsShort):
             datestr = datestr.replace(monthsShort[idx], en_month)
-        
-        try:
-            temp = datetime.strptime(datestr, "%B %d").replace(tzinfo=extractedDate.tzinfo)
-        except ValueError:
-            # Try again, allowing the year
-            temp = datetime.strptime(datestr, "%B %d %Y").replace(tzinfo=extractedDate.tzinfo)
-            
+
+        temp = datetime.strptime(datestr, "%B %d")
         if not hasYear:
             temp = temp.replace(year=extractedDate.year)
             if extractedDate < temp:

From 7460cfd490addbaff0f2713b68918dd9768f09bb Mon Sep 17 00:00:00 2001
From: empasize <swen_g@t-online.de>
Date: Mon, 29 Mar 2021 13:50:26 +0200
Subject: [PATCH 08/13] remove wrong entries in word_replacements

---
 lingua_franca/res/text/de-de/normalize.json | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/lingua_franca/res/text/de-de/normalize.json b/lingua_franca/res/text/de-de/normalize.json
index 21c30cbc..8eb4d1d4 100644
--- a/lingua_franca/res/text/de-de/normalize.json
+++ b/lingua_franca/res/text/de-de/normalize.json
@@ -24,10 +24,7 @@
     "vorm": "vor dem",
     "zum": "zu dem"
   },
-  "word_replacements": {
-    "m", "meter",
-    "km", "kilometer"
-  },
+  "word_replacements": {},
   "number_replacements": {
     "null": "0",
     "one": "1",

From 0089d23e5f9a9a932427e2c905836e13c4f75348 Mon Sep 17 00:00:00 2001
From: empasize <swen_g@t-online.de>
Date: Mon, 29 Mar 2021 13:56:45 +0200
Subject: [PATCH 09/13] set remove_articles=True

---
 lingua_franca/lang/parse_de.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/lingua_franca/lang/parse_de.py b/lingua_franca/lang/parse_de.py
index f74967c8..d67d6052 100644
--- a/lingua_franca/lang/parse_de.py
+++ b/lingua_franca/lang/parse_de.py
@@ -374,7 +374,7 @@ def date_found():
                 if wordNext and wordNext[0].isdigit() and not ":" in wordNext:
                     datestr += " " + wordNext
                     used += 1
-                    hasYear = True                         
+                    hasYear = True
                 else:
                     hasYear = False
             # Mai <day> <Year>
@@ -940,5 +940,5 @@ class GermanNormalizer(Normalizer):
     with open(resolve_resource_file("text/de-de/normalize.json")) as f:
         _default_config = json.load(f)
 
-def normalize_de(text, remove_articles):
+def normalize_de(text, remove_articles=True):
     return GermanNormalizer().normalize(text, remove_articles)

From f9287e282588eaff4bf42209239fcf609476dcd7 Mon Sep 17 00:00:00 2001
From: empasize <swen_g@t-online.de>
Date: Mon, 29 Mar 2021 14:38:09 +0200
Subject: [PATCH 10/13] add ein, eine,... to normalizer

---
 lingua_franca/res/text/de-de/normalize.json | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/lingua_franca/res/text/de-de/normalize.json b/lingua_franca/res/text/de-de/normalize.json
index 8eb4d1d4..03466dda 100644
--- a/lingua_franca/res/text/de-de/normalize.json
+++ b/lingua_franca/res/text/de-de/normalize.json
@@ -27,7 +27,13 @@
   "word_replacements": {},
   "number_replacements": {
     "null": "0",
-    "one": "1",
+    "ein" : "1",
+    "eins" : "1",
+    "eine" : "1",
+    "einer" : "1",
+    "einem" : "1",
+    "einen": "1",
+    "eines": "1",
     "zwei": "2",
     "drei": "3",
     "vier": "4",

From a222c6a86a3b3d331086ef96d9e07bf7ce15b86b Mon Sep 17 00:00:00 2001
From: empasize <swen_g@t-online.de>
Date: Mon, 29 Mar 2021 14:56:03 +0200
Subject: [PATCH 11/13] extractnumber -> extract_number

---
 lingua_franca/lang/parse_de.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/lingua_franca/lang/parse_de.py b/lingua_franca/lang/parse_de.py
index d67d6052..110eb189 100644
--- a/lingua_franca/lang/parse_de.py
+++ b/lingua_franca/lang/parse_de.py
@@ -85,7 +85,7 @@ def repl(match):
     return (duration, text)
 
 
-def extractnumber_de(text, short_scale=True, ordinals=False):
+def extract_number_de(text, short_scale=True, ordinals=False):
     """
     This function prepares the given text for parsing by making
     numbers consistent, getting rid of contractions, etc.
@@ -933,7 +933,7 @@ def extract_numbers_de(text, short_scale=True, ordinals=False):
     Returns:
         list: list of extracted numbers as floats
     """
-    return extract_numbers_generic(text, pronounce_number_de, extractnumber_de,
+    return extract_numbers_generic(text, pronounce_number_de, extract_number_de,
                                    short_scale=short_scale, ordinals=ordinals)
 
 class GermanNormalizer(Normalizer):

From 6f1161053d894bd376d90372cb00d660385804cc Mon Sep 17 00:00:00 2001
From: empasize <swen_g@t-online.de>
Date: Mon, 29 Mar 2021 20:39:56 +0200
Subject: [PATCH 12/13] add test

---
 test/test_parse_de.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/test/test_parse_de.py b/test/test_parse_de.py
index 18ed335a..c38b7027 100644
--- a/test/test_parse_de.py
+++ b/test/test_parse_de.py
@@ -102,6 +102,8 @@ def testExtract(text, expected_date, expected_leftover):
                     "2017-06-29 00:00:00", "wie ist das wetter")
         testExtract("erinnere mich um 10:45 abends",
                     "2017-06-27 22:45:00", "erinnere mich")
+        testExtract("erinnere mich an arzttermin freitag 10:45 uhr",
+                    "2017-06-30 10:45:00", "erinnere mich an arzttermin uhr")
         testExtract("was ist das Wetter am freitag morgen",
                     "2017-06-30 08:00:00", "was ist das wetter")
         testExtract("wie ist das wetter morgen",

From 4f85ac63c3ec649bf17a7738505029d2dc6b46cf Mon Sep 17 00:00:00 2001
From: empasize <swen_g@t-online.de>
Date: Mon, 29 Mar 2021 20:46:00 +0200
Subject: [PATCH 13/13] add test

---
 test/test_parse_de.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/test/test_parse_de.py b/test/test_parse_de.py
index c38b7027..a35a3772 100644
--- a/test/test_parse_de.py
+++ b/test/test_parse_de.py
@@ -103,7 +103,7 @@ def testExtract(text, expected_date, expected_leftover):
         testExtract("erinnere mich um 10:45 abends",
                     "2017-06-27 22:45:00", "erinnere mich")
         testExtract("erinnere mich an arzttermin freitag 10:45 uhr",
-                    "2017-06-30 10:45:00", "erinnere mich an arzttermin uhr")
+                    "2017-06-30 10:45:00", "erinnere mich arzttermin uhr")
         testExtract("was ist das Wetter am freitag morgen",
                     "2017-06-30 08:00:00", "was ist das wetter")
         testExtract("wie ist das wetter morgen",