BatchalignHK 0.7.22.post21__tar.gz → 0.7.22.post22__tar.gz

{batchalignhk-0.7.22.post21 → batchalignhk-0.7.22.post22}/BatchalignHK.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: BatchalignHK
-Version: 0.7.22.post21
+Version: 0.7.22.post22
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalignhk-0.7.22.post21 → batchalignhk-0.7.22.post22}/BatchalignHK.egg-info/SOURCES.txt

@@ -100,6 +100,13 @@ batchalign/pipelines/asr/tencent.py
 batchalign/pipelines/asr/utils.py
 batchalign/pipelines/asr/whisper.py
 batchalign/pipelines/asr/whisperx.py
+batchalign/pipelines/asr/num2lang/__init__.py
+batchalign/pipelines/asr/num2lang/eus.py
+batchalign/pipelines/asr/num2lang/hrv.py
+batchalign/pipelines/asr/num2lang/ind.py
+batchalign/pipelines/asr/num2lang/nld.py
+batchalign/pipelines/asr/num2lang/por.py
+batchalign/pipelines/asr/num2lang/tha.py
 batchalign/pipelines/avqi/__init__.py
 batchalign/pipelines/avqi/engine.py
 batchalign/pipelines/cleanup/__init__.py

{batchalignhk-0.7.22.post21 → batchalignhk-0.7.22.post22}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: BatchalignHK
-Version: 0.7.22.post21
+Version: 0.7.22.post22
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

batchalignhk-0.7.22.post22/batchalign/pipelines/asr/num2lang/__init__.py

@@ -0,0 +1,15 @@
+from .eus import EUS
+from .hrv import HRV
+from .nld import NLD
+from .ind import IND
+from .por import POR
+from .tha import THA
+
+NUM2LANG = {
+    "eus": EUS,
+    "hrv": HRV,
+    "nld": NLD,
+    "ind": IND,
+    "por": POR,
+    "tha": THA
+}

batchalignhk-0.7.22.post22/batchalign/pipelines/asr/num2lang/eus.py

@@ -0,0 +1,31 @@
+EUS = {
+    "1":"bat",
+    "2":"bi",
+    "3":"hiru",
+    "4":"lau",
+    "5":"bost",
+    "6":"sei",
+    "7":"zazpi",
+    "8":"zortzi",
+    "9":"bederatzi",
+    "10":"hamar",
+    "11":"hamika",
+    "12":"hamabi",
+    "13":"hamahiru",
+    "14":"hamalau",
+    "15":"hamabost",
+    "16":"hamasei",
+    "17":"hamazazpi",
+    "18":"hemezortzi",
+    "19":"hemeretzi",
+    "20":"hogie",
+    "21":"hogeita-bat",
+    "22":"hogeita-bi",
+    "23":"hogeita-hiru",
+    "24":"hogeita-lau",
+    "25":"hogeita-bost",
+    "26":"hogeita-sei",
+    "27":"hogeita-zazpi",
+    "28":"hogeita-zortzi",
+    "29":"hogeita-bederatzi"
+}

batchalignhk-0.7.22.post22/batchalign/pipelines/asr/num2lang/hrv.py

@@ -0,0 +1,111 @@
+HRV = {
+    "1": "eins",
+    "2": "zwei",
+    "3": "drei",
+    "4": "vier",
+    "5": "fünf",
+    "6": "sechs",
+    "7": "sieben",
+    "8": "acht",
+    "9": "neun",
+    "10": "zehn",
+    "11": "elf",
+    "12": "zwölf",
+    "13": "dreizehn",
+    "14": "vierzehn",
+    "15": "fünfzehn",
+    "16": "sechzehn",
+    "17": "siebzehn",
+    "18": "achtzehn",
+    "19": "neunzehn",
+    "20": "zwanzig",
+    "21": "einundzwanzig",
+    "22": "zweiundzwanzig",
+    "23": "dreiundzwanzig",
+    "24": "vierundzwanzig",
+    "25": "fünfundzwanzig",
+    "26": "sechsundzwanzig",
+    "27": "siebenundzwanzig",
+    "28": "achtundzwanzig",
+    "29": "neunundzwanzig",
+    "30": "dreißig",
+    "31": "einunddreißig",
+    "32": "zweiunddreißig",
+    "33": "dreiunddreißig",
+    "34": "vierunddreißig",
+    "35": "fünfunddreißig",
+    "36": "sechsunddreißig",
+    "37": "siebenunddreißig",
+    "38": "achtunddreißig",
+    "39": "neununddreißig",
+    "40": "vierzig",
+    "41": "einundvierzig",
+    "42": "zweiundvierzig",
+    "43": "dreiundvierzig",
+    "44": "vierundvierzig",
+    "45": "fünfundvierzig",
+    "46": "sechsundvierzig",
+    "47": "siebenundvierzig",
+    "48": "achtundvierzig",
+    "49": "neunundvierzig",
+    "50": "fünfzig",
+    "51": "einundfünfzig",
+    "52": "zweiundfünfzig",
+    "53": "dreiundfünfzig",
+    "54": "vierundfünfzig",
+    "55": "fünfundfünfzig",
+    "56": "sechsundfünfzig",
+    "57": "siebenundfünfzig",
+    "58": "achtundfünfzig",
+    "59": "neunundfünfzig",
+    "60": "sechzig",
+    "61": "einundsechzig",
+    "62": "zweiundsechzig",
+    "63": "dreiundsechzig",
+    "64": "vierundsechzig",
+    "65": "fünfundsechzig",
+    "66": "sechsundsechzig",
+    "67": "siebenundsechzig",
+    "68": "achtundsechzig",
+    "69": "neunundsechzig",
+    "70": "siebzig",
+    "71": "einundsiebzig",
+    "72": "zweiundsiebzig",
+    "73": "dreiundsiebzig",
+    "74": "vierundsiebzig",
+    "75": "fünfundsiebzig",
+    "76": "sechsundsiebzig",
+    "77": "siebenundsiebzig",
+    "78": "achtundsiebzig",
+    "79": "neunundsiebzig",
+    "80": "achtzig",
+    "81": "einundachtzig",
+    "82": "zweiundachtzig",
+    "83": "dreiundachtzig",
+    "84": "vierundachtzig",
+    "85": "fünfundachtzig",
+    "86": "sechsundachtzig",
+    "87": "siebenundachtzig",
+    "88": "achtundachtzig",
+    "89": "neunundachtzig",
+    "90": "neunzig",
+    "91": "einundneunzig",
+    "92": "zweiundneunzig",
+    "93": "dreiundneunzig",
+    "94": "vierundneunzig",
+    "95": "fünfundneunzig",
+    "96": "sechsundneunzig",
+    "97": "siebenundneunzig",
+    "98": "achtundneunzig",
+    "99": "neunundneunzig",
+    "100": "hundert",
+    "200": "zwei-hundert",
+    "300": "drei-hundert",
+    "400": "vier-hundert",
+    "500": "fünf-hundert",
+    "600": "sechs-hundert",
+    "700": "sieben-hundert",
+    "800": "acht-hundert",
+    "900": "neun-hundert",
+    "1000": "tausend"
+}

batchalignhk-0.7.22.post22/batchalign/pipelines/asr/num2lang/ind.py

@@ -0,0 +1,23 @@
+IND = {
+    "1":"satu",
+    "2":"dua",
+    "3":"tiga",
+    "4":"empat",
+    "5":"lima",
+    "6":"enam",
+    "7":"tujuh",
+    "8":"delapan",
+    "9":"sempilan",
+    "10":"sepuluh",
+    "11":"sebelas",
+    "12":"dua-belas",
+    "13":"tiga-belas",
+    "14":"empat-belas",
+    "15":"lima-belas",
+    "16":"enam-belas",
+    "17":"tujuh-belas",
+    "18":"delapan-belas",
+    "19":"sembilan-belas",
+    "20":"dua-puluh",
+    "21":"dua-puluh-satu"
+}

batchalignhk-0.7.22.post22/batchalign/pipelines/asr/num2lang/nld.py

@@ -0,0 +1,31 @@
+NLD = {
+    "1":"een",
+    "2":"twee",
+    "3":"drie",
+    "4":"vier",
+    "5":"vijf",
+    "6":"zes",
+    "7":"zeven",
+    "8":"acht",
+    "9":"negen",
+    "10":"tien",
+    "11":"elf",
+    "12":"twaalf",
+    "13":"dertien",
+    "14":"viertien",
+    "15":"vijftien",
+    "16":"zestien",
+    "17":"zeventien",
+    "18":"achttien",
+    "19":"negentien",
+    "20":"twintig",
+    "21":"eenenttwintig",
+    "22":"tweeenttwintig",
+    "23":"drieenttwintig",
+    "24":"vierenttwintig",
+    "25":"vijfenttwintig",
+    "26":"zesenttwintig",
+    "27":"zevenenttwintig",
+    "28":"achtenttwintig",
+    "29":"negenenttwintig"
+}

batchalignhk-0.7.22.post22/batchalign/pipelines/asr/num2lang/por.py

@@ -0,0 +1,41 @@
+POR = {
+    "1":"um",
+    "2":"dois",
+    "3":"três",
+    "4":"quatro",
+    "5":"cinco",
+    "6":"seis",
+    "7":"sete",
+    "8":"oito",
+    "9":"nove",
+    "10":"dez",
+    "11":"onze",
+    "12":"doze",
+    "13":"treze",
+    "14":"quatorze",
+    "15":"quinze",
+    "16":"dezesseis",
+    "17":"dezessete",
+    "18":"dezesoito",
+    "19":"dezesnove",
+    "20":"vinte",
+    "21":"vinte-e-um",
+    "22":"vinte-e-dois",
+    "23":"vinte-e-três",
+    "24":"vinte-e-quatro",
+    "25":"vinte-e-cinco",
+    "26":"vinte-e-seis",
+    "27":"vinte-e-sete",
+    "28":"vinte-e-oito",
+    "29":"vinte-e-nove",
+    "30":"trinta",
+    "40":"quarenta",
+    "50":"cinqüenta",
+    "60":"sesenta",
+    "70":"setenta",
+    "80":"oitenta",
+    "90":"noventa",
+    "100":"centenas",
+    "101":"cento-e-um",
+    "121":"cento-e-vinte-e-um"
+}

batchalignhk-0.7.22.post22/batchalign/pipelines/asr/num2lang/tha.py

@@ -0,0 +1,23 @@
+THA = {
+    "1":"นึ่ง",
+    "2":"สอง",
+    "3":"สาม",
+    "4":"สี่",
+    "5":"ห้า",
+    "6":"หก",
+    "7":"จ็ด",
+    "8":"แปด",
+    "9":"เก้า",
+    "10":"สิบ",
+    "11":"สิบเอ็ด",
+    "12":"สิบสอง",
+    "13":"สิบสาม",
+    "14":"สิบสี่",
+    "15":"สิบห้า",
+    "16":"สิบหก",
+    "17":"สิบเจ็ด",
+    "18":"สิบแปด",
+    "19":"สิบเก้า",
+    "20":"ยี่สิบ",
+    "21":"สิบเอ็ด"
+}

{batchalignhk-0.7.22.post21 → batchalignhk-0.7.22.post22}/batchalign/pipelines/asr/utils.py

@@ -1,6 +1,7 @@
 import re
 from batchalign.document import *
 from batchalign.utils import *
+from batchalign.pipelines.asr.num2lang import NUM2LANG
 
 from batchalign.constants import ENDING_PUNCT
 from batchalign.pipelines.asr.num2chinese import num2chinese
@@ -238,6 +239,10 @@ def process_generation(output, lang="eng", utterance_engine=None):
                     elif lang == "yue":
                         return num2chinese(i, simp=False)
                     else:
+                        if NUM2LANG.get(lang) is not None:
+                            n2l = NUM2LANG.get(lang)
+                            for a,b in list(reversed(n2l.items())):
+                                i = i.replace(a,b).strip()
                         return i
                 except:
                     return i

batchalignhk-0.7.22.post22/batchalign/version

@@ -0,0 +1,3 @@
+0.7.22-post.22
+October 26st, 2025
+change numbers

batchalignhk-0.7.22.post21/batchalign/version

@@ -1,3 +0,0 @@
-0.7.22-post.21
-October 21st, 2025
-oai whisper for benchmakring