Add support for Greek language (#2535)

* Add contributor agreement

* Support for Greek language

* Fix missing el_tokenizer

.github/contributors/Eleni170.md

@@ -0,0 +1,106 @@
+# spaCy contributor agreement
+
+This spaCy Contributor Agreement (**"SCA"**) is based on the
+[Oracle Contributor Agreement](http://www.oracle.com/technetwork/oca-405177.pdf).
+The SCA applies to any contribution that you make to any product or project
+managed by us (the **"project"**), and sets out the intellectual property rights
+you grant to us in the contributed materials. The term **"us"** shall mean
+[ExplosionAI UG (haftungsbeschränkt)](https://explosion.ai/legal). The term
+**"you"** shall mean the person or entity identified below.
+
+If you agree to be bound by these terms, fill in the information requested
+below and include the filled-in version with your first pull request, under the
+folder [`.github/contributors/`](/.github/contributors/). The name of the file
+should be your GitHub username, with the extension `.md`. For example, the user
+example_user would create the file `.github/contributors/example_user.md`.
+
+Read this agreement carefully before signing. These terms and conditions
+constitute a binding legal agreement.
+
+## Contributor Agreement
+
+1. The term "contribution" or "contributed materials" means any source code,
+object code, patch, tool, sample, graphic, specification, manual,
+documentation, or any other material posted or submitted by you to the project.
+
+2. With respect to any worldwide copyrights, or copyright applications and
+registrations, in your contribution:
+
+    * you hereby assign to us joint ownership, and to the extent that such
+    assignment is or becomes invalid, ineffective or unenforceable, you hereby
+    grant to us a perpetual, irrevocable, non-exclusive, worldwide, no-charge,
+    royalty-free, unrestricted license to exercise all rights under those
+    copyrights. This includes, at our option, the right to sublicense these same
+    rights to third parties through multiple levels of sublicensees or other
+    licensing arrangements;
+
+    * you agree that each of us can do all things in relation to your
+    contribution as if each of us were the sole owners, and if one of us makes
+    a derivative work of your contribution, the one who makes the derivative
+    work (or has it made will be the sole owner of that derivative work;
+
+    * you agree that you will not assert any moral rights in your contribution
+    against us, our licensees or transferees;
+
+    * you agree that we may register a copyright in your contribution and
+    exercise all ownership rights associated with it; and
+
+    * you agree that neither of us has any duty to consult with, obtain the
+    consent of, pay or render an accounting to the other for any use or
+    distribution of your contribution.
+
+3. With respect to any patents you own, or that you can license without payment
+to any third party, you hereby grant to us a perpetual, irrevocable,
+non-exclusive, worldwide, no-charge, royalty-free license to:
+
+    * make, have made, use, sell, offer to sell, import, and otherwise transfer
+    your contribution in whole or in part, alone or in combination with or
+    included in any product, work or materials arising out of the project to
+    which your contribution was submitted, and
+
+    * at our option, to sublicense these same rights to third parties through
+    multiple levels of sublicensees or other licensing arrangements.
+
+4. Except as set out above, you keep all right, title, and interest in your
+contribution. The rights that you grant to us under these terms are effective
+on the date you first submitted a contribution to us, even if your submission
+took place before the date you sign these terms.
+
+5. You covenant, represent, warrant and agree that:
+
+    * Each contribution that you submit is and shall be an original work of
+    authorship and you can legally grant the rights set out in this SCA;
+
+    * to the best of your knowledge, each contribution will not violate any
+    third party's copyrights, trademarks, patents, or other intellectual
+    property rights; and
+
+    * each contribution shall be in compliance with U.S. export control laws and
+    other applicable export and import laws. You agree to notify us if you
+    become aware of any circumstance which would make any of the foregoing
+    representations inaccurate in any respect. We may publicly disclose your
+    participation in the project, including the fact that you have signed the SCA.
+
+6. This SCA is governed by the laws of the State of California and applicable
+U.S. Federal law. Any choice of law rules will not apply.
+
+7. Please place an “x” on one of the applicable statement below. Please do NOT
+mark both statements:
+
+    * [ ] I am signing on behalf of myself as an individual and no other person
+    or entity, including my employer, has or will have rights with respect to my
+    contributions.
+
+    * [x] I am signing on behalf of my employer or a legal entity and I have the
+    actual authority to contractually bind that entity.
+
+## Contributor Details
+
+| Field                          | Entry                |
+|------------------------------- | -------------------- |
+| Name                           | Eleni Partalidou     |
+| Company name (if applicable)   | DataScouting         |
+| Title or role (if applicable)  | Software Engineer    |
+| Date                           | 06.07.2018           |
+| GitHub username                | Eleni170             |
+| Website (optional)             |                      |

spacy/lang/el/__init__.py

@@ -0,0 +1,40 @@
+# -*- coding: utf-8 -*-
+
+from __future__ import unicode_literals
+
+from .tokenizer_exceptions import TOKENIZER_EXCEPTIONS
+from .tag_map import TAG_MAP
+from .stop_words import STOP_WORDS
+from .lex_attrs import LEX_ATTRS
+from .lemmatizer import LOOKUP
+from .punctuation import TOKENIZER_PREFIXES, TOKENIZER_SUFFIXES, TOKENIZER_INFIXES
+from ..tokenizer_exceptions import BASE_EXCEPTIONS
+from ..norm_exceptions import BASE_NORMS
+from ...language import Language
+from ...attrs import LANG, NORM
+from ...util import update_exc, add_lookups
+
+
+class GreekDefaults(Language.Defaults):
+    lex_attr_getters = dict(Language.Defaults.lex_attr_getters)
+    lex_attr_getters.update(LEX_ATTRS)
+    lex_attr_getters[LANG] = lambda text: 'el'  # ISO code
+    lex_attr_getters[NORM] = add_lookups(Language.Defaults.lex_attr_getters[NORM], BASE_NORMS)
+    tokenizer_exceptions = update_exc(BASE_EXCEPTIONS, TOKENIZER_EXCEPTIONS)
+    stop_words = STOP_WORDS
+    lemma_lookup = LOOKUP
+    tag_map = TAG_MAP
+    prefixes = TOKENIZER_PREFIXES
+    suffixes = TOKENIZER_SUFFIXES
+    infixes = TOKENIZER_INFIXES
+
+
+class Greek(Language):
+
+    lang = 'el'  # ISO code
+    Defaults = GreekDefaults  # set Defaults to custom language defaults
+
+
+# set default export – this allows the language class to be lazy-loaded
+__all__ = ['Greek']
+

spacy/lang/el/examples.py

@@ -0,0 +1,19 @@
+# -*- coding: utf-8 -*-
+
+from __future__ import unicode_literals
+
+"""
+Example sentences to test spaCy and its language models.
+>>> from spacy.lang.el.examples import sentences
+>>> docs = nlp.pipe(sentences)
+"""
+
+sentences = [
+    "Η άνιση κατανομή του πλούτου και του εισοδήματος, η οποία έχει λάβει τρομερές διαστάσεις, δεν δείχνει τάσεις βελτίωσης.",
+    "Ο στόχος της σύντομης αυτής έκθεσης είναι να συνοψίσει τα κυριότερα συμπεράσματα των επισκοπήσεων κάθε μιας χώρας.",
+    "Μέχρι αργά χθες το βράδυ ο πλοιοκτήτης παρέμενε έξω από το γραφείο του γενικού γραμματέα του υπουργείου, ενώ είχε μόνον τηλεφωνική επικοινωνία με τον υπουργό.",
+    "Σύμφωνα με καλά ενημερωμένη πηγή, από την επεξεργασία του προέκυψε ότι οι δράστες της επίθεσης ήταν δύο, καθώς και ότι προσέγγισαν και αποχώρησαν από το σημείο με μοτοσικλέτα.",
+    "Η υποδομή καταλυμάτων στην Ελλάδα είναι πλήρης και ανανεώνεται συνεχώς.",
+    "Το επείγον ταχυδρομείο (ήτοι το παραδοτέο εντός 48 ωρών το πολύ) μπορεί να μεταφέρεται αεροπορικώς μόνον εφόσον εφαρμόζονται οι κανόνες ασφαλείας.",
+    "Στις ορεινές περιοχές του νησιού οι χιονοπτώσεις και οι παγετοί είναι περιορισμένοι ενώ στις παραθαλάσσιες περιοχές σημειώνονται σπανίως."
+]

spacy/lang/el/lex_attrs.py

@@ -0,0 +1,38 @@
+# -*- coding: utf-8 -*-
+
+from __future__ import unicode_literals
+
+from ...attrs import LIKE_NUM
+
+_num_words = ['μηδέν', 'ένας', 'δυο', 'δυό', 'τρεις', 'τέσσερις', 'πέντε', 'έξι', 'εφτά', 'επτά', 'οκτώ', 'οχτώ',
+              'εννιά', 'εννέα', 'δέκα', 'έντεκα', 'ένδεκα', 'δώδεκα', 'δεκατρείς', 'δεκατέσσερις', 'δεκαπέντε',
+              'δεκαέξι', 'δεκαεπτά', 'δεκαοχτώ', 'δεκαεννέα', 'δεκαεννεα', 'είκοσι', 'τριάντα', 'σαράντα', 'πενήντα',
+              'εξήντα', 'εβδομήντα', 'ογδόντα', 'ενενήντα', 'εκατό', 'διακόσιοι', 'διακόσοι', 'τριακόσιοι', 'τριακόσοι',
+              'τετρακόσιοι', 'τετρακόσοι', 'πεντακόσιοι', 'πεντακόσοι', 'εξακόσιοι', 'εξακόσοι', 'εφτακόσιοι',
+              'εφτακόσοι', 'επτακόσιοι', 'επτακόσοι', 'οχτακόσιοι', 'οχτακόσοι', 'οκτακόσιοι', 'οκτακόσοι',
+              'εννιακόσιοι', 'χίλιοι', 'χιλιάδα', 'εκατομμύριο', 'δισεκατομμύριο', 'τρισεκατομμύριο', 'τετράκις',
+              'πεντάκις', 'εξάκις', 'επτάκις', 'οκτάκις', 'εννεάκις']
+
+
+def like_num(text):
+    text = text.replace(',', '').replace('.', '')
+    if text.isdigit():
+        return True
+    if text.count('/') == 1:
+        num, denom = text.split('/')
+        if num.isdigit() and denom.isdigit():
+            return True
+    if text.count('^') == 1:
+        num, denom = text.split('^')
+        if num.isdigit() and denom.isdigit():
+            return True
+    if text.lower() in _num_words or text.lower().split(' ')[0] in _num_words:
+        return True
+    if text in _num_words:
+        return True
+    return False
+
+
+LEX_ATTRS = {
+    LIKE_NUM: like_num
+}

spacy/lang/el/punctuation.py

@@ -0,0 +1,66 @@
+# -*- coding: utf-8 -*-
+
+from __future__ import unicode_literals
+
+from ..char_classes import LIST_PUNCT, LIST_ELLIPSES, LIST_QUOTES, LIST_CURRENCY
+from ..char_classes import LIST_ICONS, ALPHA_LOWER, ALPHA_UPPER, ALPHA, HYPHENS
+from ..char_classes import QUOTES, CURRENCY
+
+_units = ('km km² km³ m m² m³ dm dm² dm³ cm cm² cm³ mm mm² mm³ ha µm nm yd in ft '
+          'kg g mg µg t lb oz m/s km/h kmh mph hPa Pa mbar mb MB kb KB gb GB tb '
+          'TB T G M K км км² км³ м м² м³ дм дм² дм³ см см² см³ мм мм² мм³ нм '
+          'кг г мг м/с км/ч кПа Па мбар Кб КБ кб Мб МБ мб Гб ГБ гб Тб ТБ тб')
+merge_chars = lambda char: char.strip().replace(' ', '|')
+UNITS = merge_chars(_units)
+
+_prefixes = (['\'\'', '§', '%', '=', r'\+[0-9]+%',  # 90%
+              r'\'([0-9]){2}([\-]\'([0-9]){2})*',  # '12'-13
+              r'\-([0-9]){1,9}\.([0-9]){1,9}',  # -12.13
+              r'\'([Α-Ωα-ωίϊΐόάέύϋΰήώ]+)\'',  # 'αβγ'
+              r'([Α-Ωα-ωίϊΐόάέύϋΰήώ]){1,3}\'',  # αβγ'
+              r'http://www.[A-Za-z]+\-[A-Za-z]+(\.[A-Za-z]+)+(\/[A-Za-z]+)*(\.[A-Za-z]+)*',
+              r'[ΈΆΊΑ-Ωα-ωίϊΐόάέύϋΰήώ]+\*',  # όνομα*
+              r'\$([0-9])+([\,\.]([0-9])+){0,1}',
+              ] + LIST_PUNCT + LIST_ELLIPSES + LIST_QUOTES +
+             LIST_CURRENCY + LIST_ICONS)
+
+_suffixes = (LIST_PUNCT + LIST_ELLIPSES + LIST_QUOTES + LIST_ICONS +
+             [r'(?<=[0-9])\+',  # 12+
+              r'([0-9])+\'',  # 12'
+              r'([A-Za-z])?\'',  # a'
+              r'^([0-9]){1,2}\.',  # 12.
+              r' ([0-9]){1,2}\.',  # 12.
+              r'([0-9]){1}\) ',  # 12)
+              r'^([0-9]){1}\)$',  # 12)
+              r'(?<=°[FfCcKk])\.',
+              r'([0-9])+\&',  # 12&
+              r'(?<=[0-9])(?:{})'.format(CURRENCY),
+              r'(?<=[0-9])(?:{})'.format(UNITS),
+              r'(?<=[0-9{}{}(?:{})])\.'.format(ALPHA_LOWER, r'²\-\)\]\+', QUOTES),
+              r'(?<=[{a}][{a}])\.'.format(a=ALPHA_UPPER),
+              r'(?<=[Α-Ωα-ωίϊΐόάέύϋΰήώ])\-',  # όνομα-
+              r'(?<=[Α-Ωα-ωίϊΐόάέύϋΰήώ])\.',
+              r'^[Α-Ω]{1}\.',
+              r'\ [Α-Ω]{1}\.',
+              r'[ΈΆΊΑΌ-Ωα-ωίϊΐόάέύϋΰήώ]+([\-]([ΈΆΊΑΌ-Ωα-ωίϊΐόάέύϋΰήώ]+))+',  # πρώτος-δεύτερος , πρώτος-δεύτερος-τρίτος
+              r'([0-9]+)mg',  # 13mg
+              r'([0-9]+)\.([0-9]+)m'  # 1.2m
+              ])
+
+_infixes = (LIST_ELLIPSES + LIST_ICONS +
+            [r'(?<=[0-9])[+\/\-\*^](?=[0-9])',  # 1/2 , 1-2 , 1*2
+             r'([a-zA-Z]+)\/([a-zA-Z]+)\/([a-zA-Z]+)',  # name1/name2/name3
+             r'([0-9])+(\.([0-9]+))*([\-]([0-9])+)+',  # 10.9 , 10.9.9 , 10.9-6
+             r'([0-9])+[,]([0-9])+[\-]([0-9])+[,]([0-9])+',  # 10,11,12
+             r'([0-9])+[ης]+([\-]([0-9])+)+',  # 1ης-2
+             r'([0-9]){1,4}[\/]([0-9]){1,2}([\/]([0-9]){0,4}){0,1}',  # 15/2 , 15/2/17 , 2017/2/15
+             r'[A-Za-z]+\@[A-Za-z]+(\-[A-Za-z]+)*\.[A-Za-z]+',  # abc@cde-fgh.a
+             r'([a-zA-Z]+)(\-([a-zA-Z]+))+',  # abc-abc
+             r'(?<=[{}])\.(?=[{}])'.format(ALPHA_LOWER, ALPHA_UPPER),
+             r'(?<=[{a}]),(?=[{a}])'.format(a=ALPHA),
+             r'(?<=[{a}])[?";:=,.]*(?:{h})(?=[{a}])'.format(a=ALPHA, h=HYPHENS),
+             r'(?<=[{a}"])[:<>=/](?=[{a}])'.format(a=ALPHA)])
+
+TOKENIZER_PREFIXES = _prefixes
+TOKENIZER_SUFFIXES = _suffixes
+TOKENIZER_INFIXES = _infixes

spacy/lang/el/stop_words.py

@@ -0,0 +1,98 @@
+# -*- coding: utf-8 -*-
+
+from __future__ import unicode_literals
+
+# Stop words
+
+# Link to greek stop words: https://www.translatum.gr/forum/index.php?topic=3550.0?topic=3550.0
+
+
+STOP_WORDS = set("""
+αδιάκοπα αι ακόμα ακόμη ακριβώς αλήθεια αληθινά αλλά αλλαχού άλλες άλλη άλλην 
+άλλης αλλιώς αλλιώτικα άλλο άλλοι αλλοιώς αλλοιώτικα άλλον άλλος άλλοτε αλλού 
+άλλους άλλων άμα άμεσα αμέσως αν ανά ανάμεσα αναμεταξύ άνευ αντί αντίπερα αντίς 
+άνω ανωτέρω άξαφνα απ απέναντι από απόψε άρα άραγε αργά αργότερο αριστερά αρκετά 
+αρχικά ας αύριο αυτά αυτές αυτή αυτήν αυτής αυτό αυτοί αυτόν αυτός αυτού αυτούς 
+αυτών αφότου αφού 
+
+βέβαια βεβαιότατα 
+
+γι για γρήγορα γύρω 
+
+δα δε δείνα δεν δεξιά δήθεν δηλαδή δι δια διαρκώς δικά δικό δικοί δικός δικού 
+δικούς διόλου δίπλα δίχως 
+
+εάν εαυτό εαυτόν εαυτού εαυτούς εαυτών έγκαιρα εγκαίρως εγώ εδώ ειδεμή είθε είμαι 
+είμαστε είναι εις είσαι είσαστε είστε είτε είχα είχαμε είχαν είχατε είχε είχες έκαστα 
+έκαστες έκαστη έκαστην έκαστης έκαστο έκαστοι έκαστον έκαστος εκάστου εκάστους εκάστων 
+εκεί εκείνα εκείνες εκείνη εκείνην εκείνης εκείνο εκείνοι εκείνον εκείνος εκείνου 
+εκείνους εκείνων εκτός εμάς εμείς εμένα εμπρός εν ένα έναν ένας ενός εντελώς εντός 
+εντωμεταξύ ενώ εξ έξαφνα εξήσ εξίσου έξω επάνω επειδή έπειτα επί επίσης επομένως εσάς 
+εσείς εσένα έστω εσύ ετέρα ετέραι ετέρας έτερες έτερη έτερης έτερο έτεροι έτερον έτερος 
+ετέρου έτερους ετέρων ετούτα ετούτες ετούτη ετούτην ετούτης ετούτο ετούτοι ετούτον 
+ετούτος ετούτου ετούτους ετούτων έτσι εύγε ευθύς ευτυχώς εφεξής έχει έχεις έχετε 
+εχθές έχομε έχουμε έχουν εχτές έχω έως 
+
+η ήδη ήμασταν ήμαστε ήμουν ήσασταν ήσαστε ήσουν ήταν ήτανε ήτοι ήττον 
+
+θα 
+
+ι ίδια ίδιαν ιδίας ίδιες ίδιο ίδιοι ίδιον ίδιοσ ιδίου ίδιους ίδιων ιδίως ιι ιιι 
+ίσαμε ίσια ίσως 
+
+κάθε καθεμία καθεμίας καθένα καθένας καθενός καθετί καθόλου καθώς και κακά κακώς καλά 
+καλώς καμία καμίαν καμίας κάμποσα κάμποσες κάμποση κάμποσην κάμποσης κάμποσο κάμποσοι 
+κάμποσον κάμποσος κάμποσου κάμποσους κάμποσων κανείς κάνεν κανένα κανέναν κανένας 
+κανενός κάποια κάποιαν κάποιας κάποιες κάποιο κάποιοι κάποιον κάποιος κάποιου κάποιους 
+κάποιων κάποτε κάπου κάπως κατ κατά κάτι κατιτί κατόπιν κάτω κιόλας κλπ κοντά κτλ κυρίως 
+
+λιγάκι λίγο λιγότερο λόγω λοιπά λοιπόν 
+
+μα μαζί μακάρι μακρυά μάλιστα μάλλον μας με μεθαύριο μείον μέλει μέλλεται μεμιάς μεν 
+μερικά μερικές μερικοί μερικούς μερικών μέσα μετ μετά μεταξύ μέχρι μη μήδε μην μήπως 
+μήτε μια μιαν μιας μόλις μολονότι μονάχα μόνες μόνη μόνην μόνης μόνο μόνοι μονομιάς 
+μόνος μόνου μόνους μόνων μου μπορεί μπορούν μπράβο μπρος 
+
+να ναι νωρίς 
+
+ξανά ξαφνικά 
+
+ο οι όλα όλες όλη όλην όλης όλο ολόγυρα όλοι όλον ολονέν όλος ολότελα όλου όλους όλων 
+όλως ολωσδιόλου όμως όποια οποιαδήποτε οποίαν οποιανδήποτε οποίας οποιασδήποτε οποιδήποτε 
+όποιες οποιεσδήποτε όποιο οποιοδηήποτε όποιοι όποιον οποιονδήποτε όποιος οποιοσδήποτε 
+οποίου οποιουδήποτε οποίους οποιουσδήποτε οποίων οποιωνδήποτε όποτε οποτεδήποτε όπου 
+οπουδήποτε όπως ορισμένα ορισμένες ορισμένων ορισμένως όσα οσαδήποτε όσες οσεσδήποτε 
+όση οσηδήποτε όσην οσηνδήποτε όσης οσησδήποτε όσο οσοδήποτε όσοι οσοιδήποτε όσον οσονδήποτε 
+όσος οσοσδήποτε όσου οσουδήποτε όσους οσουσδήποτε όσων οσωνδήποτε όταν ότι οτιδήποτε 
+ότου ου ουδέ ούτε όχι 
+
+πάλι πάντοτε παντού πάντως πάρα πέρα πέρι περίπου περισσότερο πέρσι πέρυσι πια πιθανόν 
+πιο πίσω πλάι πλέον πλην ποιά ποιάν ποιάς ποιές ποιό ποιοί ποιόν ποιός ποιού ποιούς 
+ποιών πολύ πόσες πόση πόσην πόσης πόσοι πόσος πόσους πότε πού πούθε πουθενά πρέπει 
+πριν προ προκειμένου πρόκειται πρόπερσι προς προτού προχθές προχτές πρωτύτερα πώς 
+
+σαν σας σε σεις σήμερα σιγά σου στα στη στην στης στις στο στον στου στους στων συγχρόνως 
+συν συνάμα συνεπώς συνήθως συχνά συχνάς συχνές συχνή συχνήν συχνής συχνό συχνοί συχνόν 
+συχνός συχνού συχνούς συχνών συχνώς σχεδόν σωστά 
+
+τα τάδε ταύτα ταύτες ταύτη ταύτην ταύτης ταύτοταύτον ταύτος ταύτου ταύτων τάχα τάχατε 
+τελικά τελικώς τες τέτοια τέτοιαν τέτοιας τέτοιες τέτοιο τέτοιοι τέτοιον τέτοιος τέτοιου 
+τέτοιους τέτοιων τη την της τι τίποτα τίποτε τις το τοι τον τοσ τόσα τόσες τόση τόσην 
+τόσης τόσο τόσοι τόσον τόσος τόσου τόσους τόσων τότε του τουλάχιστο τουλάχιστον τους τούτα 
+τούτες τούτη τούτην τούτης τούτο τούτοι τούτοις τούτον τούτος τούτου τούτους τούτων τυχόν 
+των τώρα 
+
+υπ υπέρ υπό υπόψη υπόψιν ύστερα 
+
+φέτος 
+
+χαμηλά χθες χτες χωρίς χωριστά 
+
+ψηλά 
+
+ω ωραία ως ωσάν ωσότου ώσπου ώστε ωστόσο ωχ
+""".split())
+
+
+
+

spacy/lang/el/tag_map.py

@@ -0,0 +1,599 @@
+# -*- coding: utf-8 -*-
+
+from __future__ import unicode_literals
+from ...symbols import POS, PUNCT, SYM, ADJ, CCONJ, SCONJ, NUM, DET, ADV, ADP, X, VERB
+from ...symbols import NOUN, PROPN, PART, INTJ,SPACE,PRON
+
+TAG_MAP = {
+    "ABBR":    {POS: NOUN, "Abbr":"Yes"},
+    "AdXxBa":    {POS: ADV, "Degree": ""},
+    "AdXxCp":    {POS: ADV, "Degree": "Cmp"},
+    "AdXxSu":    {POS: ADV, "Degree": "Sup"},
+    "AjBaFePlAc":    {POS: ADJ, "Degree": "", "Gender": "Fem", "Number": "Plur", "Case": "Acc"},
+    "AjBaFePlDa":    {POS: ADJ, "Degree": "", "Gender": "Fem", "Number": "Plur", "Case": "Dat"},
+    "AjBaFePlGe":    {POS: ADJ, "Degree": "", "Gender": "Fem", "Number": "Plur", "Case": "Gen"},
+    "AjBaFePlNm":    {POS: ADJ, "Degree": "", "Gender": "Fem", "Number": "Plur", "Case": "Nom"},
+    "AjBaFePlVo":    {POS: ADJ, "Degree": "", "Gender": "Fem", "Number": "Plur", "Case": "Voc"},
+    "AjBaFeSgAc":    {POS: ADJ, "Degree": "", "Gender": "Fem", "Number": "Sing", "Case": "Acc"},
+    "AjBaFeSgDa":    {POS: ADJ, "Degree": "", "Gender": "Fem", "Number": "Sing", "Case": "Dat"},
+    "AjBaFeSgGe":    {POS: ADJ, "Degree": "", "Gender": "Fem", "Number": "Sing", "Case": "Gen"},
+    "AjBaFeSgNm":    {POS: ADJ, "Degree": "", "Gender": "Fem", "Number": "Sing", "Case": "Nom"},
+    "AjBaFeSgVo":    {POS: ADJ, "Degree": "", "Gender": "Fem", "Number": "Sing", "Case": "Voc"},
+    "AjBaMaPlAc":    {POS: ADJ, "Degree": "", "Gender": "Masc", "Number": "Plur", "Case": "Acc"},
+    "AjBaMaPlDa":    {POS: ADJ, "Degree": "", "Gender": "Masc", "Number": "Plur", "Case": "Dat"},
+    "AjBaMaPlGe":    {POS: ADJ, "Degree": "", "Gender": "Masc", "Number": "Plur", "Case": "Gen"},
+    "AjBaMaPlNm":    {POS: ADJ, "Degree": "", "Gender": "Masc", "Number": "Plur", "Case": "Nom"},
+    "AjBaMaPlVo":    {POS: ADJ, "Degree": "", "Gender": "Masc", "Number": "Plur", "Case": "Voc"},
+    "AjBaMaSgAc":    {POS: ADJ, "Degree": "", "Gender": "Masc", "Number": "Sing", "Case": "Acc"},
+    "AjBaMaSgDa":    {POS: ADJ, "Degree": "", "Gender": "Masc", "Number": "Sing", "Case": "Dat"},
+    "AjBaMaSgGe":    {POS: ADJ, "Degree": "", "Gender": "Masc", "Number": "Sing", "Case": "Gen"},
+    "AjBaMaSgNm":    {POS: ADJ, "Degree": "", "Gender": "Masc", "Number": "Sing", "Case": "Nom"},
+    "AjBaMaSgVo":    {POS: ADJ, "Degree": "", "Gender": "Masc", "Number": "Sing", "Case": "Voc"},
+    "AjBaNePlAc":    {POS: ADJ, "Degree": "", "Gender": "Neut", "Number": "Plur", "Case": "Acc"},
+    "AjBaNePlDa":    {POS: ADJ, "Degree": "", "Gender": "Neut", "Number": "Plur", "Case": "Dat"},
+    "AjBaNePlGe":    {POS: ADJ, "Degree": "", "Gender": "Neut", "Number": "Plur", "Case": "Gen"},
+    "AjBaNePlNm":    {POS: ADJ, "Degree": "", "Gender": "Neut", "Number": "Plur", "Case": "Nom"},
+    "AjBaNePlVo":    {POS: ADJ, "Degree": "", "Gender": "Neut", "Number": "Plur", "Case": "Voc"},
+    "AjBaNeSgAc":    {POS: ADJ, "Degree": "", "Gender": "Neut", "Number": "Sing", "Case": "Acc"},
+    "AjBaNeSgDa":    {POS: ADJ, "Degree": "", "Gender": "Neut", "Number": "Sing", "Case": "Dat"},
+    "AjBaNeSgGe":    {POS: ADJ, "Degree": "", "Gender": "Neut", "Number": "Sing", "Case": "Gen"},
+    "AjBaNeSgNm":    {POS: ADJ, "Degree": "", "Gender": "Neut", "Number": "Sing", "Case": "Nom"},
+    "AjBaNeSgVo":    {POS: ADJ, "Degree": "", "Gender": "Neut", "Number": "Sing", "Case": "Voc"},
+    "AjCpFePlAc":    {POS: ADJ, "Degree": "Cmp", "Gender": "Fem", "Number": "Plur", "Case": "Acc"},
+    "AjCpFePlDa":    {POS: ADJ, "Degree": "Cmp", "Gender": "Fem", "Number": "Plur", "Case": "Dat"},
+    "AjCpFePlGe":    {POS: ADJ, "Degree": "Cmp", "Gender": "Fem", "Number": "Plur", "Case": "Gen"},
+    "AjCpFePlNm":    {POS: ADJ, "Degree": "Cmp", "Gender": "Fem", "Number": "Plur", "Case": "Nom"},
+    "AjCpFePlVo":    {POS: ADJ, "Degree": "Cmp", "Gender": "Fem", "Number": "Plur", "Case": "Voc"},
+    "AjCpFeSgAc":    {POS: ADJ, "Degree": "Cmp", "Gender": "Fem", "Number": "Sing", "Case": "Acc"},
+    "AjCpFeSgDa":    {POS: ADJ, "Degree": "Cmp", "Gender": "Fem", "Number": "Sing", "Case": "Dat"},
+    "AjCpFeSgGe":    {POS: ADJ, "Degree": "Cmp", "Gender": "Fem", "Number": "Sing", "Case": "Gen"},
+    "AjCpFeSgNm":    {POS: ADJ, "Degree": "Cmp", "Gender": "Fem", "Number": "Sing", "Case": "Nom"},
+    "AjCpFeSgVo":    {POS: ADJ, "Degree": "Cmp", "Gender": "Fem", "Number": "Sing", "Case": "Voc"},
+    "AjCpMaPlAc":    {POS: ADJ, "Degree": "Cmp", "Gender": "Masc", "Number": "Plur", "Case": "Acc"},
+    "AjCpMaPlDa":    {POS: ADJ, "Degree": "Cmp", "Gender": "Masc", "Number": "Plur", "Case": "Dat"},
+    "AjCpMaPlGe":    {POS: ADJ, "Degree": "Cmp", "Gender": "Masc", "Number": "Plur", "Case": "Gen"},
+    "AjCpMaPlNm":    {POS: ADJ, "Degree": "Cmp", "Gender": "Masc", "Number": "Plur", "Case": "Nom"},
+    "AjCpMaPlVo":    {POS: ADJ, "Degree": "Cmp", "Gender": "Masc", "Number": "Plur", "Case": "Voc"},
+    "AjCpMaSgAc":    {POS: ADJ, "Degree": "Cmp", "Gender": "Masc", "Number": "Sing", "Case": "Acc"},
+    "AjCpMaSgDa":    {POS: ADJ, "Degree": "Cmp", "Gender": "Masc", "Number": "Sing", "Case": "Dat"},
+    "AjCpMaSgGe":    {POS: ADJ, "Degree": "Cmp", "Gender": "Masc", "Number": "Sing", "Case": "Gen"},
+    "AjCpMaSgNm":    {POS: ADJ, "Degree": "Cmp", "Gender": "Masc", "Number": "Sing", "Case": "Nom"},
+    "AjCpMaSgVo":    {POS: ADJ, "Degree": "Cmp", "Gender": "Masc", "Number": "Sing", "Case": "Voc"},
+    "AjCpNePlAc":    {POS: ADJ, "Degree": "Cmp", "Gender": "Neut", "Number": "Plur", "Case": "Acc"},
+    "AjCpNePlDa":    {POS: ADJ, "Degree": "Cmp", "Gender": "Neut", "Number": "Plur", "Case": "Dat"},
+    "AjCpNePlGe":    {POS: ADJ, "Degree": "Cmp", "Gender": "Neut", "Number": "Plur", "Case": "Gen"},
+    "AjCpNePlNm":    {POS: ADJ, "Degree": "Cmp", "Gender": "Neut", "Number": "Plur", "Case": "Nom"},
+    "AjCpNePlVo":    {POS: ADJ, "Degree": "Cmp", "Gender": "Neut", "Number": "Plur", "Case": "Voc"},
+    "AjCpNeSgAc":    {POS: ADJ, "Degree": "Cmp", "Gender": "Neut", "Number": "Sing", "Case": "Acc"},
+    "AjCpNeSgDa":    {POS: ADJ, "Degree": "Cmp", "Gender": "Neut", "Number": "Sing", "Case": "Dat"},
+    "AjCpNeSgGe":    {POS: ADJ, "Degree": "Cmp", "Gender": "Neut", "Number": "Sing", "Case": "Gen"},
+    "AjCpNeSgNm":    {POS: ADJ, "Degree": "Cmp", "Gender": "Neut", "Number": "Sing", "Case": "Nom"},
+    "AjCpNeSgVo":    {POS: ADJ, "Degree": "Cmp", "Gender": "Neut", "Number": "Sing", "Case": "Voc"},
+    "AjSuFePlAc":    {POS: ADJ, "Degree": "Sup", "Gender": "Fem", "Number": "Plur", "Case": "Acc"},
+    "AjSuFePlDa":    {POS: ADJ, "Degree": "Sup", "Gender": "Fem", "Number": "Plur", "Case": "Dat"},
+    "AjSuFePlGe":    {POS: ADJ, "Degree": "Sup", "Gender": "Fem", "Number": "Plur", "Case": "Gen"},
+    "AjSuFePlNm":    {POS: ADJ, "Degree": "Sup", "Gender": "Fem", "Number": "Plur", "Case": "Nom"},
+    "AjSuFePlVo":    {POS: ADJ, "Degree": "Sup", "Gender": "Fem", "Number": "Plur", "Case": "Voc"},
+    "AjSuFeSgAc":    {POS: ADJ, "Degree": "Sup", "Gender": "Fem", "Number": "Sing", "Case": "Acc"},
+    "AjSuFeSgDa":    {POS: ADJ, "Degree": "Sup", "Gender": "Fem", "Number": "Sing", "Case": "Dat"},
+    "AjSuFeSgGe":    {POS: ADJ, "Degree": "Sup", "Gender": "Fem", "Number": "Sing", "Case": "Gen"},
+    "AjSuFeSgNm":    {POS: ADJ, "Degree": "Sup", "Gender": "Fem", "Number": "Sing", "Case": "Nom"},
+    "AjSuFeSgVo":    {POS: ADJ, "Degree": "Sup", "Gender": "Fem", "Number": "Sing", "Case": "Voc"},
+    "AjSuMaPlAc":    {POS: ADJ, "Degree": "Sup", "Gender": "Masc", "Number": "Plur", "Case": "Acc"},
+    "AjSuMaPlDa":    {POS: ADJ, "Degree": "Sup", "Gender": "Masc", "Number": "Plur", "Case": "Dat"},
+    "AjSuMaPlGe":    {POS: ADJ, "Degree": "Sup", "Gender": "Masc", "Number": "Plur", "Case": "Gen"},
+    "AjSuMaPlNm":    {POS: ADJ, "Degree": "Sup", "Gender": "Masc", "Number": "Plur", "Case": "Nom"},
+    "AjSuMaPlVo":    {POS: ADJ, "Degree": "Sup", "Gender": "Masc", "Number": "Plur", "Case": "Voc"},
+    "AjSuMaSgAc":    {POS: ADJ, "Degree": "Sup", "Gender": "Masc", "Number": "Sing", "Case": "Acc"},
+    "AjSuMaSgDa":    {POS: ADJ, "Degree": "Sup", "Gender": "Masc", "Number": "Sing", "Case": "Dat"},
+    "AjSuMaSgGe":    {POS: ADJ, "Degree": "Sup", "Gender": "Masc", "Number": "Sing", "Case": "Gen"},
+    "AjSuMaSgNm":    {POS: ADJ, "Degree": "Sup", "Gender": "Masc", "Number": "Sing", "Case": "Nom"},
+    "AjSuMaSgVo":    {POS: ADJ, "Degree": "Sup", "Gender": "Masc", "Number": "Sing", "Case": "Voc"},
+    "AjSuNePlAc":    {POS: ADJ, "Degree": "Sup", "Gender": "Neut", "Number": "Plur", "Case": "Acc"},
+    "AjSuNePlDa":    {POS: ADJ, "Degree": "Sup", "Gender": "Neut", "Number": "Plur", "Case": "Dat"},
+    "AjSuNePlGe":    {POS: ADJ, "Degree": "Sup", "Gender": "Neut", "Number": "Plur", "Case": "Gen"},
+    "AjSuNePlNm":    {POS: ADJ, "Degree": "Sup", "Gender": "Neut", "Number": "Plur", "Case": "Nom"},
+    "AjSuNePlVo":    {POS: ADJ, "Degree": "Sup", "Gender": "Neut", "Number": "Plur", "Case": "Voc"},
+    "AjSuNeSgAc":    {POS: ADJ, "Degree": "Sup", "Gender": "Neut", "Number": "Sing", "Case": "Acc"},
+    "AjSuNeSgDa":    {POS: ADJ, "Degree": "Sup", "Gender": "Neut", "Number": "Sing", "Case": "Dat"},
+    "AjSuNeSgGe":    {POS: ADJ, "Degree": "Sup", "Gender": "Neut", "Number": "Sing", "Case": "Gen"},
+    "AjSuNeSgNm":    {POS: ADJ, "Degree": "Sup", "Gender": "Neut", "Number": "Sing", "Case": "Nom"},
+    "AjSuNeSgVo":    {POS: ADJ, "Degree": "Sup", "Gender": "Neut", "Number": "Sing", "Case": "Voc"},
+    "AsPpPaFePlAc":    {POS: ADP, "Gender": "Fem", "Number": "Plur", "Case": "Acc"},
+    "AsPpPaFePlGe":    {POS: ADP, "Gender": "Fem", "Number": "Plur", "Case": "Gen"},
+    "AsPpPaFeSgAc":    {POS: ADP, "Gender": "Fem", "Number": "Sing", "Case": "Acc"},
+    "AsPpPaFeSgGe":    {POS: ADP, "Gender": "Fem", "Number": "Sing", "Case": "Gen"},
+    "AsPpPaMaPlAc":    {POS: ADP, "Gender": "Masc", "Number": "Plur", "Case": "Acc"},
+    "AsPpPaMaPlGe":    {POS: ADP, "Gender": "Masc", "Number": "Plur", "Case": "Gen"},
+    "AsPpPaMaSgAc":    {POS: ADP, "Gender": "Masc", "Number": "Sing", "Case": "Acc"},
+    "AsPpPaMaSgGe":    {POS: ADP, "Gender": "Masc", "Number": "Sing", "Case": "Gen"},
+    "AsPpPaNePlAc":    {POS: ADP, "Gender": "Neut", "Number": "Plur", "Case": "Acc"},
+    "AsPpPaNePlGe":    {POS: ADP, "Gender": "Neut", "Number": "Plur", "Case": "Gen"},
+    "AsPpPaNeSgAc":    {POS: ADP, "Gender": "Neut", "Number": "Sing", "Case": "Acc"},
+    "AsPpPaNeSgGe":    {POS: ADP, "Gender": "Neut", "Number": "Sing", "Case": "Gen"},
+    "AsPpSp":    {POS: ADP},
+    "AtDfFePlAc":    {POS: DET, "PronType": "Art", "Gender": "Fem", "Number": "Plur", "Case": "Acc", "Other":{"Definite": "Def"}},
+    "AtDfFePlGe":    {POS: DET, "PronType": "Art", "Gender": "Fem", "Number": "Plur", "Case": "Gen", "Other":{"Definite": "Def"}},
+    "AtDfFePlNm":    {POS: DET, "PronType": "Art", "Gender": "Fem", "Number": "Plur", "Case": "Nom", "Other":{"Definite": "Def"}},
+    "AtDfFeSgAc":    {POS: DET, "PronType": "Art", "Gender": "Fem", "Number": "Sing", "Case": "Acc", "Other":{"Definite": "Def"}},
+    "AtDfFeSgDa":    {POS: DET, "PronType": "Art", "Gender": "Fem", "Number": "Sing", "Case": "Dat", "Other":{"Definite": "Def"}},
+    "AtDfFeSgGe":    {POS: DET, "PronType": "Art", "Gender": "Fem", "Number": "Sing", "Case": "Gen", "Other":{"Definite": "Def"}},
+    "AtDfFeSgNm":    {POS: DET, "PronType": "Art", "Gender": "Fem", "Number": "Sing", "Case": "Nom", "Other":{"Definite": "Def"}},
+    "AtDfMaPlAc":    {POS: DET, "PronType": "Art", "Gender": "Masc", "Number": "Plur", "Case": "Acc", "Other":{"Definite": "Def"}},
+    "AtDfMaPlGe":    {POS: DET, "PronType": "Art", "Gender": "Masc", "Number": "Plur", "Case": "Gen", "Other":{"Definite": "Def"}},
+    "AtDfMaPlNm":    {POS: DET, "PronType": "Art", "Gender": "Masc", "Number": "Plur", "Case": "Nom", "Other":{"Definite": "Def"}},
+    "AtDfMaSgAc":    {POS: DET, "PronType": "Art", "Gender": "Masc", "Number": "Sing", "Case": "Acc", "Other":{"Definite": "Def"}},
+    "AtDfMaSgDa":    {POS: DET, "PronType": "Art", "Gender": "Masc", "Number": "Sing", "Case": "Dat", "Other":{"Definite": "Def"}},
+    "AtDfMaSgGe":    {POS: DET, "PronType": "Art", "Gender": "Masc", "Number": "Sing", "Case": "Gen", "Other":{"Definite": "Def"}},
+    "AtDfMaSgNm":    {POS: DET, "PronType": "Art", "Gender": "Masc", "Number": "Sing", "Case": "Nom", "Other":{"Definite": "Def"}},
+    "AtDfNePlAc":    {POS: DET, "PronType": "Art", "Gender": "Neut", "Number": "Plur", "Case": "Acc", "Other":{"Definite": "Def"}},
+    "AtDfNePlDa":    {POS: DET, "PronType": "Art", "Gender": "Neut", "Number": "Plur", "Case": "Dat", "Other":{"Definite": "Def"}},
+    "AtDfNePlGe":    {POS: DET, "PronType": "Art", "Gender": "Neut", "Number": "Plur", "Case": "Gen", "Other":{"Definite": "Def"}},
+    "AtDfNePlNm":    {POS: DET, "PronType": "Art", "Gender": "Neut", "Number": "Plur", "Case": "Nom", "Other":{"Definite": "Def"}},
+    "AtDfNeSgAc":    {POS: DET, "PronType": "Art", "Gender": "Neut", "Number": "Sing", "Case": "Acc", "Other":{"Definite": "Def"}},
+    "AtDfNeSgDa":    {POS: DET, "PronType": "Art", "Gender": "Neut", "Number": "Sing", "Case": "Dat", "Other":{"Definite": "Def"}},
+    "AtDfNeSgGe":    {POS: DET, "PronType": "Art", "Gender": "Neut", "Number": "Sing", "Case": "Gen", "Other":{"Definite": "Def"}},
+    "AtDfNeSgNm":    {POS: DET, "PronType": "Art", "Gender": "Neut", "Number": "Sing", "Case": "Nom", "Other":{"Definite": "Def"}},
+    "AtIdFeSgAc":    {POS: DET, "PronType": "Art", "Gender": "Fem", "Number": "Sing", "Case": "Acc", "Other":{"Definite": "Ind"}},
+    "AtIdFeSgDa":    {POS: DET, "PronType": "Art", "Gender": "Fem", "Number": "Sing", "Case": "Dat", "Other":{"Definite": "Ind"}},
+    "AtIdFeSgGe":    {POS: DET, "PronType": "Art", "Gender": "Fem", "Number": "Sing", "Case": "Gen", "Other":{"Definite": "Ind"}},
+    "AtIdFeSgNm":    {POS: DET, "PronType": "Art", "Gender": "Fem", "Number": "Sing", "Case": "Nom", "Other":{"Definite": "Ind"}},
+    "AtIdMaSgAc":    {POS: DET, "PronType": "Art", "Gender": "Masc", "Number": "Sing", "Case": "Acc", "Other":{"Definite": "Ind"}},
+    "AtIdMaSgGe":    {POS: DET, "PronType": "Art", "Gender": "Masc", "Number": "Sing", "Case": "Gen", "Other":{"Definite": "Ind"}},
+    "AtIdMaSgNm":    {POS: DET, "PronType": "Art", "Gender": "Masc", "Number": "Sing", "Case": "Nom", "Other":{"Definite": "Ind"}},
+    "AtIdNeSgAc":    {POS: DET, "PronType": "Art", "Gender": "Neut", "Number": "Sing", "Case": "Acc", "Other":{"Definite": "Ind"}},
+    "AtIdNeSgGe":    {POS: DET, "PronType": "Art", "Gender": "Neut", "Number": "Sing", "Case": "Gen", "Other":{"Definite": "Ind"}},
+    "AtIdNeSgNm":    {POS: DET, "PronType": "Art", "Gender": "Neut", "Number": "Sing", "Case": "Nom", "Other":{"Definite": "Ind"}},
+    "CjCo":    {POS: CCONJ},
+    "CjSb":    {POS: SCONJ},
+    "CPUNCT":    {POS: PUNCT},
+    "DATE":    {POS: NUM},
+    "DIG":    {POS: NUM},
+    "ENUM":    {POS: NUM},
+    "Ij":    {POS: INTJ},
+    "INIT":    {POS: SYM},
+    "NBABBR":    {POS: NOUN, "Abbr":"Yes"},
+    "NmAnFePlAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Acc"},
+    "NmAnFePlGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Gen"},
+    "NmAnFePlNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Nom"},
+    "NmAnFePlVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Voc"},
+    "NmAnFeSgAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Sing", "Case": "Acc"},
+    "NmAnFeSgGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Sing", "Case": "Gen"},
+    "NmAnFeSgNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Sing", "Case": "Nom"},
+    "NmAnFeSgVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Sing", "Case": "Voc"},
+    "NmAnMaPlAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Plur", "Case": "Acc"},
+    "NmAnMaPlGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Plur", "Case": "Gen"},
+    "NmAnMaPlNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Plur", "Case": "Nom"},
+    "NmAnMaPlVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Plur", "Case": "Voc"},
+    "NmAnMaSgAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Sing", "Case": "Acc"},
+    "NmAnMaSgGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Sing", "Case": "Gen"},
+    "NmAnMaSgNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Sing", "Case": "Nom"},
+    "NmAnMaSgVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Sing", "Case": "Voc"},
+    "NmAnNePlAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Plur", "Case": "Acc"},
+    "NmAnNePlGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Plur", "Case": "Gen"},
+    "NmAnNePlNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Plur", "Case": "Nom"},
+    "NmAnNePlVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Plur", "Case": "Voc"},
+    "NmAnNeSgAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Sing", "Case": "Acc"},
+    "NmAnNeSgGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Sing", "Case": "Gen"},
+    "NmAnNeSgNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Sing", "Case": "Nom"},
+    "NmAnNeSgVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Sing", "Case": "Voc"},
+    "NmAnXxXxXxAd":    {POS: NUM, "NumType": "Mult", "Gender": "Masc|Fem|Neut", "Number": "Sing|Plur", "Case": "Acc|Gen|Nom|Voc"},
+    "NmCdFePlAcAj":    {POS: NUM, "NumType": "Card", "Gender": "Fem", "Number": "Plur", "Case": "Acc"},
+    "NmCdFePlGeAj":    {POS: NUM, "NumType": "Card", "Gender": "Fem", "Number": "Plur", "Case": "Gen"},
+    "NmCdFePlNmAj":    {POS: NUM, "NumType": "Card", "Gender": "Fem", "Number": "Plur", "Case": "Nom"},
+    "NmCdFePlVoAj":    {POS: NUM, "NumType": "Card", "Gender": "Fem", "Number": "Plur", "Case": "Voc"},
+    "NmCdFeSgAcAj":    {POS: NUM, "NumType": "Card", "Gender": "Fem", "Number": "Sing", "Case": "Acc"},
+    "NmCdFeSgDaAj":    {POS: NUM, "NumType": "Card", "Gender": "Fem", "Number": "Sing", "Case": "Dat"},
+    "NmCdFeSgGeAj":    {POS: NUM, "NumType": "Card", "Gender": "Fem", "Number": "Sing", "Case": "Gen"},
+    "NmCdFeSgNmAj":    {POS: NUM, "NumType": "Card", "Gender": "Fem", "Number": "Sing", "Case": "Nom"},
+    "NmCdMaPlAcAj":    {POS: NUM, "NumType": "Card", "Gender": "Masc", "Number": "Plur", "Case": "Acc"},
+    "NmCdMaPlGeAj":    {POS: NUM, "NumType": "Card", "Gender": "Masc", "Number": "Plur", "Case": "Gen"},
+    "NmCdMaPlNmAj":    {POS: NUM, "NumType": "Card", "Gender": "Masc", "Number": "Plur", "Case": "Nom"},
+    "NmCdMaPlVoAj":    {POS: NUM, "NumType": "Card", "Gender": "Masc", "Number": "Plur", "Case": "Voc"},
+    "NmCdMaSgAcAj":    {POS: NUM, "NumType": "Card", "Gender": "Masc", "Number": "Sing", "Case": "Acc"},
+    "NmCdMaSgGeAj":    {POS: NUM, "NumType": "Card", "Gender": "Masc", "Number": "Sing", "Case": "Gen"},
+    "NmCdMaSgNmAj":    {POS: NUM, "NumType": "Card", "Gender": "Masc", "Number": "Sing", "Case": "Nom"},
+    "NmCdNePlAcAj":    {POS: NUM, "NumType": "Card", "Gender": "Neut", "Number": "Plur", "Case": "Acc"},
+    "NmCdNePlDaAj":    {POS: NUM, "NumType": "Card", "Gender": "Neut", "Number": "Plur", "Case": "Dat"},
+    "NmCdNePlGeAj":    {POS: NUM, "NumType": "Card", "Gender": "Neut", "Number": "Plur", "Case": "Gen"},
+    "NmCdNePlNmAj":    {POS: NUM, "NumType": "Card", "Gender": "Neut", "Number": "Plur", "Case": "Nom"},
+    "NmCdNePlVoAj":    {POS: NUM, "NumType": "Card", "Gender": "Neut", "Number": "Plur", "Case": "Voc"},
+    "NmCdNeSgAcAj":    {POS: NUM, "NumType": "Card", "Gender": "Neut", "Number": "Sing", "Case": "Acc"},
+    "NmCdNeSgGeAj":    {POS: NUM, "NumType": "Card", "Gender": "Neut", "Number": "Sing", "Case": "Gen"},
+    "NmCdNeSgNmAj":    {POS: NUM, "NumType": "Card", "Gender": "Neut", "Number": "Sing", "Case": "Nom"},
+    "NmCtFePlAcNo":    {POS: NUM, "NumType": "Sets", "Gender": "Fem", "Number": "Plur", "Case": "Acc"},
+    "NmCtFePlGeNo":    {POS: NUM, "NumType": "Sets", "Gender": "Fem", "Number": "Plur", "Case": "Gen"},
+    "NmCtFePlNmNo":    {POS: NUM, "NumType": "Sets", "Gender": "Fem", "Number": "Plur", "Case": "Nom"},
+    "NmCtFePlVoNo":    {POS: NUM, "NumType": "Sets", "Gender": "Fem", "Number": "Plur", "Case": "Voc"},
+    "NmCtFeSgAcNo":    {POS: NUM, "NumType": "Sets", "Gender": "Fem", "Number": "Sing", "Case": "Acc"},
+    "NmCtFeSgGeNo":    {POS: NUM, "NumType": "Sets", "Gender": "Fem", "Number": "Sing", "Case": "Gen"},
+    "NmCtFeSgNmNo":    {POS: NUM, "NumType": "Sets", "Gender": "Fem", "Number": "Sing", "Case": "Nom"},
+    "NmCtFeSgVoNo":    {POS: NUM, "NumType": "Sets", "Gender": "Fem", "Number": "Sing", "Case": "Voc"},
+    "NmMlFePlAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Acc"},
+    "NmMlFePlGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Gen"},
+    "NmMlFePlNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Nom"},
+    "NmMlFePlVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Voc"},
+    "NmMlFeSgAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Sing", "Case": "Acc"},
+    "NmMlFeSgGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Sing", "Case": "Gen"},
+    "NmMlFeSgNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Sing", "Case": "Nom"},
+    "NmMlFeSgVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Sing", "Case": "Voc"},
+    "NmMlMaPlAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Plur", "Case": "Acc"},
+    "NmMlMaPlGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Plur", "Case": "Gen"},
+    "NmMlMaPlNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Plur", "Case": "Nom"},
+    "NmMlMaPlVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Plur", "Case": "Voc"},
+    "NmMlMaSgAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Sing", "Case": "Acc"},
+    "NmMlMaSgGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Sing", "Case": "Gen"},
+    "NmMlMaSgNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Sing", "Case": "Nom"},
+    "NmMlMaSgVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Masc", "Number": "Sing", "Case": "Voc"},
+    "NmMlNePlAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Plur", "Case": "Acc"},
+    "NmMlNePlGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Plur", "Case": "Gen"},
+    "NmMlNePlNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Plur", "Case": "Nom"},
+    "NmMlNePlVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Plur", "Case": "Voc"},
+    "NmMlNeSgAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Sing", "Case": "Acc"},
+    "NmMlNeSgGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Sing", "Case": "Gen"},
+    "NmMlNeSgNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Sing", "Case": "Nom"},
+    "NmMlNeSgVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Neut", "Number": "Sing", "Case": "Voc"},
+    "NmMlXxXxXxAd":    {POS: NUM, "NumType": "Mult", "Gender": "Masc|Fem|Neut", "Number": "Sing|Plur", "Case": "Acc|Gen|Nom|Voc"},
+    "NmOdFePlAcAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Acc"},
+    "NmOdFePlGeAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Gen"},
+    "NmOdFePlNmAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Nom"},
+    "NmOdFePlVoAj":    {POS: NUM, "NumType": "Mult", "Gender": "Fem", "Number": "Plur", "Case": "Voc"},
+    "NmOdFeSgAcAj":    {POS: NUM, "NumType": "Ord", "Gender": "Fem", "Number": "Sing", "Case": "Acc"},
+    "NmOdFeSgGeAj":    {POS: NUM, "NumType": "Ord", "Gender": "Fem", "Number": "Sing", "Case": "Gen"},
+    "NmOdFeSgNmAj":    {POS: NUM, "NumType": "Ord", "Gender": "Fem", "Number": "Sing", "Case": "Nom"},
+    "NmOdFeSgVoAj":    {POS: NUM, "NumType": "Ord", "Gender": "Fem", "Number": "Sing", "Case": "Voc"},
+    "NmOdMaPlAcAj":    {POS: NUM, "NumType": "Ord", "Gender": "Masc", "Number": "Plur", "Case": "Acc"},
+    "NmOdMaPlGeAj":    {POS: NUM, "NumType": "Ord", "Gender": "Masc", "Number": "Plur", "Case": "Gen"},
+    "NmOdMaPlNmAj":    {POS: NUM, "NumType": "Ord", "Gender": "Masc", "Number": "Plur", "Case": "Nom"},
+    "NmOdMaPlVoAj":    {POS: NUM, "NumType": "Ord", "Gender": "Masc", "Number": "Plur", "Case": "Voc"},
+    "NmOdMaSgAcAj":    {POS: NUM, "NumType": "Ord", "Gender": "Masc", "Number": "Sing", "Case": "Acc"},
+    "NmOdMaSgGeAj":    {POS: NUM, "NumType": "Ord", "Gender": "Masc", "Number": "Sing", "Case": "Gen"},
+    "NmOdMaSgNmAj":    {POS: NUM, "NumType": "Ord", "Gender": "Masc", "Number": "Sing", "Case": "Nom"},
+    "NmOdMaSgVoAj":    {POS: NUM, "NumType": "Ord", "Gender": "Masc", "Number": "Sing", "Case": "Voc"},
+    "NmOdNePlAcAj":    {POS: NUM, "NumType": "Ord", "Gender": "Neut", "Number": "Plur", "Case": "Acc"},
+    "NmOdNePlGeAj":    {POS: NUM, "NumType": "Ord", "Gender": "Neut", "Number": "Plur", "Case": "Gen"},
+    "NmOdNePlNmAj":    {POS: NUM, "NumType": "Ord", "Gender": "Neut", "Number": "Plur", "Case": "Nom"},
+    "NmOdNePlVoAj":    {POS: NUM, "NumType": "Ord", "Gender": "Neut", "Number": "Plur", "Case": "Voc"},
+    "NmOdNeSgAcAj":    {POS: NUM, "NumType": "Ord", "Gender": "Neut", "Number": "Sing", "Case": "Acc"},
+    "NmOdNeSgGeAj":    {POS: NUM, "NumType": "Ord", "Gender": "Neut", "Number": "Sing", "Case": "Gen"},
+    "NmOdNeSgNmAj":    {POS: NUM, "NumType": "Ord", "Gender": "Neut", "Number": "Sing", "Case": "Nom"},
+    "NmOdNeSgVoAj":    {POS: NUM, "NumType": "Ord", "Gender": "Neut", "Number": "Sing", "Case": "Voc"},
+    "NoCmFePlAc":    {POS: NOUN, "Gender": "Fem", "Number": "Plur", "Case": "Acc"},
+    "NoCmFePlDa":    {POS: NOUN, "Gender": "Fem", "Number": "Plur", "Case": "Dat"},
+    "NoCmFePlGe":    {POS: NOUN, "Gender": "Fem", "Number": "Plur", "Case": "Gen"},
+    "NoCmFePlNm":    {POS: NOUN, "Gender": "Fem", "Number": "Plur", "Case": "Nom"},
+    "NoCmFePlVo":    {POS: NOUN, "Gender": "Fem", "Number": "Plur", "Case": "Voc"},
+    "NoCmFeSgAc":    {POS: NOUN, "Gender": "Fem", "Number": "Sing", "Case": "Acc"},
+    "NoCmFeSgDa":    {POS: NOUN, "Gender": "Fem", "Number": "Sing", "Case": "Dat"},
+    "NoCmFeSgGe":    {POS: NOUN, "Gender": "Fem", "Number": "Sing", "Case": "Gen"},
+    "NoCmFeSgNm":    {POS: NOUN, "Gender": "Fem", "Number": "Sing", "Case": "Nom"},
+    "NoCmFeSgVo":    {POS: NOUN, "Gender": "Fem", "Number": "Sing", "Case": "Voc"},
+    "NoCmMaPlAc":    {POS: NOUN, "Gender": "Masc", "Number": "Plur", "Case": "Acc"},
+    "NoCmMaPlDa":    {POS: NOUN, "Gender": "Masc", "Number": "Plur", "Case": "Dat"},
+    "NoCmMaPlGe":    {POS: NOUN, "Gender": "Masc", "Number": "Plur", "Case": "Gen"},
+    "NoCmMaPlNm":    {POS: NOUN, "Gender": "Masc", "Number": "Plur", "Case": "Nom"},
+    "NoCmMaPlVo":    {POS: NOUN, "Gender": "Masc", "Number": "Plur", "Case": "Voc"},
+    "NoCmMaSgAc":    {POS: NOUN, "Gender": "Masc", "Number": "Sing", "Case": "Acc"},
+    "NoCmMaSgDa":    {POS: NOUN, "Gender": "Masc", "Number": "Sing", "Case": "Dat"},
+    "NoCmMaSgGe":    {POS: NOUN, "Gender": "Masc", "Number": "Sing", "Case": "Gen"},
+    "NoCmMaSgNm":    {POS: NOUN, "Gender": "Masc", "Number": "Sing", "Case": "Nom"},
+    "NoCmMaSgVo":    {POS: NOUN, "Gender": "Masc", "Number": "Sing", "Case": "Voc"},
+    "NoCmNePlAc":    {POS: NOUN, "Gender": "Neut", "Number": "Plur", "Case": "Acc"},
+    "NoCmNePlDa":    {POS: NOUN, "Gender": "Neut", "Number": "Plur", "Case": "Dat"},
+    "NoCmNePlGe":    {POS: NOUN, "Gender": "Neut", "Number": "Plur", "Case": "Gen"},
+    "NoCmNePlNm":    {POS: NOUN, "Gender": "Neut", "Number": "Plur", "Case": "Nom"},
+    "NoCmNePlVo":    {POS: NOUN, "Gender": "Neut", "Number": "Plur", "Case": "Voc"},
+    "NoCmNeSgAc":    {POS: NOUN, "Gender": "Neut", "Number": "Sing", "Case": "Acc"},
+    "NoCmNeSgDa":    {POS: NOUN, "Gender": "Neut", "Number": "Sing", "Case": "Dat"},
+    "NoCmNeSgGe":    {POS: NOUN, "Gender": "Neut", "Number": "Sing", "Case": "Gen"},
+    "NoCmNeSgNm":    {POS: NOUN, "Gender": "Neut", "Number": "Sing", "Case": "Nom"},
+    "NoCmNeSgVo":    {POS: NOUN, "Gender": "Neut", "Number": "Sing", "Case": "Voc"},
+    "NoPrFePlAc":    {POS: PROPN, "Gender": "Fem", "Number": "Plur", "Case": "Acc"},
+    "NoPrFePlDa":    {POS: PROPN, "Gender": "Fem", "Number": "Plur", "Case": "Dat"},
+    "NoPrFePlGe":    {POS: PROPN, "Gender": "Fem", "Number": "Plur", "Case": "Gen"},
+    "NoPrFePlNm":    {POS: PROPN, "Gender": "Fem", "Number": "Plur", "Case": "Nom"},
+    "NoPrFePlVo":    {POS: PROPN, "Gender": "Fem", "Number": "Plur", "Case": "Voc"},
+    "NoPrFeSgAc":    {POS: PROPN, "Gender": "Fem", "Number": "Sing", "Case": "Acc"},
+    "NoPrFeSgDa":    {POS: PROPN, "Gender": "Fem", "Number": "Sing", "Case": "Dat"},
+    "NoPrFeSgGe":    {POS: PROPN, "Gender": "Fem", "Number": "Sing", "Case": "Gen"},
+    "NoPrFeSgNm":    {POS: PROPN, "Gender": "Fem", "Number": "Sing", "Case": "Nom"},
+    "NoPrFeSgVo":    {POS: PROPN, "Gender": "Fem", "Number": "Sing", "Case": "Voc"},
+    "NoPrMaPlAc":    {POS: PROPN, "Gender": "Masc", "Number": "Plur", "Case": "Acc"},
+    "NoPrMaPlGe":    {POS: PROPN, "Gender": "Masc", "Number": "Plur", "Case": "Gen"},
+    "NoPrMaPlNm":    {POS: PROPN, "Gender": "Masc", "Number": "Plur", "Case": "Nom"},
+    "NoPrMaPlVo":    {POS: PROPN, "Gender": "Masc", "Number": "Plur", "Case": "Voc"},
+    "NoPrMaSgAc":    {POS: PROPN, "Gender": "Masc", "Number": "Sing", "Case": "Acc"},
+    "NoPrMaSgDa":    {POS: PROPN, "Gender": "Masc", "Number": "Sing", "Case": "Dat"},
+    "NoPrMaSgGe":    {POS: PROPN, "Gender": "Masc", "Number": "Sing", "Case": "Gen"},
+    "NoPrMaSgNm":    {POS: PROPN, "Gender": "Masc", "Number": "Sing", "Case": "Nom"},
+    "NoPrMaSgVo":    {POS: PROPN, "Gender": "Masc", "Number": "Sing", "Case": "Voc"},
+    "NoPrNePlAc":    {POS: PROPN, "Gender": "Neut", "Number": "Plur", "Case": "Acc"},
+    "NoPrNePlGe":    {POS: PROPN, "Gender": "Neut", "Number": "Plur", "Case": "Gen"},
+    "NoPrNePlNm":    {POS: PROPN, "Gender": "Neut", "Number": "Plur", "Case": "Nom"},
+    "NoPrNeSgAc":    {POS: PROPN, "Gender": "Neut", "Number": "Sing", "Case": "Acc"},
+    "NoPrNeSgGe":    {POS: PROPN, "Gender": "Neut", "Number": "Sing", "Case": "Gen"},
+    "NoPrNeSgNm":    {POS: PROPN, "Gender": "Neut", "Number": "Sing", "Case": "Nom"},
+    "OPUNCT":    {POS: PUNCT},
+    "PnDfFe03PlAcXx":    {POS: PRON, "PronType": "", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnDfFe03SgAcXx":    {POS: PRON, "PronType": "", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnDfMa03PlGeXx":    {POS: PRON, "PronType": "", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnDmFe03PlAcXx":    {POS: PRON, "PronType": "Dem", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnDmFe03PlGeXx":    {POS: PRON, "PronType": "Dem", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnDmFe03PlNmXx":    {POS: PRON, "PronType": "Dem", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnDmFe03SgAcXx":    {POS: PRON, "PronType": "Dem", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnDmFe03SgDaXx":    {POS: PRON, "PronType": "Dem", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Dat"},
+    "PnDmFe03SgGeXx":    {POS: PRON, "PronType": "Dem", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnDmFe03SgNmXx":    {POS: PRON, "PronType": "Dem", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnDmMa03PlAcXx":    {POS: PRON, "PronType": "Dem", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnDmMa03PlDaXx":    {POS: PRON, "PronType": "Dem", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Dat"},
+    "PnDmMa03PlGeXx":    {POS: PRON, "PronType": "Dem", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnDmMa03PlNmXx":    {POS: PRON, "PronType": "Dem", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnDmMa03SgAcXx":    {POS: PRON, "PronType": "Dem", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnDmMa03SgGeXx":    {POS: PRON, "PronType": "Dem", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnDmMa03SgNmXx":    {POS: PRON, "PronType": "Dem", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnDmNe03PlAcXx":    {POS: PRON, "PronType": "Dem", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnDmNe03PlDaXx":    {POS: PRON, "PronType": "Dem", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Dat"},
+    "PnDmNe03PlGeXx":    {POS: PRON, "PronType": "Dem", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnDmNe03PlNmXx":    {POS: PRON, "PronType": "Dem", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnDmNe03SgAcXx":    {POS: PRON, "PronType": "Dem", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnDmNe03SgDaXx":    {POS: PRON, "PronType": "Dem", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Dat"},
+    "PnDmNe03SgGeXx":    {POS: PRON, "PronType": "Dem", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnDmNe03SgNmXx":    {POS: PRON, "PronType": "Dem", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnIdFe03PlAcXx":    {POS: PRON, "PronType": "Ind", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnIdFe03PlGeXx":    {POS: PRON, "PronType": "Ind", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnIdFe03PlNmXx":    {POS: PRON, "PronType": "Ind", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnIdFe03SgAcXx":    {POS: PRON, "PronType": "Ind", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnIdFe03SgGeXx":    {POS: PRON, "PronType": "Ind", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnIdFe03SgNmXx":    {POS: PRON, "PronType": "Ind", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnIdMa03PlAcXx":    {POS: PRON, "PronType": "Ind", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnIdMa03PlGeXx":    {POS: PRON, "PronType": "Ind", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnIdMa03PlNmXx":    {POS: PRON, "PronType": "Ind", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnIdMa03SgAcXx":    {POS: PRON, "PronType": "Ind", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnIdMa03SgGeXx":    {POS: PRON, "PronType": "Ind", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnIdMa03SgNmXx":    {POS: PRON, "PronType": "Ind", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnIdNe03PlAcXx":    {POS: PRON, "PronType": "Ind", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnIdNe03PlGeXx":    {POS: PRON, "PronType": "Ind", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnIdNe03PlNmXx":    {POS: PRON, "PronType": "Ind", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnIdNe03SgAcXx":    {POS: PRON, "PronType": "Ind", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnIdNe03SgDaXx":    {POS: PRON, "PronType": "Ind", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Dat"},
+    "PnIdNe03SgGeXx":    {POS: PRON, "PronType": "Ind", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnIdNe03SgNmXx":    {POS: PRON, "PronType": "Ind", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnIrFe03PlAcXx":    {POS: PRON, "PronType": "Int", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnIrFe03PlGeXx":    {POS: PRON, "PronType": "Int", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnIrFe03PlNmXx":    {POS: PRON, "PronType": "Int", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnIrFe03SgAcXx":    {POS: PRON, "PronType": "Int", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnIrFe03SgGeXx":    {POS: PRON, "PronType": "Int", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnIrFe03SgNmXx":    {POS: PRON, "PronType": "Int", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnIrMa03PlAcXx":    {POS: PRON, "PronType": "Int", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnIrMa03PlGeXx":    {POS: PRON, "PronType": "Int", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnIrMa03PlNmXx":    {POS: PRON, "PronType": "Int", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnIrMa03SgAcXx":    {POS: PRON, "PronType": "Int", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnIrMa03SgGeXx":    {POS: PRON, "PronType": "Int", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnIrMa03SgNmXx":    {POS: PRON, "PronType": "Int", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnIrNe03PlAcXx":    {POS: PRON, "PronType": "Int", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnIrNe03PlGeXx":    {POS: PRON, "PronType": "Int", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnIrNe03PlNmXx":    {POS: PRON, "PronType": "Int", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnIrNe03SgAcXx":    {POS: PRON, "PronType": "Int", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnIrNe03SgGeXx":    {POS: PRON, "PronType": "Int", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnIrNe03SgNmXx":    {POS: PRON, "PronType": "Int", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnPeFe01PlAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "1", "Number": "Plur", "Case": "Acc"},
+    "PnPeFe01PlAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "1", "Number": "Plur", "Case": "Acc"},
+    "PnPeFe01PlGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "1", "Number": "Plur", "Case": "Gen"},
+    "PnPeFe01PlNmSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "1", "Number": "Plur", "Case": "Nom"},
+    "PnPeFe01SgAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "1", "Number": "Sing", "Case": "Acc"},
+    "PnPeFe01SgAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "1", "Number": "Sing", "Case": "Acc"},
+    "PnPeFe01SgGeSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "1", "Number": "Sing", "Case": "Gen"},
+    "PnPeFe01SgGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "1", "Number": "Sing", "Case": "Gen"},
+    "PnPeFe01SgNmSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "1", "Number": "Sing", "Case": "Nom"},
+    "PnPeFe02PlAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "2", "Number": "Plur", "Case": "Acc"},
+    "PnPeFe02PlAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "2", "Number": "Plur", "Case": "Acc"},
+    "PnPeFe02PlGeSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "2", "Number": "Plur", "Case": "Gen"},
+    "PnPeFe02PlGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "2", "Number": "Plur", "Case": "Gen"},
+    "PnPeFe02PlNmSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "2", "Number": "Plur", "Case": "Nom"},
+    "PnPeFe02SgAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "2", "Number": "Sing", "Case": "Acc"},
+    "PnPeFe02SgAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "2", "Number": "Sing", "Case": "Acc"},
+    "PnPeFe02SgGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "2", "Number": "Sing", "Case": "Gen"},
+    "PnPeFe02SgNmSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "2", "Number": "Sing", "Case": "Nom"},
+    "PnPeFe03PlAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnPeFe03PlAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnPeFe03PlGeSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnPeFe03PlGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnPeFe03PlNmSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnPeFe03SgAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnPeFe03SgAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnPeFe03SgGeSt":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnPeFe03SgGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnPeMa01PlAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "1", "Number": "Plur", "Case": "Acc"},
+    "PnPeMa01PlAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "1", "Number": "Plur", "Case": "Acc"},
+    "PnPeMa01PlDaSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "1", "Number": "Plur", "Case": "Dat"},
+    "PnPeMa01PlGeSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "1", "Number": "Plur", "Case": "Gen"},
+    "PnPeMa01PlGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "1", "Number": "Plur", "Case": "Gen"},
+    "PnPeMa01PlNmSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "1", "Number": "Plur", "Case": "Nom"},
+    "PnPeMa01SgAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "1", "Number": "Sing", "Case": "Acc"},
+    "PnPeMa01SgAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "1", "Number": "Sing", "Case": "Acc"},
+    "PnPeMa01SgGeSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "1", "Number": "Sing", "Case": "Gen"},
+    "PnPeMa01SgGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "1", "Number": "Sing", "Case": "Gen"},
+    "PnPeMa01SgNmSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "1", "Number": "Sing", "Case": "Nom"},
+    "PnPeMa02PlAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "2", "Number": "Plur", "Case": "Acc"},
+    "PnPeMa02PlAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "2", "Number": "Plur", "Case": "Acc"},
+    "PnPeMa02PlGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "2", "Number": "Plur", "Case": "Gen"},
+    "PnPeMa02PlNmSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "2", "Number": "Plur", "Case": "Nom"},
+    "PnPeMa02PlVoSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "2", "Number": "Plur", "Case": "Voc"},
+    "PnPeMa02SgAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "2", "Number": "Sing", "Case": "Acc"},
+    "PnPeMa02SgAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "2", "Number": "Sing", "Case": "Acc"},
+    "PnPeMa02SgGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "2", "Number": "Sing", "Case": "Gen"},
+    "PnPeMa02SgNmSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "2", "Number": "Sing", "Case": "Nom"},
+    "PnPeMa03PlAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnPeMa03PlGeSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnPeMa03PlGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnPeMa03PlNmSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnPeMa03SgAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnPeMa03SgAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnPeMa03SgGeSt":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnPeMa03SgGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnPeMa03SgNmWe":    {POS: PRON, "PronType": "Prs", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnPeNe03PlAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnPeNe03PlGeSt":    {POS: PRON, "PronType": "Prs", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnPeNe03PlGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnPeNe03SgAcSt":    {POS: PRON, "PronType": "Prs", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnPeNe03SgAcWe":    {POS: PRON, "PronType": "Prs", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnPeNe03SgGeSt":    {POS: PRON, "PronType": "Prs", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnPeNe03SgGeWe":    {POS: PRON, "PronType": "Prs", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnPoFe01PlGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Fem", "Person": "1", "Number": "Plur", "Case": "Gen"},
+    "PnPoFe01SgGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Fem", "Person": "1", "Number": "Sing", "Case": "Gen"},
+    "PnPoFe02PlGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Fem", "Person": "2", "Number": "Plur", "Case": "Gen"},
+    "PnPoFe02SgGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Fem", "Person": "2", "Number": "Sing", "Case": "Gen"},
+    "PnPoFe03PlGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnPoFe03SgGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnPoMa01PlGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Masc", "Person": "1", "Number": "Plur", "Case": "Gen"},
+    "PnPoMa01SgGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Masc", "Person": "1", "Number": "Sing", "Case": "Gen"},
+    "PnPoMa02PlGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Masc", "Person": "2", "Number": "Plur", "Case": "Gen"},
+    "PnPoMa02SgGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Masc", "Person": "2", "Number": "Sing", "Case": "Gen"},
+    "PnPoMa03PlGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnPoMa03SgGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnPoNe03PlGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnPoNe03SgGeXx":    {POS: PRON, "Poss": "Yes", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnReFe03PlAcXx":    {POS: PRON, "PronType": "Rel", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnReFe03PlGeXx":    {POS: PRON, "PronType": "Rel", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnReFe03PlNmXx":    {POS: PRON, "PronType": "Rel", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnReFe03SgAcXx":    {POS: PRON, "PronType": "Rel", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnReFe03SgGeXx":    {POS: PRON, "PronType": "Rel", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnReFe03SgNmXx":    {POS: PRON, "PronType": "Rel", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnReMa03PlAcXx":    {POS: PRON, "PronType": "Rel", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnReMa03PlGeXx":    {POS: PRON, "PronType": "Rel", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnReMa03PlNmXx":    {POS: PRON, "PronType": "Rel", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnReMa03SgAcXx":    {POS: PRON, "PronType": "Rel", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnReMa03SgGeXx":    {POS: PRON, "PronType": "Rel", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnReMa03SgNmXx":    {POS: PRON, "PronType": "Rel", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnReNe03PlAcXx":    {POS: PRON, "PronType": "Rel", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnReNe03PlGeXx":    {POS: PRON, "PronType": "Rel", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnReNe03PlNmXx":    {POS: PRON, "PronType": "Rel", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnReNe03SgAcXx":    {POS: PRON, "PronType": "Rel", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnReNe03SgGeXx":    {POS: PRON, "PronType": "Rel", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnReNe03SgNmXx":    {POS: PRON, "PronType": "Rel", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnRiFe03PlAcXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnRiFe03PlGeXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnRiFe03PlNmXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Fem", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnRiFe03SgAcXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnRiFe03SgGeXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnRiFe03SgNmXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Fem", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnRiMa03PlAcXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnRiMa03PlGeXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnRiMa03PlNmXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Masc", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnRiMa03SgAcXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnRiMa03SgGeXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnRiMa03SgNmXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Masc", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PnRiNe03PlAcXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Acc"},
+    "PnRiNe03PlGeXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Gen"},
+    "PnRiNe03PlNmXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Neut", "Person": "3", "Number": "Plur", "Case": "Nom"},
+    "PnRiNe03SgAcXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Acc"},
+    "PnRiNe03SgGeXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Gen"},
+    "PnRiNe03SgNmXx":    {POS: PRON, "PronType": "Ind,Rel", "Gender": "Neut", "Person": "3", "Number": "Sing", "Case": "Nom"},
+    "PTERM_P":    {POS: PUNCT},
+    "PtFu":    {POS: PART},
+    "PtNg":    {POS: PART},
+    "PtOt":    {POS: PART},
+    "PtSj":    {POS: PART},
+    "Pu":    {POS: SYM},
+    "PUNCT":    {POS: PUNCT},
+    "RgAbXx":    {POS: X},
+    "RgAnXx":    {POS: X},
+    "RgFwOr":    {POS: X, "Foreign": "Yes"},
+    "RgFwTr":    {POS: X, "Foreign": "Yes"},
+    "RgSyXx":    {POS: SYM},
+    "VbIsIdPa03SgXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbIsIdPa03SgXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbIsIdPa03SgXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbIsIdPa03SgXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbIsIdPr03SgXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbIsIdPr03SgXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbIsIdXx03SgXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbIsIdXx03SgXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbIsNfXxXxXxXxPeAvXx":    {POS: VERB, "VerbForm": "Inf", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing|Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa01PlXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "1", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa01PlXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "1", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa01PlXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "1", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa01PlXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "1", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa01SgXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "1", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa01SgXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "1", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa01SgXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa01SgXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "1", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa02PlXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa02PlXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa02PlXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa02PlXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa02SgXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa02SgXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa02SgXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa02SgXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa03PlXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa03PlXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp", "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa03PlXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf", "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa03PlXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa03SgXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa03SgXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa03SgXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPa03SgXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr01PlXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "1", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr01PlXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "1", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr01SgXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "1", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr01SgXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "1", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr02PlXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr02PlXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr02SgXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr02SgXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr03PlXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "3", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr03PlXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "3", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr03SgXxIpAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdPr03SgXxIpPvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx01PlXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "1", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx01PlXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "1", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx01SgXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "1", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx01SgXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "1", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx02PlXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx02PlXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx02SgXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx02SgXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx03PlXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "3", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx03PlXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "3", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx03SgXxPeAvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnIdXx03SgXxPePvXx":    {POS: VERB, "VerbForm": "Fin", "Mood": "Ind", "Tense": "Pres|Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnMpXx02PlXxIpAvXx":    {POS: VERB, "VerbForm": "", "Mood": "Imp", "Tense": "Pres|Past", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnMpXx02PlXxIpPvXx":    {POS: VERB, "VerbForm": "", "Mood": "Imp", "Tense": "Pres|Past", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnMpXx02PlXxPeAvXx":    {POS: VERB, "VerbForm": "", "Mood": "Imp", "Tense": "Pres|Past", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnMpXx02PlXxPePvXx":    {POS: VERB, "VerbForm": "", "Mood": "Imp", "Tense": "Pres|Past", "Person": "2", "Number": "Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnMpXx02SgXxIpAvXx":    {POS: VERB, "VerbForm": "", "Mood": "Imp", "Tense": "Pres|Past", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnMpXx02SgXxIpPvXx":    {POS: VERB, "VerbForm": "", "Mood": "Imp", "Tense": "Pres|Past", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnMpXx02SgXxPeAvXx":    {POS: VERB, "VerbForm": "", "Mood": "Imp", "Tense": "Pres|Past", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnMpXx02SgXxPePvXx":    {POS: VERB, "VerbForm": "", "Mood": "Imp", "Tense": "Pres|Past", "Person": "2", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnMpXx03SgXxIpPvXx":    {POS: VERB, "VerbForm": "", "Mood": "Imp", "Tense": "Pres|Past", "Person": "3", "Number": "Sing", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnNfXxXxXxXxPeAvXx":    {POS: VERB, "VerbForm": "Inf", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing|Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnNfXxXxXxXxPePvXx":    {POS: VERB, "VerbForm": "Inf", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing|Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnPpPrXxXxXxIpAvXx":    {POS: VERB, "VerbForm": "Conv", "Mood": "", "Tense": "Pres", "Person": "1|2|3", "Number": "Sing|Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"},
+    "VbMnPpXxXxPlFePePvAc":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Fem", "Aspect": "Perf" , "Voice": "Pass", "Case": "Acc"},
+    "VbMnPpXxXxPlFePePvGe":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Fem", "Aspect": "Perf" , "Voice": "Pass", "Case": "Gen"},
+    "VbMnPpXxXxPlFePePvNm":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Fem", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom"},
+    "VbMnPpXxXxPlFePePvVo":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Fem", "Aspect": "Perf" , "Voice": "Pass", "Case": "Voc"},
+    "VbMnPpXxXxPlMaPePvAc":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Masc", "Aspect": "Perf" , "Voice": "Pass", "Case": "Acc"},
+    "VbMnPpXxXxPlMaPePvGe":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Masc", "Aspect": "Perf" , "Voice": "Pass", "Case": "Gen"},
+    "VbMnPpXxXxPlMaPePvNm":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Masc", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom"},
+    "VbMnPpXxXxPlMaPePvVo":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Masc", "Aspect": "Perf" , "Voice": "Pass", "Case": "Voc"},
+    "VbMnPpXxXxPlNePePvAc":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Acc"},
+    "VbMnPpXxXxPlNePePvGe":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Gen"},
+    "VbMnPpXxXxPlNePePvNm":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom"},
+    "VbMnPpXxXxPlNePePvVo":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Plur", "Gender": "Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Voc"},
+    "VbMnPpXxXxSgFePePvAc":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Fem", "Aspect": "Perf" , "Voice": "Pass", "Case": "Acc"},
+    "VbMnPpXxXxSgFePePvGe":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Fem", "Aspect": "Perf" , "Voice": "Pass", "Case": "Gen"},
+    "VbMnPpXxXxSgFePePvNm":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Fem", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom"},
+    "VbMnPpXxXxSgFePePvVo":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Fem", "Aspect": "Perf" , "Voice": "Pass", "Case": "Voc"},
+    "VbMnPpXxXxSgMaPePvAc":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Masc", "Aspect": "Perf" , "Voice": "Pass", "Case": "Acc"},
+    "VbMnPpXxXxSgMaPePvGe":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Masc", "Aspect": "Perf" , "Voice": "Pass", "Case": "Gen"},
+    "VbMnPpXxXxSgMaPePvNm":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Masc", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom"},
+    "VbMnPpXxXxSgMaPePvVo":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Masc", "Aspect": "Perf" , "Voice": "Pass", "Case": "Voc"},
+    "VbMnPpXxXxSgNePePvAc":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Acc"},
+    "VbMnPpXxXxSgNePePvGe":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Gen"},
+    "VbMnPpXxXxSgNePePvNm":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Nom"},
+    "VbMnPpXxXxSgNePePvVo":    {POS: VERB, "VerbForm": "Part", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing", "Gender": "Neut", "Aspect": "Perf" , "Voice": "Pass", "Case": "Voc"},
+    "VbMnPpXxXxXxXxIpAvXx":    {POS: VERB, "VerbForm": "Conv", "Mood": "", "Tense": "Pres|Past", "Person": "1|2|3", "Number": "Sing|Plur", "Gender": "Masc|Fem|Neut", "Aspect": "Imp" , "Voice": "Act", "Case": "Nom|Gen|Dat|Acc|Voc"}
+}
+

spacy/lang/el/tag_map_general.py

@@ -0,0 +1,24 @@
+# -*- coding: utf-8 -*-
+
+from __future__ import unicode_literals
+from ...symbols import POS, PUNCT, SYM, ADJ, CCONJ, SCONJ, NUM, DET, ADV, ADP, X, VERB
+from ...symbols import NOUN, PROPN, PART, INTJ,SPACE,PRON
+
+TAG_MAP = {
+    "Adjective":    {POS:ADJ},
+    "Adposition":    {POS:ADP},
+    "Adverb":    {POS:ADV},
+    "Conjuction_Coordinating":    {POS:CCONJ},
+    "Conjuction_Subordinating":    {POS:SCONJ},
+    "Determiner":    {POS:DET},
+    "Interjection":    {POS:INTJ},
+    "Noun_Common":    {POS:NOUN},
+    "Noun_Proper":    {POS:PROPN},
+    "Numeral":    {POS:NUM},
+    "Other":    {POS:X},
+    "Particle":    {POS:PART},
+    "Pronoun":    {POS:PRON},
+    "Punctuation":    {POS:PUNCT},
+    "Symbol":    {POS:SYM},
+    "Verb":    {POS:VERB}
+}

spacy/lang/el/tokenizer_exceptions.py

@@ -0,0 +1,382 @@
+# -*- coding: utf-8 -*-
+
+from __future__ import unicode_literals
+
+from ...symbols import ORTH, LEMMA, TAG, NORM, ADP, DET
+
+_exc = {}
+
+for token in ["Απ'", "ΑΠ'", "αφ'", "Αφ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "από", NORM: "από"}
+    ]
+
+for token in ["Αλλ'", "αλλ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "αλλά", NORM: "αλλά"}
+    ]
+
+for token in ["παρ'", "Παρ'", "ΠΑΡ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "παρά", NORM: "παρά"}
+    ]
+
+for token in ["καθ'", "Καθ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "κάθε", NORM: "κάθε"}
+    ]
+
+for token in ["κατ'", "Κατ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "κατά", NORM: "κατά"}
+    ]
+
+for token in ["'ΣΟΥΝ", "'ναι", "'ταν", "'τανε", "'μαστε", "'μουνα", "'μουν"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "είμαι", NORM: "είμαι"}
+    ]
+
+for token in ["Επ'", "επ'", "εφ'", "Εφ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "επί", NORM: "επί"}
+    ]
+
+for token in ["Δι'", "δι'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "δια", NORM: "δια"}
+    ]
+
+for token in ["'χουν", "'χουμε", "'χαμε", "'χα", "'χε", "'χεις", "'χει"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "έχω", NORM: "έχω"}
+    ]
+
+for token in ["υπ'", "Υπ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "υπό", NORM: "υπό"}
+    ]
+
+for token in ["Μετ'", "ΜΕΤ'", "'μετ"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "μετά", NORM: "μετά"}
+    ]
+
+for token in ["Μ'", "μ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "με", NORM: "με"}
+    ]
+
+for token in ["Γι'", "ΓΙ'", "γι'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "για", NORM: "για"}
+    ]
+
+for token in ["Σ'", "σ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "σε", NORM: "σε"}
+    ]
+
+for token in ["Θ'", "θ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "θα", NORM: "θα"}
+    ]
+
+for token in ["Ν'", "ν'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "να", NORM: "να"}
+    ]
+
+for token in ["Τ'", "τ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "να", NORM: "να"}
+    ]
+
+for token in ["'γω", "'σένα", "'μεις"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "εγώ", NORM: "εγώ"}
+    ]
+
+for token in ["Τ'", "τ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "το", NORM: "το"}
+    ]
+
+for token in ["Φέρ'", "Φερ'", "φέρ'", "φερ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "φέρνω", NORM: "φέρνω"}
+    ]
+
+for token in ["'ρθούνε", "'ρθουν", "'ρθει", "'ρθεί", "'ρθε", "'ρχεται"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "έρχομαι", NORM: "έρχομαι"}
+    ]
+
+for token in ["'πανε", "'λεγε", "'λεγαν", "'πε", "'λεγα"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "λέγω", NORM: "λέγω"}
+    ]
+
+for token in ["Πάρ'", "πάρ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "παίρνω", NORM: "παίρνω"}
+    ]
+
+for token in ["μέσ'", "Μέσ'", "μεσ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "μέσα", NORM: "μέσα"}
+    ]
+
+for token in ["Δέσ'", "Δεσ'", "δεσ'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "δένω", NORM: "δένω"}
+    ]
+
+for token in ["'κανε", "Κάν'"]:
+    _exc[token] = [
+        {ORTH: token, LEMMA: "κάνω", NORM: "κάνω"}
+    ]
+
+_other_exc = {
+
+    "κι": [
+        {ORTH: "κι", LEMMA: "και", NORM: "και"},
+    ],
+
+    "Παίξ'": [
+        {ORTH: "Παίξ'", LEMMA: "παίζω", NORM: "παίζω"},
+    ],
+
+    "Αντ'": [
+        {ORTH: "Αντ'", LEMMA: "αντί", NORM: "αντί"},
+    ],
+
+    "ολ'": [
+        {ORTH: "ολ'", LEMMA: "όλος", NORM: "όλος"},
+    ],
+
+    "ύστερ'": [
+        {ORTH: "ύστερ'", LEMMA: "ύστερα", NORM: "ύστερα"},
+    ],
+
+    "'πρεπε": [
+        {ORTH: "'πρεπε", LEMMA: "πρέπει", NORM: "πρέπει"},
+    ],
+
+    "Δύσκολ'": [
+        {ORTH: "Δύσκολ'", LEMMA: "δύσκολος", NORM: "δύσκολος"},
+    ],
+
+    "'θελα": [
+        {ORTH: "'θελα", LEMMA: "θέλω", NORM: "θέλω"},
+    ],
+
+    "'γραφα": [
+        {ORTH: "'γραφα", LEMMA: "γράφω", NORM: "γράφω"},
+    ],
+
+    "'παιρνα": [
+        {ORTH: "'παιρνα", LEMMA: "παίρνω", NORM: "παίρνω"},
+    ],
+
+    "'δειξε": [
+        {ORTH: "'δειξε", LEMMA: "δείχνω", NORM: "δείχνω"},
+    ],
+
+    "όμουρφ'": [
+        {ORTH: "όμουρφ'", LEMMA: "όμορφος", NORM: "όμορφος"},
+    ],
+
+    "κ'τσή": [
+        {ORTH: "κ'τσή", LEMMA: "κουτσός", NORM: "κουτσός"},
+    ],
+
+    "μηδ'": [
+        {ORTH: "μηδ'", LEMMA: "μήδε", NORM: "μήδε"},
+    ],
+
+    "'ξομολογήθηκε": [
+        {ORTH: "'ξομολογήθηκε", LEMMA: "εξομολογούμαι", NORM: "εξομολογούμαι"},
+    ],
+
+    "'μας": [
+        {ORTH: "'μας", LEMMA: "εμάς", NORM: "εμάς"},
+    ],
+
+    "'ξερες": [
+        {ORTH: "'ξερες", LEMMA: "ξέρω", NORM: "ξέρω"},
+    ],
+
+    "έφθασ'": [
+        {ORTH: "έφθασ'", LEMMA: "φθάνω", NORM: "φθάνω"},
+    ],
+
+    "εξ'": [
+        {ORTH: "εξ'", LEMMA: "εκ", NORM: "εκ"},
+    ],
+
+    "δώσ'": [
+        {ORTH: "δώσ'", LEMMA: "δίνω", NORM: "δίνω"},
+    ],
+
+    "τίποτ'": [
+        {ORTH: "τίποτ'", LEMMA: "τίποτα", NORM: "τίποτα"},
+    ],
+
+    "Λήξ'": [
+        {ORTH: "Λήξ'", LEMMA: "λήγω", NORM: "λήγω"},
+    ],
+
+    "άσ'": [
+        {ORTH: "άσ'", LEMMA: "αφήνω", NORM: "αφήνω"},
+    ],
+
+    "Στ'": [
+        {ORTH: "Στ'", LEMMA: "στο", NORM: "στο"},
+
+    ],
+
+    "Δωσ'": [
+        {ORTH: "Δωσ'", LEMMA: "δίνω", NORM: "δίνω"},
+    ],
+
+    "Βάψ'": [
+        {ORTH: "Βάψ'", LEMMA: "βάφω", NORM: "βάφω"},
+    ],
+
+    "Αλλ'": [
+        {ORTH: "Αλλ'", LEMMA: "αλλά", NORM: "αλλά"},
+    ],
+
+    "Αμ'": [
+        {ORTH: "Αμ'", LEMMA: "άμα", NORM: "άμα"},
+    ],
+
+    "Αγόρασ'": [
+        {ORTH: "Αγόρασ'", LEMMA: "αγοράζω", NORM: "αγοράζω"},
+    ],
+
+    "'φύγε": [
+        {ORTH: "'φύγε", LEMMA: "φεύγω", NORM: "φεύγω"},
+    ],
+
+    "'φερε": [
+        {ORTH: "'φερε", LEMMA: "φέρνω", NORM: "φέρνω"},
+    ],
+
+    "'φαγε": [
+        {ORTH: "'φαγε", LEMMA: "τρώω", NORM: "τρώω"},
+    ],
+
+    "'σπαγαν": [
+        {ORTH: "'σπαγαν", LEMMA: "σπάω", NORM: "σπάω"},
+    ],
+
+    "'σκασε": [
+        {ORTH: "'σκασε", LEMMA: "σκάω", NORM: "σκάω"},
+    ],
+
+    "'σβηνε": [
+        {ORTH: "'σβηνε", LEMMA: "σβήνω", NORM: "σβήνω"},
+    ],
+
+    "'ριξε": [
+        {ORTH: "'ριξε", LEMMA: "ρίχνω", NORM: "ρίχνω"},
+    ],
+
+    "'κλεβε": [
+        {ORTH: "'κλεβε", LEMMA: "κλέβω", NORM: "κλέβω"},
+    ],
+
+    "'κει": [
+        {ORTH: "'κει", LEMMA: "εκεί", NORM: "εκεί"},
+    ],
+
+    "'βλεπε": [
+        {ORTH: "'βλεπε", LEMMA: "βλέπω", NORM: "βλέπω"},
+    ],
+
+    "'βγαινε": [
+        {ORTH: "'βγαινε", LEMMA: "βγαίνω", NORM: "βγαίνω"},
+    ]
+}
+
+_exc.update(_other_exc)
+
+for h in range(1, 12 + 1):
+
+    for period in ["π.μ.", "πμ"]:
+        _exc["%d%s" % (h, period)] = [
+            {ORTH: "%d" % h},
+            {ORTH: period, LEMMA: "π.μ.", NORM: "π.μ."}]
+
+    for period in ["μ.μ.", "μμ"]:
+        _exc["%d%s" % (h, period)] = [
+            {ORTH: "%d" % h},
+            {ORTH: period, LEMMA: "μ.μ.", NORM: "μ.μ."}]
+
+for exc_data in [
+    {ORTH: "ΑΓΡ.", LEMMA: "Αγροτικός", NORM: "Αγροτικός"},
+    {ORTH: "Αγ. Γρ.", LEMMA: "Αγία Γραφή", NORM: "Αγία Γραφή"},
+    {ORTH: "Αθ.", LEMMA: "Αθανάσιος", NORM: "Αθανάσιος"},
+    {ORTH: "Αλεξ.", LEMMA: "Αλέξανδρος", NORM: "Αλέξανδρος"},
+    {ORTH: "Απρ.", LEMMA: "Απρίλιος", NORM: "Απρίλιος"},
+    {ORTH: "Αύγ.", LEMMA: "Αύγουστος", NORM: "Αύγουστος"},
+    {ORTH: "Δεκ.", LEMMA: "Δεκέμβριος", NORM: "Δεκέμβριος"},
+    {ORTH: "Δημ.", LEMMA: "Δήμος", NORM: "Δήμος"},
+    {ORTH: "Ιαν.", LEMMA: "Ιανουάριος", NORM: "Ιανουάριος"},
+    {ORTH: "Ιούλ.", LEMMA: "Ιούλιος", NORM: "Ιούλιος"},
+    {ORTH: "Ιούν.", LEMMA: "Ιούνιος", NORM: "Ιούνιος"},
+    {ORTH: "Ιωαν.", LEMMA: "Ιωάννης", NORM: "Ιωάννης"},
+    {ORTH: "Μ. Ασία", LEMMA: "Μικρά Ασία", NORM: "Μικρά Ασία"},
+    {ORTH: "Μάρτ.", LEMMA: "Μάρτιος", NORM: "Μάρτιος"},
+    {ORTH: "Μάρτ'", LEMMA: "Μάρτιος", NORM: "Μάρτιος"},
+    {ORTH: "Νοέμβρ.", LEMMA: "Νοέμβριος", NORM: "Νοέμβριος"},
+    {ORTH: "Οκτ.", LEMMA: "Οκτώβριος", NORM: "Οκτώβριος"},
+    {ORTH: "Σεπτ.", LEMMA: "Σεπτέμβριος", NORM: "Σεπτέμβριος"},
+    {ORTH: "Φεβρ.", LEMMA: "Φεβρουάριος", NORM: "Φεβρουάριος"},
+]:
+    _exc[exc_data[ORTH]] = [exc_data]
+
+for orth in [
+    "$ΗΠΑ",
+    "Α'", "Α.Ε.", "Α.Ε.Β.Ε.", "Α.Ε.Ι.", "Α.Ε.Π.", "Α.Μ.Α.", "Α.Π.Θ.", "Α.Τ.", "Α.Χ.", "ΑΝ.", "Αγ.", "Αλ.", "Αν.",
+    "Αντ.", "Απ.",
+    "Β'", "Β)", "Β.Ζ.", "Β.Ι.Ο.", "Β.Κ.", "Β.Μ.Α.", "Βασ.",
+    "Γ'", "Γ)", "Γ.Γ.", "Γ.Δ.", "Γκ.",
+    "Δ.Ε.Η.", "Δ.Ε.Σ.Ε.", "Δ.Ν.", "Δ.Ο.Υ.", "Δ.Σ.", "Δ.Υ.", "ΔΙ.ΚΑ.Τ.Σ.Α.", "Δηλ.", "Διον.",
+    "Ε.Α.", "Ε.Α.Κ.", "Ε.Α.Π.", "Ε.Ε.", "Ε.Κ.", "Ε.ΚΕ.ΠΙΣ.", "Ε.Λ.Α.", "Ε.Λ.Ι.Α.", "Ε.Π.Σ.", "Ε.Π.Τ.Α.", "Ε.Σ.Ε.Ε.Κ.",
+    "Ε.Υ.Κ.", "ΕΕ.", "ΕΚ.", "ΕΛ.", "ΕΛ.ΑΣ.", "Εθν.", "Ελ.", "Εμ.", "Επ.", "Ευ.",
+    "Η'", "Η.Π.Α.",
+    "ΘΕ.", "Θεμ.", "Θεοδ.", "Θρ.",
+    "Ι.Ε.Κ.", "Ι.Κ.Α.", "Ι.Κ.Υ.", "Ι.Σ.Θ.", "Ι.Χ.", "ΙΖ'", "ΙΧ.",
+    "Κ.Α.Α.", "Κ.Α.Ε.", "Κ.Β.Σ.", "Κ.Δ.", "Κ.Ε.", "Κ.Ε.Κ.", "Κ.Ι.", "Κ.Κ.", "Κ.Ι.Θ.", "Κ.Ι.Θ.", "Κ.ΚΕΚ.", "Κ.Ο.",
+    "Κ.Π.Ρ.", "ΚΑΤ.", "ΚΚ.", "Καν.", "Καρ.", "Κατ.", "Κυρ.", "Κων.",
+    "Λ.Α.", "Λ.χ.", "Λ.Χ.", "Λεωφ.", "Λι.",
+    "Μ.Δ.Ε.", "Μ.Ε.Ο.", "Μ.Ζ.", "Μ.Μ.Ε.", "Μ.Ο.", "Μεγ.", "Μιλτ.", "Μιχ.",
+    "Ν.Δ.", "Ν.Ε.Α.", "Ν.Κ.", "Ν.Ο.", "Ν.Ο.Θ.", "Ν.Π.Δ.Δ.", "Ν.Υ.", "ΝΔ.", "Νικ.", "Ντ'", "Ντ.",
+    "Ο'", "Ο.Α.", "Ο.Α.Ε.Δ.", "Ο.Δ.", "Ο.Ε.Ε.", "Ο.Ε.Ε.Κ.", "Ο.Η.Ε.", "Ο.Κ.",
+    "Π.Δ.", "Π.Ε.Κ.Δ.Υ.", "Π.Ε.Π.", "Π.Μ.Σ.", "ΠΟΛ.", "Π.Χ.", "Παρ.", "Πλ.", "Πρ.",
+    "Σ.Δ.Ο.Ε.", "Σ.Ε.", "Σ.Ε.Κ.", "Σ.Π.Δ.Ω.Β.", "Σ.Τ.", "Σαβ.", "Στ.", "ΣτΕ.", "Στρ.",
+    "Τ.Α.", "Τ.Ε.Ε.", "Τ.Ε.Ι.", "ΤΡ.", "Τζ.", "Τηλ.",
+    "Υ.Γ.", "ΥΓ.", "ΥΠ.Ε.Π.Θ.",
+    "Φ.Α.Β.Ε.", "Φ.Κ.", "Φ.Σ.", "Φ.Χ.", "Φ.Π.Α.", "Φιλ.",
+    "Χ.Α.Α.", "ΧΡ.", "Χ.Χ.", "Χαρ.", "Χιλ.", "Χρ.",
+    "άγ.", "άρθρ.", "αι.", "αν.", "απ.", "αρ.", "αριθ.", "αριθμ.",
+    "β'", "βλ.",
+    "γ.γ.", "γεν.", "γραμμ.",
+    "δ.δ.", "δ.σ.", "δηλ.", "δισ.", "δολ.", "δρχ.",
+    "εκ.", "εκατ.", "ελ.",
+    "θιν'",
+    "κ.", "κ.ά.", "κ.α.", "κ.κ.", "κ.λπ.", "κ.ο.κ.", "κ.τ.λ.", "κλπ.", "κτλ.", "κυβ.",
+    "λ.χ.",
+    "μ.", "μ.Χ.", "μ.μ.", "μιλ.",
+    "ντ'",
+    "π.Χ.", "π.β.", "π.δ.", "π.μ.", "π.χ.",
+    "σ.", "σ.α.λ.", "σ.σ.", "σελ.", "στρ.",
+    "τ'ς", "τ.μ.", "τετ.", "τετρ.", "τηλ.", "τρισ.", "τόν.",
+    "υπ.",
+    "χ.μ.", "χγρ.", "χιλ.", "χλμ."
+]:
+    _exc[orth] = [{ORTH: orth}]
+
+TOKENIZER_EXCEPTIONS = _exc

spacy/tests/conftest.py

@@ -14,7 +14,7 @@ from .. import util
 # These languages are used for generic tokenizer tests – only add a language
 # here if it's using spaCy's tokenizer (not a different library)
 # TODO: re-implement generic tokenizer tests
-_languages = ['bn', 'da', 'de', 'en', 'es', 'fi', 'fr', 'ga', 'he', 'hu', 'id',
+_languages = ['bn', 'da', 'de', 'el', 'en', 'es', 'fi', 'fr', 'ga', 'he', 'hu', 'id',
               'it', 'nb', 'nl', 'pl', 'pt', 'ro', 'ru', 'sv', 'tr', 'ar', 'ut', 'tt',
               'xx']
 
@@ -158,6 +158,10 @@ def tr_tokenizer():
 def tt_tokenizer():
     return util.get_lang_class('tt').Defaults.create_tokenizer()
 
+@pytest.fixture
+def el_tokenizer():
+    return util.get_lang_class('el').Defaults.create_tokenizer()
+
 @pytest.fixture
 def ar_tokenizer():
     return util.get_lang_class('ar').Defaults.create_tokenizer()

spacy/tests/lang/el/test_exception.py

@@ -0,0 +1,18 @@
+# -*- coding: utf-8 -*-
+
+from __future__ import unicode_literals
+
+import pytest
+
+
+@pytest.mark.parametrize('text', ["αριθ.", "τρισ.", "δισ.", "σελ."])
+def test_tokenizer_handles_abbr(el_tokenizer, text):
+    tokens = el_tokenizer(text)
+    assert len(tokens) == 1
+
+
+def test_tokenizer_handles_exc_in_text(el_tokenizer):
+    text = "Στα 14 τρισ. δολάρια το κόστος από την άνοδο της στάθμης της θάλασσας."
+    tokens = el_tokenizer(text)
+    assert len(tokens) == 14
+    assert tokens[2].text == "τρισ."

spacy/tests/lang/el/test_text.py

@@ -0,0 +1,25 @@
+# -*- coding: utf-8 -*-
+
+from __future__ import unicode_literals
+
+import pytest
+
+
+def test_tokenizer_handles_long_text(el_tokenizer):
+    text = """Η Ελλάδα (παλαιότερα Ελλάς), επίσημα γνωστή ως Ελληνική Δημοκρατία,\
+    είναι χώρα της νοτιοανατολικής Ευρώπης στο νοτιότερο άκρο της Βαλκανικής χερσονήσου.\
+    Συνορεύει στα βορειοδυτικά με την Αλβανία, στα βόρεια με την πρώην\
+    Γιουγκοσλαβική Δημοκρατία της Μακεδονίας και τη Βουλγαρία και στα βορειοανατολικά με την Τουρκία."""
+    tokens = el_tokenizer(text)
+    assert len(tokens) == 54
+
+
+@pytest.mark.parametrize('text,length',[
+    ("Διοικητικά η Ελλάδα διαιρείται σε 13 Περιφέρειες.", 8),
+    ("Η εκπαίδευση στην Ελλάδα χωρίζεται κυρίως σε τρία επίπεδα.", 10),
+    ("Η Ελλάδα είναι μία από τις χώρες της Ευρωπαϊκής Ένωσης (ΕΕ) που διαθέτει σηµαντικό ορυκτό πλούτο.", 19),
+    ("Η ναυτιλία αποτέλεσε ένα σημαντικό στοιχείο της Ελληνικής οικονομικής δραστηριότητας από τα αρχαία χρόνια.", 15),
+    ("Η Ελλάδα είναι μέλος σε αρκετούς διεθνείς οργανισμούς.", 9)])
+def test_tokenizer_handles_cnts(el_tokenizer,text, length):
+    tokens = el_tokenizer(text)
+    assert len(tokens) == length