add num. verbalizer

text/LICENSE

@@ -0,0 +1,19 @@
+Copyright (c) 2017 Keith Ito
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in
+all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+THE SOFTWARE.

text/__init__.py

@@ -0,0 +1,74 @@
+""" from https://github.com/keithito/tacotron """
+import re
+from text import cleaners
+from text.symbols import symbols
+
+
+# Mappings from symbol to numeric ID and vice versa:
+_symbol_to_id = {s: i for i, s in enumerate(symbols)}
+_id_to_symbol = {i: s for i, s in enumerate(symbols)}
+
+# Regular expression matching text enclosed in curly braces:
+_curly_re = re.compile(r'(.*?)\{(.+?)\}(.*)')
+
+
+def text_to_sequence(text, cleaner_names):
+  '''Converts a string of text to a sequence of IDs corresponding to the symbols in the text.
+
+    The text can optionally have ARPAbet sequences enclosed in curly braces embedded
+    in it. For example, "Turn left on {HH AW1 S S T AH0 N} Street."
+
+    Args:
+      text: string to convert to a sequence
+      cleaner_names: names of the cleaner functions to run the text through
+
+    Returns:
+      List of integers corresponding to the symbols in the text
+  '''
+  sequence = []
+
+  # Check for curly braces and treat their contents as ARPAbet:
+  while len(text):
+    m = _curly_re.match(text)
+    if not m:
+      sequence += _symbols_to_sequence(_clean_text(text, cleaner_names))
+      break
+    sequence += _symbols_to_sequence(_clean_text(m.group(1), cleaner_names))
+    sequence += _arpabet_to_sequence(m.group(2))
+    text = m.group(3)
+
+  return sequence
+
+
+def sequence_to_text(sequence):
+  '''Converts a sequence of IDs back to a string'''
+  result = ''
+  for symbol_id in sequence:
+    if symbol_id in _id_to_symbol:
+      s = _id_to_symbol[symbol_id]
+      # Enclose ARPAbet back in curly braces:
+      if len(s) > 1 and s[0] == '@':
+        s = '{%s}' % s[1:]
+      result += s
+  return result.replace('}{', ' ')
+
+
+def _clean_text(text, cleaner_names):
+  for name in cleaner_names:
+    cleaner = getattr(cleaners, name)
+    if not cleaner:
+      raise Exception('Unknown cleaner: %s' % name)
+    text = cleaner(text)
+  return text
+
+
+def _symbols_to_sequence(symbols):
+  return [_symbol_to_id[s] for s in symbols if _should_keep_symbol(s)]
+
+
+def _arpabet_to_sequence(text):
+  return _symbols_to_sequence(['@' + s for s in text.split()])
+
+
+def _should_keep_symbol(s):
+  return s in _symbol_to_id and s is not '_' and s is not '~'

text/ca.sor

@@ -0,0 +1,485 @@
+^0 zero
+1$ u
+1 un
+2 dos
+3 tres
+4 quatre
+5 cinc
+6 sis
+7 set
+8 huit # [:ca-valencia:] [:ca-ES-valencia:]
+8 vuit
+9 nou
+#10-19
+10 deu
+11 onze
+12 dotze
+13 tretze
+14 catorze
+15 quinze
+16 setze
+17 dèsset # [:ca-valencia:] [:ca-ES-valencia:]
+17 desset # [:ca-balear:] [:ca-ES-balear:]
+17 disset
+18 devuit # [:ca-balear:] [:ca-ES-balear:]
+18 díhuit # [:ca-valencia:] [:ca-ES-valencia:]
+19 denou # [:ca-balear:] [:ca-ES-balear:]
+19 dènou # [:ca-valencia:] [:ca-ES-valencia:]
+1(\d) di$1
+# 20-29
+20 vint
+2(\d) vint-i-$1
+# 30, 40, 50, 60, 70, 80, 90
+30 trenta
+40 quaranta
+50 cinquanta
+60 seixanta
+70 setanta
+80 huitanta # [:ca-valencia:] [:ca-ES-valencia:]
+80 vuitanta
+90 noranta
+(\d)(\d) $(\10)-$2
+
+#100-199
+100 cent
+1(\d\d) cent $1
+#200-999
+(\d)00 $1-cents
+(\d)(\d\d) $1-cents $2
+
+#1000-1999
+1000 mil
+1(\d{3}) mil $1
+
+#2000-999999
+(\d{1,3})000 $1 mil
+(\d{1,3})(\d{3}) $1 mil $2
+
+# our limit is number <10^606
+(\d{606,}) ""
+
+# x-lions
+# 10000000=10^6 -> un milió
+1((0{6})+) un $(pre:$(count:\1))lió
+1((\d{6})+) un $(pre:$(count:\1))lió $1
+# 2000000=2·10^6 -> dos milions
+(\d{1,6})((0{6})+) $1 $(pre:$(count:\2))lions
+(\d{1,6})((\d{6})+) $1 $(pre:$(count:\2))lions $2
+
+
+# count number of 10^6, usefull for x-lions, and x-liards prefixes.
+count:.{0,5}? 0
+count:.{6}.{0,5} 1
+count:(.{12}).{0,5} 2
+count:(.{18}).{0,5} 3
+count:(.{24}).{0,5} 4
+count:(.{30}).{0,5} 5
+count:(.{36}).{0,5} 6
+count:(.{42}).{0,5} 7
+count:(.{48}).{0,5} 8
+count:(.{54}).{0,5} 9
+count:(.{60})(.{0,59}) 1|$(count:\2)
+count:(.{120})(.{0,59}) 2|$(count:\2)
+count:(.{180})(.{0,59}) 3|$(count:\2)
+count:(.{240})(.{0,59}) 4|$(count:\2)
+count:(.{300})(.{0,59}) 5|$(count:\2)
+count:(.{360})(.{0,59}) 6|$(count:\2)
+count:(.{420})(.{0,59}) 7|$(count:\2)
+count:(.{480})(.{0,59}) 8|$(count:\2)
+count:(.{540})(.{0,59}) 9|$(count:\2)
+count:(.{600})(.{0,5}) 10|$(count:\2) # our limit is 10^606-1
+
+# prefixes needed for x-lions and x-liards, up to 10^606-1
+pre:1 mi
+pre:2 bi
+pre:3 tri
+pre:4 quadri
+pre:5 quinti
+pre:6 sexti
+pre:7 septi
+pre:8 octi
+pre:9 noni
+pre:10 deci
+pre:1(\d) $(pre2:\1)|deci
+pre:(\d)0 $(pre3:\1)
+pre:(\d)(\d) $(pre2:\2)|$(pre3:\1)
+pre:100 centi
+
+pre2:1 uno
+pre2:2 duo
+pre2:3 tre
+pre2:4 quattour
+pre2:5 quin
+pre2:6 sex
+pre2:7 septen
+pre2:8 octo
+pre2:9 novem
+
+pre3:1 deci
+pre3:2 viginti
+pre3:3 triginti
+pre3:4 quadraginti
+pre3:5 quinquaginti
+pre3:6 sexaginti
+pre3:7 septuaginti
+pre3:8 octoginti
+pre3:9 nonoginti
+pre3:10 centi
+
+# negative number
+[--](\d+) menys |$1
+
+# decimals
+"([^,]*\d)[.]((\d{3})+)([,][^,.]*)?" $(\1\2\4)
+"([--]?\d+)([,]0*)?" $1
+"([--]?\d+)[,](\d*)" $(\1·\2)
+"([--]?\d+·0*)([^0]00?)0*" $1| |$2
+"([--]?\d+·0*)([^0])" $1| |$2
+"([--]?\d+·0*)([^0]\d)" $1| |$2
+"([--]?\d+·0*)([^0]\d\d)" $1| |$2
+"([--]?\d+·0*)([^0]\d\d)0*" $1| |$2
+
+"([--]?\d+·0*)(([^0]|[^0]\d*[^0]))0*" $1| $(read:\2)
+"([--]?\d+)·(\d*)(\d)" $(\1·\2)| |$3
+"([--]?\d+)·" $1| coma
+
+# used for decimal part
+#read:(\d*[^0])0*$ $(read:\1)
+read:(\d*[1-9])(00+)([1-9]\d*) $(read:\1)| |$(read:\2) |$(read:\3)
+read:(\d$) $1
+read:0(\d+) $(read:0)| |$(read:\1)
+read:([1-9]\d) $1
+read:([1-9]\d\d) $1
+read:(\d\d\d) $1
+read:(\d\d)((\d\d)+) $(read:\1)| |$(read:\2)
+read:(\d\d)((\d\d)*)(\d\d\d) $(read:\1)| |$(read:\2)| |$(read:\4)
+
+
+# convert masculine forms to feminine forms
+# it can be run after: standard number conversion; and after ordinal, partitive functions.
+## runned with feminine function.
+f:(.*iliard)(.*) \1$(f:\2) # convert only <1,000,000,000
+f:(.*ili)(.*) \1$(f:\2) # convert only <100,0000
+f:(.*d)o(s[^èé]*) $(f:\1ue\2) # 2 -> dos -> dues
+f:(.*cent)(s.*) $(f:\1e\2) # cents -> centes
+f:(((.*)[^a-zèé]|))u$ \1una # vint-i-u -> vint-i-una
+## runned after ord function.
+f:(.*[^0-9])n$ \1na # segon -> segona
+f:(.*[^0-9]r)$ \1a # tercer -> tercera
+f:(.*[^0-9]r)t$ \1ta # quart -> quarta
+f:(.*[^0-9])è$ \1ena # sisè -> sisena
+f:(.*[^0-9])é$ \1ena # sisé -> sisena
+## runned after ord2 function.
+f:(.*[0-9])[nrtè]$ \1a # 2n -> 2a
+## runnded after part function.
+f:(.*ter)ç$ \1cera # terç -> tercera
+f:(.*è[sc]i)m$ \1ma # milionèsim -> milionèsima
+f:(.*[^0-9]i)g$ \1tja # mig -> mitja
+
+
+no-centes:(.*)centes(.*) \1cents\2
+no-centes:(.*) \1
+
+# convert ordinal numbers (1st, 2nd, 3rd,... nth) to partitive (1, 1/2, 1/3, .... 1/n)
+p:(.*)primer$ \1unitat
+p:(.*)segon$ \1mig
+p:(.*)tercer$ \1terç
+p:(.*quart)$ \1
+p:(.*)des[èé]$ \1dècim
+p:((.*)cent)[èé]$ \1èsim
+p:((.*)mil)[èé]$ \1·lèsim
+p:((.*)ilion)[èé]$ \1èsim
+p:((.*)iliard)[èé]$ \1èsim
+
+
+# fallback, ignore 1-letter not-defined fuctions
+.:(.*) \1
+
+# runned after ordinal and partitive fuctions
+pl:(.*[^\d][nrtnec])$ \1s
+pl:(.*[^\d])ig$ \1igs # mig -> mitjos
+pl:(.*[^\d])ja$ \1ges
+pl:(.*[^\d])a$ \1es
+pl:(.*[^\d])[èé]$ \1ens
+# after ord2: 1r->1rs, 2n->2ns, 5è->5ns, ...
+pl:(\d+[rnrt])$ \1s # 1r -> 1rs, 2n -> 2ns, 4t -> 4ts
+pl:(\d+)[èé]$ \1ns # 5è -> 5ns
+pl:(\d+)a$ \1es # 2a -> 2es
+# after partitive
+pl:([^[0-9]*[sç])$ \1os # dos -> dosos, terç > terços
+pl:([^[0-9]*è[sc]im)$ \1s # dècim -> dècims
+#fallback
+pl:(.*) \1
+
+
+# unit/subunit singular/plural
+# million or greater part of the number name separated by "ili" pattern
+# before masculine to feminine conversion
+us(.).:([^,]*),([^,]*),([^,]*),([^,]*),([^,]*),([^,]*) $(\1:\7)| \2
+up(.).:([^,]*),([^,]*),([^,]*),([^,]*),([^,]*),([^,]*) $(\1:\7)| \3
+ud(.).:([^,]*),([^,]*),([^,]*),([^,]*),([^,]*),([^,]*) $(\1:\7)| \4
+ss.(.):([^,]*),([^,]*),([^,]*),([^,]*),([^,]*),([^,]*) $(\1:\7)| \5
+sp.(.):([^,]*),([^,]*),([^,]*),([^,]*),([^,]*),([^,]*) $(\1:\7)| \6
+
+# "mm" means masculine unit and masculine subunit
+# Usually used by Catalan users
+CHF:(.+),(.+) $(\2mm: franc suís, francs suïssos, de francs suïssos, cèntim, cèntims, \1)
+EUR:(.+),(.+) $(\2mm: euro, euros, d'euros, cèntim, cèntims, \1)
+GBP:(.+),(.+) $(\2fm: lliura esterlina, lliures esterlines, de lliures esterlines, penic, penics, \1)
+JPY:(.+),(.+) $(\2mm: ien, iens, de iens, sen, sen, \1)
+USD:(.+),(.+) $(\2mm: dòlar dels EUA, dòlars dels EUA, de dòlars dels EUA, centau, centaus, \1)
+# ACTIVE ISO 4217 CODES--A--
+AED:(.+),(.+) $(\2mm: dírham dels Emirats Àrabs Units, dírhams dels Emirats Àrabs Units, de dírhams dels Emirats Àrabs Units, fils, fulús, \1)
+AFN:(.+),(.+) $(\2mm: afgani, afganis, d'afganis, puli, puli, \1)
+ALL:(.+),(.+) $(\2mm: lek, lekë, de lekë, qindarka, qindarka, \1)
+AMD:(.+),(.+) $(\2mm: dram, drams, de drams, luma, luma, \1)
+ANG:(.+),(.+) $(\2mm: florí de les Antilles Neerlandeses, florins de les Antilles Neerlandeses, de florins de les Antilles Neerlandeses, cèntim, cèntims, \1)
+AOA:(.+),(.+) $(\2fm: kwanza, kwanzes, de kwanzes, cèntim cèntims, \1)
+ARS:(.+),(.+) $(\2mm: peso argentí, pesos argentins, de pesos argentins, centau, centaus, \1)
+AUD(.+),(.+) $(\2mm: dòlar australià, dòlars australians, de dòlars australians, centau, centaus, \1)
+AWG:(.+),(.+) $(\2mm: florí d'Aruba, florins d'Aruba, de florins d'Aruba, cèntim, cèntims, \1)
+AZN:(.+),(.+) $(\2mm: manat azerbaidjanès, manats azerbaidjanesos, de manats azerbaidjanesos, qəpik, qəpik, \1)
+# ACTIVE ISO 4217 CODES --X--
+#XAF Franc CFA emès pel BEAC (Banc dels Estats de l'Àfrica Central)
+XAG:(.+),(.+) $(\2fm: unça de plata, unces de plata, d'unces de plata, cèntim, cèntims, \1)
+XAU:(.+),(.+) $(\2fm: unça d'or, unces d'or, d'unces d'or, cèntim, cèntims, \1)
+#XBA Unitat compensatòria europea (EURCO) (unitat per al mercat d'obligacions)
+#XBB Unitat monetària europea (EMU-6) (unitat per al mercat d'obligacions)
+#XBC Unitat de compte europea 9 (EUA-9) (unitat per al mercat d'obligacions)
+#XBD Unitat de compte europea 17 (EUA-17) (unitat per al mercat d'obligacions)
+#XCD Dòlar del Carib Oriental
+#XDR Drets especials de gir (del Fons Monetari Internacional)
+#XFU Franc UIC (divisa especial)
+#XOF Franc CFA emès pel BCEAO (Banc Central dels Estats de l'Àfrica Occidental)
+XPD:(.+),(.+) $(\2fm: unça de pal·ladi, unces de pal·ladi, d'unces de pal·ladi, cèntim, cèntims, \1)
+#XPF Franc CFP (per als territoris francesos del Pacífic)
+XPT:(.+),(.+) $(\2fm: unça de platí, unces de platí, d'unces de platí, cèntim, cèntims, \1)
+#XTS Codi reservat per a proves
+#XXX Sense moneda, sense transacció monetària
+# OBSOLETE ISO 4217 CODES --Replaced by EUR--
+ADF:(.+),(.+) $(\2mm: franc andorrà, francs andorrans, de francs andorrans, cèntim, cèntims, \1)
+ADP:(.+),(.+) $(\2fm: pesseta andorrana, pessetes andorranes, de pessetes andorranes, cèntim, cèntims, \1)
+ATS:(.+),(.+) $(\2mm: xíling austríac, xílings austríacs, de xílings austríacs, groschen, groschen, \1)
+BEF:(.+),(.+) $(\2mm: franc belga, francs belgues, de francs belgues, cèntim, cèntims, \1)
+CYP:(.+),(.+) $(\2mm: lliura xipriota, lliures xipriotes, de lliures xipriotes, cèntim, cèntims, \1)
+DEM:(.+),(.+) $(\2mm: marc alemany, marcs alemanys, de marcs alemanys, penic, penics, \1)
+ESP:(.+),(.+) $(\2fm: pesseta, pessetes, de pessetes, cèntim, cèntims, \1)
+FIM:(.+),(.+) $(\2mm: marc finlandès, marcs finlandesos, de marcs finlandesos, penic, penics, \1)
+FRF:(.+),(.+) $(\2mm: franc francès, francs francesos, de francs francesos, cèntim, cèntims, \1)
+GRD:(.+),(.+) $(\2fm: dracma grega, dracmes gregues, leptó, leptà, \1)
+IEP:(.+),(.+) $(\2fm: lliura irlandesa, lliures irlandeses, de lliures irlandeses, penic, penics, \1)
+ITL:(.+),(.+) $(\2fm: lira italiana, lires italianes, de lires italianes, cèntim, cèntims, \1)
+LUF:(.+),(.+) $(\2mm: franc luxemburguès, francs luxemburguesos, de francs luxemburguesos, cèntim, cèntims, \1)
+MCF:(.+),(.+) $(\2mm: franc monegasc, francs monegascs, de francs monegascs, cèntim, cèntims, \1)
+MTL:(.+),(.+) $(\2fm: lira maltesa, lires malteses, de lires malteses, cèntim, cèntims, \1)
+NLG:(.+),(.+) $(\2mm: florí neerlandès, florins neerlandesos, de florins neerlandesos, cèntim, cèntims, \1)
+PTE:(.+),(.+) $(\2mm: escut portuguès, escuts portuguesos, de escuts portuguesos, centau, centaus, \1)
+SIT:(.+),(.+) $(\2mm: tolar eslovè, tolars eslovens, de tolars eslovens, stotin, stotinov, \1)
+SKK:(.+),(.+) $(\2fm: corona eslovaca, corones eslovaques, de corones eslovaques, halier, halierov, \1)
+SML:(.+),(.+) $(\2fm: lira de San Marino, lires de San Marino, de lires de San Marino, cèntim, cèntims, \1)
+VAL:(.+),(.+) $(\2fm: lira vaticana, lires vaticanes, de lires vaticanes, cèntim, cèntims, \1)
+XEU:(.+),(.+) $(\2mm: ecu, ecus, d'ecus, cèntim, cèntims, \1)
+
+#crypto-currencies
+XMR:(.+),(.+) $(\2mm: monero, moneros, de moneros, piconero, piconeros, \1) #TODO: 1,000,000,000,000 piconeros = 1 monero
+XBT:(.+),(.+) $(\2mm: bitcoin, bitcoins, de bitcoins, satoshi, satoshis, \1) # TODO: 10,000,000 satoshis = 1,000 millibitcoin = 1 bitcoin
+
+# unknow currency
+[A-Z]{3}:.* ""
+
+
+"([A-Z]{3}) ([-−]?1)([.,]00?)?"$(\1:|$2,us)
+"([A-Z]{3}) ([-−]?\d+0{6,})([.,]00?)?"$(\1:|$2,ud)
+"([A-Z]{3}) ([-−]?\d+)([.,]00?)?"$(\1:|$2,up)
+"(([A-Z]{3}) [-−]?\d+)[.,](01)" $1 amb$(\2:un,ss)
+"(([A-Z]{3}) [-−]?\d+)[.,](\d)" $1 amb$(\2:|$(\30),sp)
+"(([A-Z]{3}) [-−]?\d+)[.,](\d\d)" $1 amb$(\2:|$3,sp)
+
+
+# detects number followed by currency code
+"([-−]?\d+)([.,]\d+)? ([A-Z]{3})" $(\3 \1\2)
+
+
+# currency symbols
+"€[ ]?([^ ]*)" $(EUR \1)
+"£[ ]?([^ ]*)" $(GBP \1)
+"\$[ ]?([^ ]*)" $(USD \1)
+"¥[ ]?([^ ]*)" $(JPY \1)
+"₩[ ]?([^ ]*)" $(KRW \1)
+"₽[ ]?([^ ]*)" $(RUB \1)
+"ɱ[ ]?([^ ]*)" $(XMR \1)
+"₿[ ]?([^ ]*)" $(XBT \1)
+
+"([^ ]+)[ ]?€$" $(EUR \1)
+"([^ ]+)[ ]?£$" $(GBP \1)
+"([^ ]+)[ ]?\$$" $(USD \1)
+"([^ ]+)[ ]?¥$" $(JPY \1)
+"([^ ]+)[ ]?₩$" $(KRW \1)
+"([^ ]+)[ ]?₽$" $(RUB \1)
+"([^ ]+)[ ]?ɱ$" $(XMR \1)
+"([^ ]+)[ ]?₿$" $(XBT \1)
+
+== feminine ==
+
+1 una
+(.*) $(f:|$1)
+
+== masculine ==
+
+1 un
+(.*) $1
+
+== ordinal(-masculine)? ==
+
+([-−]\d+) ""
+\d+[,.] ""
+0 zeroé # [:ca-valencia:] [:ca-ES-valencia:]
+0 zeroè
+1 primer
+2 segon
+3 tercer
+4 quart
+(\d+)$ $(ordinal $2)
+"un ([^ ]*(ilió|iliard))$" $(ordinal \2)
+(.*li)ó$ \2oné # [:ca-valencia:] [:ca-ES-valencia:]
+(.*li)ó$ \2onè
+(.*(cent|mil|ion|iliard))s?$ \2é # [:ca-valencia:] [:ca-ES-valencia:]
+(.*(cent|mil|ion|iliard))s?$ \2è 
+"(.* )u$" \2uné # [:ca-valencia:] [:ca-ES-valencia:]
+"(.* )u$" \2unè
+(.*-)u$ \2uné # [:ca-valencia:] [:ca-ES-valencia:]
+(.*-)u$ \2unè
+"u" primer
+"un" primer
+"dos" segon
+"tres" terç
+"quatre" quart
+(.*)cinc$ \2cinqué # [:ca-valencia:] [:ca-ES-valencia:]
+(.*)cinc$ \2cinquè
+(.*)dènou$ \2denové # [:ca-valencia:] [:ca-ES-valencia:]
+(.*)nou$ \2nové # [:ca-valencia:] [:ca-ES-valencia:]
+(.*)nou$ \2novè
+(.*)deu$ \2desé # [:ca-valencia:] [:ca-ES-valencia:]
+(.*)deu$ \2desè
+(.*)dèsset$ \2desseté # [:ca-valencia:] [:ca-ES-valencia:]
+(.*)díhuit$ \2dihuité # [:ca-valencia:] [:ca-ES-valencia:]
+(.*)[ae]$ \2é # [:ca-valencia:] [:ca-ES-valencia:]
+(.*)[ae]$ \2è
+(.*\D)$ \2é # [:ca-valencia:] [:ca-ES-valencia:]
+(.*\D)$ \2è
+
+== ordinal-feminine ==
+([-−]\d+) ""
+\d+[,.] ""
+(\d+)$ $(no-centes:$(f:$(ordinal \1)))
+
+== ordinal-masculine-plural ==
+
+([-−]?\d+) $(ordinal-masculine-plural $(ordinal \1))
+primer primers
+segon segons
+(.*)è  \1ens
+(.*)er   \1ers
+
+== ordinal-feminine-plural ==
+
+([-−]?\d+) $(ordinal-feminine-plural $(ordinal-feminine \1))
+(.*)a  \1es
+
+== ordinal-number(-masculine)? ==
+
+#(\d+) $(o:\2)
+1$ 1r
+2$ 2n
+3$ 3r
+4$ 4t
+(\d+)$ \2é # [:ca-valencia:] [:ca-ES-valencia:]
+(\d+)$ \2è
+
+== ordinal-number-feminine ==
+(\d+)$ \1a
+
+== partitive(-masculine)? ==
+([--]?\d+) $(p:$(ordinal \2))
+
+== partitive-feminine ==
+([--]?\d+) $(no-centes:$(f:$(p:$(ordinal \1))))
+
+
+== partitive(-masculine)?-plural ==
+([--]?\d+) $(pl:$(p:$(ordinal $2)))
+
+== partitive-feminine-plural ==
+([--]?\d+) $(no-centes:$(pl:$(f:$(p:$(ordinal $1)))))
+
+== fraction(-masculine)? ==
+([--]?1)(/1)? $2
+([--]?1)/2 mig
+([--]?1)/([3-9]\d*) $(masculine \2)| $(partitive \3)
+([--]?\d+)(/1)? $2
+([--]?\d+)/([1-9]\d*) $2| $(partitive-plural \3)
+
+== fraction-feminine ==
+([--]?1)(/1)? $(f:$1)| unitat
+([--]?1)/([1-9]\d*) $(f:$1)| $(partitive-feminine \2)| part
+([--]?\d+)(/1)? $(f:$1)| unitats
+([--]?\d+)/([1-9]\d*) $(f:$1)| $(partitive-feminine-plural \2)| parts
+
+== collective ==
+2 parell, parella o duo
+3 tern, terna, tercet, trio, tríada o treset
+4 qüern, tètrada, quartet, quarteta o quàdruple
+5 quintern, quintet, cinquet o quíntuple
+6 sextet, siset o sèxtuple
+7 septet, setet o sèptuple
+8 octet o òctuple
+9 nònuple
+10 dècada o dècuple
+12 dotzena
+100 centenar
+144 grossa
+1000 miler
+10000 miríada
+
+== years ==
+2 bienni
+3 trienni
+4 quadrienni
+5 quinquenni o lustre
+6 sesenni
+7 septenni
+10 dècada o decenni
+12 duodecenni
+15 quindecenni
+20 vintenni o vicenni
+30 trentenni o tricenni
+40 quarantenni
+50 cinquantenni
+60 seixantenni
+70 setantenni
+80 huitantenni [:ca-valencia:] [:ca-ES-valencia:]
+80 vuitantenni
+90 norantenni
+100 segle o centenni
+1000 mil·lenni
+
+== multiplicative ==
+2 doble o duple
+3 triple
+4 quàdruple
+5 quíntuple
+6 sèxtuple
+7 sèptuple
+8 òctuple
+9 nònuple
+10 dècuple
+12 duodècuple
+100 cèntuple
+1/10 subdècuple
+1/2 súbduple
+
+== help ==
+
+"" $(1)|, $(2), $(3)\n$(help feminine)$(help masculine)$(help ordinal-number-masculine)$(help ordinal-number-feminine)$(help ordinal-feminine)$(help ordinal-masculine)
+(feminine|masculine|ordinal(-number)?(-feminine|-masculine)?) \1: $(\1 1), $(\1 2), $(\1 3)\n
+

text/cleaners.py

@@ -0,0 +1,150 @@
+""" from https://github.com/keithito/tacotron """
+
+'''
+Cleaners are transformations that run over the input text at both training and eval time.
+
+Cleaners can be selected by passing a comma-delimited list of cleaner names as the "cleaners"
+hyperparameter. Some cleaners are English-specific. You'll typically want to use:
+  1. "english_cleaners" for English text
+  2. "transliteration_cleaners" for non-English text that can be transliterated to ASCII using
+     the Unidecode library (https://pypi.python.org/pypi/Unidecode)
+  3. "basic_cleaners" if you do not want to transliterate (in this case, you should also update
+     the symbols in symbols.py to match your data).
+'''
+
+import re
+from unidecode import unidecode
+from text.numbers import normalize_numbers
+from text.numbers_ca import normalize_numbers_ca
+from text.symbols import symbols
+
+# Regular expression matching whitespace:
+_whitespace_re = re.compile(r'\s+')
+
+# List of (regular expression, replacement) pairs for abbreviations:
+_abbreviations_en = [(re.compile('\\b%s\\.' % x[0], re.IGNORECASE), x[1]) for x in [
+  ('mrs', 'misess'),
+  ('mr', 'mister'),
+  ('dr', 'doctor'),
+  ('st', 'saint'),
+  ('co', 'company'),
+  ('jr', 'junior'),
+  ('maj', 'major'),
+  ('gen', 'general'),
+  ('drs', 'doctors'),
+  ('rev', 'reverend'),
+  ('lt', 'lieutenant'),
+  ('hon', 'honorable'),
+  ('sgt', 'sergeant'),
+  ('capt', 'captain'),
+  ('esq', 'esquire'),
+  ('ltd', 'limited'),
+  ('col', 'colonel'),
+  ('ft', 'fort'),
+]]
+
+# List of (regular expression, replacement) pairs for catalan abbreviations:
+_abbreviations_ca = [(re.compile('\\b%s\\b' % x[0], re.IGNORECASE), x[1]) for x in [
+  ('tv3', 't v tres'),
+  ('8tv', 'vuit t v'),
+  ('pp', 'p p'),
+  ('psoe', 'p soe'),
+  ('sr.?', 'senyor'),
+  ('sra.?', 'senyora'),
+  ('srta.?', 'senyoreta')
+]]
+
+_replacements_ca = [(re.compile('%s' % x[0], re.IGNORECASE), x[1]) for x in [
+  (';', ','),
+  (':', '\.'),
+  ('\.\.\.,', ','),
+  ('\.\.\.', '…'),
+  ('ñ','ny')
+]]
+
+
+def expand_abbreviations(text, lang='ca'):
+  if lang == 'en':
+    _abbreviations = _abbreviations_en
+  elif lang == 'ca':
+    _abbreviations = _abbreviations_ca
+  else:
+    raise ValueError('no %s language for abbreviations'%lang)
+  for regex, replacement in _abbreviations:
+    text = re.sub(regex, replacement, text)
+  return text
+
+
+def convert_characters(text, lang='ca'):
+  if lang == 'ca':
+    _replacements = _replacements_ca
+  else:
+    raise ValueError('no %s language for punctuation conversion'%lang)
+  for regex, replacement in _replacements_ca:
+    text = re.sub(regex, replacement, text)
+  return text
+
+
+def expand_numbers(text, lang="ca"):
+  if lang == 'ca':
+    return normalize_numbers_ca(text)
+  else:
+    return normalize_numbers(text)
+
+
+def lowercase(text):
+  return text.lower()
+
+
+def collapse_whitespace(text):
+  return re.sub(_whitespace_re, ' ', text)
+
+
+def convert_to_ascii(text, lang="ca"):
+  if lang == 'en':
+    return unidecode(text)
+  elif lang == 'ca':
+    char_replace = []
+    for t in set(list(text)):
+      if t not in symbols:
+        char_replace.append([t, unidecode(t)])
+    for target, replace in char_replace:
+      text = text.replace(target, replace)
+    return text
+  else:
+    raise ValueError('no %s language for punctuation conversion'%lang)
+
+
+def basic_cleaners(text):
+  '''Basic pipeline that lowercases and collapses whitespace without transliteration.'''
+  text = lowercase(text)
+  text = collapse_whitespace(text)
+  return text
+
+
+def transliteration_cleaners(text):
+  '''Pipeline for non-English text that transliterates to ASCII.'''
+  text = convert_to_ascii(text)
+  text = lowercase(text)
+  text = collapse_whitespace(text)
+  return text
+
+
+def english_cleaners(text):
+  '''Pipeline for English text, including number and abbreviation expansion.'''
+  text = convert_to_ascii(text)
+  text = lowercase(text)
+  text = expand_numbers(text, lang='en')
+  text = expand_abbreviations(text, lang='en')
+  text = collapse_whitespace(text)
+  return text
+
+
+def catalan_cleaners(text):
+  text = lowercase(text)
+  text = expand_numbers(text, lang="ca")
+  text = convert_characters(text, lang="ca")
+  text = convert_to_ascii(text, lang="ca")
+  text = expand_abbreviations(text, lang="ca")
+  text = collapse_whitespace(text)
+  return text

text/cmudict.py

@@ -0,0 +1,65 @@
+""" from https://github.com/keithito/tacotron """
+
+import re
+
+
+valid_symbols = [
+  'AA', 'AA0', 'AA1', 'AA2', 'AE', 'AE0', 'AE1', 'AE2', 'AH', 'AH0', 'AH1', 'AH2',
+  'AO', 'AO0', 'AO1', 'AO2', 'AW', 'AW0', 'AW1', 'AW2', 'AY', 'AY0', 'AY1', 'AY2',
+  'B', 'CH', 'D', 'DH', 'EH', 'EH0', 'EH1', 'EH2', 'ER', 'ER0', 'ER1', 'ER2', 'EY',
+  'EY0', 'EY1', 'EY2', 'F', 'G', 'HH', 'IH', 'IH0', 'IH1', 'IH2', 'IY', 'IY0', 'IY1',
+  'IY2', 'JH', 'K', 'L', 'M', 'N', 'NG', 'OW', 'OW0', 'OW1', 'OW2', 'OY', 'OY0',
+  'OY1', 'OY2', 'P', 'R', 'S', 'SH', 'T', 'TH', 'UH', 'UH0', 'UH1', 'UH2', 'UW',
+  'UW0', 'UW1', 'UW2', 'V', 'W', 'Y', 'Z', 'ZH'
+]
+
+_valid_symbol_set = set(valid_symbols)
+
+
+class CMUDict:
+  '''Thin wrapper around CMUDict data. http://www.speech.cs.cmu.edu/cgi-bin/cmudict'''
+  def __init__(self, file_or_path, keep_ambiguous=True):
+    if isinstance(file_or_path, str):
+      with open(file_or_path, encoding='latin-1') as f:
+        entries = _parse_cmudict(f)
+    else:
+      entries = _parse_cmudict(file_or_path)
+    if not keep_ambiguous:
+      entries = {word: pron for word, pron in entries.items() if len(pron) == 1}
+    self._entries = entries
+
+
+  def __len__(self):
+    return len(self._entries)
+
+
+  def lookup(self, word):
+    '''Returns list of ARPAbet pronunciations of the given word.'''
+    return self._entries.get(word.upper())
+
+
+
+_alt_re = re.compile(r'\([0-9]+\)')
+
+
+def _parse_cmudict(file):
+  cmudict = {}
+  for line in file:
+    if len(line) and (line[0] >= 'A' and line[0] <= 'Z' or line[0] == "'"):
+      parts = line.split('  ')
+      word = re.sub(_alt_re, '', parts[0])
+      pronunciation = _get_pronunciation(parts[1])
+      if pronunciation:
+        if word in cmudict:
+          cmudict[word].append(pronunciation)
+        else:
+          cmudict[word] = [pronunciation]
+  return cmudict
+
+
+def _get_pronunciation(s):
+  parts = s.strip().split(' ')
+  for part in parts:
+    if part not in _valid_symbol_set:
+      return None
+  return ' '.join(parts)

text/numbers.py

@@ -0,0 +1,71 @@
+""" from https://github.com/keithito/tacotron """
+
+import inflect
+import re
+
+
+_inflect = inflect.engine()
+_comma_number_re = re.compile(r'([0-9][0-9\,]+[0-9])')
+_decimal_number_re = re.compile(r'([0-9]+\.[0-9]+)')
+_pounds_re = re.compile(r'£([0-9\,]*[0-9]+)')
+_dollars_re = re.compile(r'\$([0-9\.\,]*[0-9]+)')
+_ordinal_re = re.compile(r'[0-9]+(st|nd|rd|th)')
+_number_re = re.compile(r'[0-9]+')
+
+
+def _remove_commas(m):
+  return m.group(1).replace(',', '')
+
+
+def _expand_decimal_point(m):
+  return m.group(1).replace('.', ' point ')
+
+
+def _expand_dollars(m):
+  match = m.group(1)
+  parts = match.split('.')
+  if len(parts) > 2:
+    return match + ' dollars'  # Unexpected format
+  dollars = int(parts[0]) if parts[0] else 0
+  cents = int(parts[1]) if len(parts) > 1 and parts[1] else 0
+  if dollars and cents:
+    dollar_unit = 'dollar' if dollars == 1 else 'dollars'
+    cent_unit = 'cent' if cents == 1 else 'cents'
+    return '%s %s, %s %s' % (dollars, dollar_unit, cents, cent_unit)
+  elif dollars:
+    dollar_unit = 'dollar' if dollars == 1 else 'dollars'
+    return '%s %s' % (dollars, dollar_unit)
+  elif cents:
+    cent_unit = 'cent' if cents == 1 else 'cents'
+    return '%s %s' % (cents, cent_unit)
+  else:
+    return 'zero dollars'
+
+
+def _expand_ordinal(m):
+  return _inflect.number_to_words(m.group(0))
+
+
+def _expand_number(m):
+  num = int(m.group(0))
+  if num > 1000 and num < 3000:
+    if num == 2000:
+      return 'two thousand'
+    elif num > 2000 and num < 2010:
+      return 'two thousand ' + _inflect.number_to_words(num % 100)
+    elif num % 100 == 0:
+      return _inflect.number_to_words(num // 100) + ' hundred'
+    else:
+      return _inflect.number_to_words(num, andword='', zero='oh', group=2).replace(', ', ' ')
+  else:
+    return _inflect.number_to_words(num, andword='')
+
+
+def normalize_numbers(text):
+  text = re.sub(_comma_number_re, _remove_commas, text)
+  text = re.sub(_pounds_re, r'\1 pounds', text)
+  text = re.sub(_dollars_re, _expand_dollars, text)
+  text = re.sub(_decimal_number_re, _expand_decimal_point, text)
+  text = re.sub(_ordinal_re, _expand_ordinal, text)
+  text = re.sub(_number_re, _expand_number, text)
+  return text

text/numbers_ca.py

@@ -0,0 +1,54 @@
+import re
+import io
+import pathlib
+from text.soros import compile
+
+filepath = pathlib.Path(__file__).parent.absolute()
+with io.open(f"{filepath}/ca.sor", 'r', encoding="utf-8") as prg:
+    num2text = compile(prg.read(), 'ca')
+
+_separador_milers_re = re.compile(r'([0-9][0-9\.]+[0-9]{3})')
+_decimal_re = re.compile(r'([0-9]+\,[0-9]+)')
+_ordinal_ms_re = re.compile(r'([0-9]+)(r|er|n|on|t|rt|è|e|ne|nè)+(\b)')
+_ordinal_mp_re = re.compile(r'([0-9]+)(rs|ns|ts|ns)+(\b)')
+_ordinal_fs_re = re.compile(r'([0-9]+)(a|ra|na|ta)+(\b)')
+_ordinal_fp_re = re.compile(r'([0-9]+)(es)+(\b)')
+_cardinal_re = re.compile(r'[0-9]+')
+_fraccions_re = re.compile(r'(\b)([0-9]+\/[0-9]+)(\b)')
+_hores_re = re.compile(r'(\b)([0-9]{1,2}):([0-9]{2})(\b)')
+
+def _esborra_separador_milers(m):
+  return m.group(1).replace('.', '')
+
+def _num2text(m):
+  return num2text.run(m.group(0))
+
+def _ordinal_ms(m):
+  return num2text.run(f"ordinal {m.group(1)}") + m.group(3)
+
+def _ordinal_mp(m):
+  return num2text.run(f"ordinal-masculine-plural {m.group(1)}") + m.group(3)
+
+def _ordinal_fs(m):
+  return num2text.run(f"ordinal-feminine {m.group(1)}") + m.group(3)
+
+def _ordinal_fp(m):
+  return num2text.run(f"ordinal-feminine-plural {m.group(1)}") + m.group(3)
+
+def _fraccions(m):
+  return m.group(1) + num2text.run(f"fraction {m.group(2)}") + m.group(3)
+
+def _hores(m):
+  return m.group(1) + num2text.run(m.group(2)) + " i " + num2text.run(m.group(3))  + m.group(4)
+
+def normalize_numbers_ca(text):
+  text = re.sub(_separador_milers_re, _esborra_separador_milers, text)
+  text = re.sub(_decimal_re, _num2text, text)
+  text = re.sub(_ordinal_ms_re, _ordinal_ms, text)
+  text = re.sub(_ordinal_mp_re, _ordinal_mp, text)
+  text = re.sub(_ordinal_fs_re, _ordinal_fs, text)
+  text = re.sub(_ordinal_fp_re, _ordinal_fp, text)
+  text = re.sub(_fraccions_re, _fraccions, text)
+  text = re.sub(_hores_re, _hores, text)
+  text = re.sub(_cardinal_re, _num2text, text)
+  return text

text/numbers_ca_test.py

@@ -0,0 +1,106 @@
+import unittest
+
+from text.numbers_ca import normalize_numbers_ca
+
+
+class NumbersCa(unittest.TestCase):
+    def test_cardinals(self):
+        """
+        Converteix cardinals simples en una frase
+        """
+        self.assertEqual(normalize_numbers_ca("Va nèixer el 23 de desembre de 1988"), "Va nèixer el vint-i-tres de desembre de mil nou-cents vuitanta-vuit")
+        self.assertEqual(normalize_numbers_ca("tinc 3 preguntes"), "tinc tres preguntes")
+    
+    def test_separador_milers(self):
+        """
+        Ignora separadors de milers
+        """
+        self.assertEqual(normalize_numbers_ca("1.000"), "mil")
+        self.assertEqual(normalize_numbers_ca("323.400"), "tres-cents vint-i-tres mil quatre-cents")
+        self.assertEqual(normalize_numbers_ca("900.323.400"), "nou-cents milions tres-cents vint-i-tres mil quatre-cents")
+    
+    def test_decimals(self):
+        """
+        Converteix decimals
+        """
+        self.assertEqual(normalize_numbers_ca("1,33"), "u coma trenta-tres")
+        self.assertEqual(normalize_numbers_ca("75,5"), "setanta-cinc coma cinc")
+        self.assertEqual(normalize_numbers_ca("75,555"), "setanta-cinc coma cinc-cents cinquanta-cinc")
+        self.assertEqual(normalize_numbers_ca("999.999.999,99"), "nou-cents noranta-nou milions nou-cents noranta-nou mil nou-cents noranta-nou coma noranta-nou")
+        self.assertEqual(normalize_numbers_ca("1,12345678900"), "u coma dotze trenta-quatre cinquanta-sis set-cents vuitanta-nou")
+
+    def test_decimals_2(self):
+        """
+        Ignora comes que no pertànyen a un número decimal
+        """
+        self.assertEqual(normalize_numbers_ca("Va comprar pa, vi i llonganisses"), "Va comprar pa, vi i llonganisses")
+        self.assertEqual(normalize_numbers_ca("El número guanyador és 1, 23, 55, 34"), "El número guanyador és u, vint-i-tres, cinquanta-cinc, trenta-quatre")
+
+    def test_ordinals_ms(self):
+        """
+        Converteix ordinals masculins singulars
+        """
+        self.assertEqual(normalize_numbers_ca("Va arribar 4t de 5"), "Va arribar quart de cinc")
+        self.assertEqual(normalize_numbers_ca("el 1r va ser ell"), "el primer va ser ell")
+        self.assertEqual(normalize_numbers_ca("el 3er, no va aguantar"), "el tercer, no va aguantar")
+        self.assertEqual(normalize_numbers_ca("2n"), "segon")
+        self.assertEqual(normalize_numbers_ca("2on"), "segon")
+        self.assertEqual(normalize_numbers_ca("4t"), "quart")
+        self.assertEqual(normalize_numbers_ca("4rt"), "quart")
+        self.assertEqual(normalize_numbers_ca("5è: remogueu la barreja"), "cinquè: remogueu la barreja")
+        self.assertEqual(normalize_numbers_ca("6e"), "sisè")
+        self.assertEqual(normalize_numbers_ca("6e"), "sisè")
+        self.assertEqual(normalize_numbers_ca("21nè"), "vint-i-unè")
+        self.assertEqual(normalize_numbers_ca("un 81ne de Palamós"), "un vuitanta-unè de Palamós")
+
+    def test_ordinals_fs(self):
+        """
+        Converteix ordinals femenins singulars
+        """
+        self.assertEqual(normalize_numbers_ca("1a"), "primera")
+        self.assertEqual(normalize_numbers_ca("3ra"), "tercera")
+        self.assertEqual(normalize_numbers_ca("2a"), "segona")
+        self.assertEqual(normalize_numbers_ca("2na"), "segona")
+        self.assertEqual(normalize_numbers_ca("4a."), "quarta.")
+        self.assertEqual(normalize_numbers_ca("pugi a la 4ta, després giri a l'esquerra"), "pugi a la quarta, després giri a l'esquerra")
+        self.assertEqual(normalize_numbers_ca("va quedar 5a en la classificació"), "va quedar cinquena en la classificació")
+        self.assertEqual(normalize_numbers_ca("la 5na vegada"), "la cinquena vegada")
+
+    def test_ordinals_mp(self):
+        """
+        Converteix ordinals masculins plurals
+        """
+        self.assertEqual(normalize_numbers_ca("1rs"), "primers")
+        self.assertEqual(normalize_numbers_ca("van arribar 2ns"), "van arribar segons")
+    
+    def test_ordinals_fp(self):
+        """
+        Converteix ordinals femenins plurals
+        """
+        self.assertEqual(normalize_numbers_ca("1es"), "primeres")
+    
+    def test_fraccions_s(self):
+        """
+        Converteix fraccions singulars
+        """
+        self.assertEqual(normalize_numbers_ca("1/2 got de vi"), "mig got de vi")
+        self.assertEqual(normalize_numbers_ca("1/3 de farina"), "un terç de farina")
+        self.assertEqual(normalize_numbers_ca("1/8"), "un vuitè")
+        
+    def test_fraccions_p(self):
+        """
+        Converteix fraccions plurals
+        """
+        self.assertEqual(normalize_numbers_ca("4/2 gots de vi"), "quatre migs gots de vi")
+        self.assertEqual(normalize_numbers_ca("2/3 de farina"), "dos terços de farina")
+        self.assertEqual(normalize_numbers_ca("3/8"), "tres vuitens")
+
+    def test_hores(self):
+        """
+        Converteix hores de manera simplificada
+        """
+        self.assertEqual(normalize_numbers_ca("a les 11:45"), "a les onze i quaranta-cinc")
+        self.assertEqual(normalize_numbers_ca("a partir de les 23:12"), "a partir de les vint-i-tres i dotze")
+        
+if __name__ == '__main__':
+    unittest.main()

text/soros.py

@@ -0,0 +1,140 @@
+"Soros interpreter (see http://numbertext.org)"
+from __future__ import unicode_literals
+from __future__ import print_function
+import re
+import sys
+
+
+def run(program, data, lang):
+    return compile(program, lang).run(data)
+
+
+def compile(program, lang):
+    return _Soros(program, lang)
+
+# conversion function
+
+
+def _tr(text, chars, chars2, delim):
+    for i in range(0, len(chars)):
+        text = text.replace(delim + chars[i], chars2[i])
+    return text
+
+
+# string literals for metacharacter encoding
+_m = "\\\";#$()|[]"
+# Unicode private area
+_c = u"\uE000\uE001\uE002\uE003\uE004\uE005\uE006\uE007\uE008\uE009"
+_pipe = u"\uE003"
+# separator prefix = \uE00A
+
+# pattern to recognize function calls in the replacement string
+_func = re.compile(_tr(r"""(?:\|?(?:\$\()+)?  # optional nested calls
+                (\|?\$\(([^\(\)]*)\)\|?)      # inner call (2 subgroups)
+                (?:\)+\|?)?""",               # optional nested calls
+                       _m[4:8], _c[:4], "\\"), re.X)  # \$, \(, \), \| -> \uE000..\uE003
+
+
+class _Soros:
+    def __init__(self, prg, lang):
+        self.lines = []
+        if prg.find("__numbertext__") == -1:
+            prg = "__numbertext__;" + prg
+        # default left zero deletion
+        # and separator function (no separation, if subcall returns with empty string)
+        prg = prg.replace("__numbertext__", u"""0+(0|[1-9]\\d*) $1
+\"([a-z][-a-z]* )0+(0|[1-9]\\d*)\" $(\\1\\2)
+\"\uE00A(.*)\uE00A(.+)\uE00A(.*)\" \\1\\2\\3
+\"\uE00A.*\uE00A\uE00A.*\"
+""")
+        prg = _tr(prg, _m[:4], _c[:4],
+                  "\\")  # \\, \", \;, \# -> \uE000..\uE003
+        # switch off all country-dependent lines, and switch on the requested ones
+        prg = re.sub(
+            r"(^|[\n;])([^\n;#]*#[^\n]*[\[]:[^\n:\]]*:][^\n]*)", r"\1#\2", prg)
+        prg = re.sub(r"(^|[\n;])#([^\n;#]*#[^\n]*[\[]:" +
+                     lang.replace("_", "-") + r":][^\n]*)", r"\1\2", prg)
+        matchline = re.compile("^\s*(\"[^\"]*\"|[^\s]*)\s*(.*[^\s])?\s*$")
+        prefix = ""
+        for s in re.sub("(#[^\n]*)?(\n|$)", ";", prg).split(";"):
+            macro = re.match("== *(.*[^ ]?) ==", s)
+            if macro != None:
+                prefix = macro.group(1)
+                continue
+            m = matchline.match(s)
+            if prefix != "" and s != "" and m != None:
+                s = m.group(1).strip("\"")
+                space = " " if s != "" else ""
+                caret = ""
+                if s[0:1] == "^":
+                    s = s[1:]
+                    caret = "^"
+                s2 = m.group(2) if m.group(2) != None else ""
+                s = "\"" + caret + prefix + space + s + "\" " + s2
+                m = matchline.match(s)
+            if m != None:
+                s = _tr(m.group(1).strip("\""), _c[1:4], _m[1:4], "") \
+                    .replace(_c[_m.find("\\")], "\\\\")  # -> \\, ", ;, #
+                if m.group(2) != None:
+                    s2 = m.group(2).strip("\"")
+                else:
+                    s2 = ""
+                # \$, \(, \), \|, \[, \] -> \uE004..\uE009
+                s2 = _tr(s2, _m[4:], _c[4:], "\\")
+                # call inner separator: [ ... $1 ... ] -> $(\uE00A ... \uE00A$1\uE00A ... )
+                s2 = re.sub(r"[\[]\$(\d\d?|\([^\)]+\))",
+                            u"$(\uE00A\uE00A|$\\1\uE00A", s2)
+                s2 = re.sub(r"[\[]([^\$[\\]*)\$(\d\d?|\([^\)]+\))",
+                            u"$(\uE00A\\1\uE00A$\\2\uE00A", s2)
+                # add "|" in terminating position
+                s2 = re.sub(r"\uE00A]$", "|\uE00A)", s2)
+                s2 = re.sub(r"]", ")", s2)
+                s2 = re.sub(r"(\$\d|\))\|\$", r"\1||$",
+                            s2)  # $()|$() -> $()||$()
+                # \uE000..\uE003-> \, ", ;, #
+                s2 = _tr(s2, _c[:4], _m[:4], "")
+                # $, (, ), | -> \uE000..\uE003
+                s2 = _tr(s2, _m[4:8], _c[:4], "")
+                # \uE004..\uE009 -> $, (, ), |, [, ]
+                s2 = _tr(s2, _c[4:], _m[4:], "")
+                s2 = re.sub(r"\\(\d)", r"\\g<\1>",
+                            re.sub(r"\uE000(\d)", "\uE000\uE001\\\\g<\\1>\uE002", s2))
+                try:
+                    self.lines = self.lines + [[
+                        re.compile("^" + s.lstrip("^").rstrip("$") + "$"),
+                        s2, s[:1] == "^", s[-1:] == "$"]]
+                except:
+                    print("Error in following regex line: " + s, file=sys.stderr)
+                    raise
+
+    def run(self, data):
+        return self._run(data, True, True)
+
+    def _run(self, data, begin, end):
+        for i in self.lines:
+            if not ((begin == False and i[2]) or (end == False and i[3])):
+                m = i[0].match(data)
+                if m:
+                    try:
+                        s = m.expand(i[1])
+                    except:
+                        print("Error for the following input: " +
+                              data, file=sys.stderr)
+                        raise
+                    n = _func.search(s)
+                    while n:
+                        b = False
+                        e = False
+                        if n.group(1)[0:1] == _pipe or n.group()[0:1] == _pipe:
+                            b = True
+                        elif n.start() == 0:
+                            b = begin
+                        if n.group(1)[-1:] == _pipe or n.group()[-1:] == _pipe:
+                            e = True
+                        elif n.end() == len(s):
+                            e = end
+                        s = s[:n.start(1)] + self._run(n.group(2),
+                                                       b, e) + s[n.end(1):]
+                        n = _func.search(s)
+                    return s
+        return ""

text/symbols.py

@@ -0,0 +1,17 @@
+""" from https://github.com/keithito/tacotron """
+
+'''
+Defines the set of symbols used in text input to the model.
+
+The default is a set of ASCII characters that works well for English or text that has been run through Unidecode. For other data, you can modify _characters. See TRAINING_DATA.md for details. '''
+from text import cmudict
+
+_pad = '_' # in principle not used in tacotron2
+_punctuation = '\'!,.?…· '
+_letters = 'AÀÁBCÇDEÉÈFGHIÍÏJKLMNOÓÒPQRSTUÜÚVWXYZaàábcçdeéèfghiíïjklmnoóòpqrstuüúvwxyz'
+
+# Prepend "@" to ARPAbet symbols to ensure uniqueness (some are the same as uppercase letters):
+_arpabet = ['@' + s for s in cmudict.valid_symbols]
+
+# Export all symbols:
+symbols = [_pad] + list(_punctuation) + list(_letters) + _arpabet

text/symbols_en.py

@@ -0,0 +1,18 @@
+""" from https://github.com/keithito/tacotron """
+
+'''
+Defines the set of symbols used in text input to the model.
+
+The default is a set of ASCII characters that works well for English or text that has been run through Unidecode. For other data, you can modify _characters. See TRAINING_DATA.md for details. '''
+from text import cmudict
+
+_pad        = '_'
+_punctuation = '!\'(),.:;? '
+_special = '-'
+_letters = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz'
+
+# Prepend "@" to ARPAbet symbols to ensure uniqueness (some are the same as uppercase letters):
+_arpabet = ['@' + s for s in cmudict.valid_symbols]
+
+# Export all symbols:
+symbols = [_pad] + list(_special) + list(_punctuation) + list(_letters) + _arpabet