bg-text-normalizer 1.0.0__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- bg_text_normalizer-1.0.0/LICENSE +21 -0
- bg_text_normalizer-1.0.0/PKG-INFO +147 -0
- bg_text_normalizer-1.0.0/README.md +118 -0
- bg_text_normalizer-1.0.0/pyproject.toml +42 -0
- bg_text_normalizer-1.0.0/setup.cfg +4 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer/__init__.py +40 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer/bg_abbreviations.py +238 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer/bg_currency.py +124 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer/bg_dates.py +97 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer/bg_normalizer.py +342 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer/bg_numbers.py +407 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer/bg_phone.py +82 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer/bg_roman.py +58 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer/bg_time.py +62 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer.egg-info/PKG-INFO +147 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer.egg-info/SOURCES.txt +16 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer.egg-info/dependency_links.txt +1 -0
- bg_text_normalizer-1.0.0/src/bg_text_normalizer.egg-info/top_level.txt +1 -0
|
@@ -0,0 +1,21 @@
|
|
|
1
|
+
MIT License
|
|
2
|
+
|
|
3
|
+
Copyright (c) 2025 Radi Totev
|
|
4
|
+
|
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
6
|
+
of this software and associated documentation files (the "Software"), to deal
|
|
7
|
+
in the Software without restriction, including without limitation the rights
|
|
8
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
9
|
+
copies of the Software, and to permit persons to whom the Software is
|
|
10
|
+
furnished to do so, subject to the following conditions:
|
|
11
|
+
|
|
12
|
+
The above copyright notice and this permission notice shall be included in all
|
|
13
|
+
copies or substantial portions of the Software.
|
|
14
|
+
|
|
15
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
16
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
17
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
18
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
19
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
20
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|
21
|
+
SOFTWARE.
|
|
@@ -0,0 +1,147 @@
|
|
|
1
|
+
Metadata-Version: 2.4
|
|
2
|
+
Name: bg-text-normalizer
|
|
3
|
+
Version: 1.0.0
|
|
4
|
+
Summary: Bulgarian text normalization for TTS — converts numbers, dates, currency, abbreviations to spoken form
|
|
5
|
+
Author: Radi Totev
|
|
6
|
+
License-Expression: MIT
|
|
7
|
+
Project-URL: Homepage, https://github.com/raditotev/bg-text-normalizer
|
|
8
|
+
Project-URL: Repository, https://github.com/raditotev/bg-text-normalizer
|
|
9
|
+
Project-URL: Issues, https://github.com/raditotev/bg-text-normalizer/issues
|
|
10
|
+
Keywords: bulgarian,tts,text-normalization,nlp,speech-synthesis,bulgarian-language
|
|
11
|
+
Classifier: Development Status :: 4 - Beta
|
|
12
|
+
Classifier: Intended Audience :: Developers
|
|
13
|
+
Classifier: Intended Audience :: Science/Research
|
|
14
|
+
Classifier: Programming Language :: Python :: 3
|
|
15
|
+
Classifier: Programming Language :: Python :: 3.8
|
|
16
|
+
Classifier: Programming Language :: Python :: 3.9
|
|
17
|
+
Classifier: Programming Language :: Python :: 3.10
|
|
18
|
+
Classifier: Programming Language :: Python :: 3.11
|
|
19
|
+
Classifier: Programming Language :: Python :: 3.12
|
|
20
|
+
Classifier: Programming Language :: Python :: 3.13
|
|
21
|
+
Classifier: Topic :: Multimedia :: Sound/Audio :: Speech
|
|
22
|
+
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
|
|
23
|
+
Classifier: Topic :: Text Processing :: Linguistic
|
|
24
|
+
Classifier: Natural Language :: Bulgarian
|
|
25
|
+
Requires-Python: >=3.8
|
|
26
|
+
Description-Content-Type: text/markdown
|
|
27
|
+
License-File: LICENSE
|
|
28
|
+
Dynamic: license-file
|
|
29
|
+
|
|
30
|
+
# Bulgarian Text Normalizer for TTS
|
|
31
|
+
|
|
32
|
+
A comprehensive text normalization package that converts written Bulgarian text into its spoken form, designed as a preprocessing step for Text-to-Speech (TTS) systems.
|
|
33
|
+
|
|
34
|
+
## Features
|
|
35
|
+
|
|
36
|
+
| Category | Examples |
|
|
37
|
+
|----------|----------|
|
|
38
|
+
| **Numbers** | `1500` → `хиляда и петстотин` |
|
|
39
|
+
| **Dates** | `15.02.2026 г.` → `петнадесети февруари две хиляди двадесет и шеста година` |
|
|
40
|
+
| **Time** | `14:30 ч.` → `четиринадесет и тридесет часа` |
|
|
41
|
+
| **Currency** | `99.99 лв.` → `деветдесет и девет лева и деветдесет и девет стотинки` |
|
|
42
|
+
| **Percentages** | `15.5%` → `петнадесет цяло и пет десети процента` |
|
|
43
|
+
| **Ordinals** | `21-ви` → `двадесет и първи` |
|
|
44
|
+
| **Abbreviations** | `бул. Витоша, гр. София` → `булевард Витоша, град София` |
|
|
45
|
+
| **Phone numbers** | `+359 888 123 456` → digit-by-digit reading |
|
|
46
|
+
| **Roman numerals** | `век XXI` → `век двадесет и първи` |
|
|
47
|
+
| **Symbols** | `№10` → `номер десет` |
|
|
48
|
+
|
|
49
|
+
## Grammatical Correctness
|
|
50
|
+
|
|
51
|
+
- **Gender agreement**: Handles masculine/feminine/neuter (`един`/`една`/`едно`, `два`/`две`)
|
|
52
|
+
- **Ordinal forms**: Full gender-aware ordinals (`първи`/`първа`/`първо`)
|
|
53
|
+
- **Year reading**: Ordinal feminine form matching "година" (`две хиляди двадесет и шеста`)
|
|
54
|
+
- **Space-separated thousands**: `7 000 000` → `седем милиона`
|
|
55
|
+
|
|
56
|
+
## Usage
|
|
57
|
+
|
|
58
|
+
### Quick usage
|
|
59
|
+
```python
|
|
60
|
+
from bg_text_normalizer import normalize_text
|
|
61
|
+
|
|
62
|
+
result = normalize_text("На 15.02.2026 г. в 14:30 ч. цената е 1500.50 лв.")
|
|
63
|
+
# "На петнадесети февруари две хиляди двадесет и шеста година в четиринадесет
|
|
64
|
+
# и тридесет часа цената е хиляда и петстотин лева и петдесет стотинки."
|
|
65
|
+
```
|
|
66
|
+
|
|
67
|
+
### Class-based usage
|
|
68
|
+
```python
|
|
69
|
+
from bg_text_normalizer import BulgarianTextNormalizer
|
|
70
|
+
|
|
71
|
+
normalizer = BulgarianTextNormalizer(expand_abbrevs=True, verbose=False)
|
|
72
|
+
result = normalizer.normalize("бул. Витоша №10, гр. София")
|
|
73
|
+
# "булевард Витоша номер десет, град София"
|
|
74
|
+
```
|
|
75
|
+
|
|
76
|
+
### Individual modules
|
|
77
|
+
```python
|
|
78
|
+
from bg_text_normalizer.bg_numbers import number_to_words_cardinal, number_to_words_ordinal
|
|
79
|
+
from bg_text_normalizer.bg_dates import normalize_date
|
|
80
|
+
from bg_text_normalizer.bg_currency import normalize_currency
|
|
81
|
+
|
|
82
|
+
number_to_words_cardinal(2500, gender='m') # "две хиляди и петстотин"
|
|
83
|
+
number_to_words_ordinal(15, gender='m') # "петнадесети"
|
|
84
|
+
normalize_date(15, 2, 2026) # "петнадесети февруари две хиляди двадесет и шеста"
|
|
85
|
+
normalize_currency("99.99", "BGN") # "деветдесет и девет лева и деветдесет и девет стотинки"
|
|
86
|
+
```
|
|
87
|
+
|
|
88
|
+
## Integration with TTS Training (Qwen3-TTS)
|
|
89
|
+
|
|
90
|
+
Use this normalizer as a preprocessing step when preparing your training data:
|
|
91
|
+
|
|
92
|
+
```python
|
|
93
|
+
import json
|
|
94
|
+
from bg_text_normalizer import normalize_text
|
|
95
|
+
|
|
96
|
+
# Process your JSONL training data
|
|
97
|
+
with open('raw_data.jsonl', 'r') as f_in, open('normalized_data.jsonl', 'w') as f_out:
|
|
98
|
+
for line in f_in:
|
|
99
|
+
entry = json.loads(line)
|
|
100
|
+
entry['text'] = normalize_text(entry['text'])
|
|
101
|
+
f_out.write(json.dumps(entry, ensure_ascii=False) + '\n')
|
|
102
|
+
```
|
|
103
|
+
|
|
104
|
+
For inference (runtime TTS), add normalization before synthesis:
|
|
105
|
+
|
|
106
|
+
```python
|
|
107
|
+
from bg_text_normalizer import normalize_text
|
|
108
|
+
|
|
109
|
+
def synthesize(text: str):
|
|
110
|
+
normalized = normalize_text(text)
|
|
111
|
+
# ... pass normalized text to TTS model
|
|
112
|
+
```
|
|
113
|
+
|
|
114
|
+
## File Structure
|
|
115
|
+
|
|
116
|
+
```
|
|
117
|
+
bg-text-normalizer/
|
|
118
|
+
├── src/
|
|
119
|
+
│ └── bg_text_normalizer/
|
|
120
|
+
│ ├── __init__.py # Package entry point
|
|
121
|
+
│ ├── bg_normalizer.py # Main orchestrator
|
|
122
|
+
│ ├── bg_numbers.py # Cardinal, ordinal, decimal numbers
|
|
123
|
+
│ ├── bg_dates.py # Date normalization
|
|
124
|
+
│ ├── bg_time.py # Time normalization
|
|
125
|
+
│ ├── bg_currency.py # Currency (BGN, EUR, USD, GBP)
|
|
126
|
+
│ ├── bg_abbreviations.py # 100+ Bulgarian abbreviations
|
|
127
|
+
│ ├── bg_phone.py # Phone number reading
|
|
128
|
+
│ └── bg_roman.py # Roman numeral conversion
|
|
129
|
+
├── test_normalizer.py # Test suite
|
|
130
|
+
├── pyproject.toml
|
|
131
|
+
└── README.md
|
|
132
|
+
```
|
|
133
|
+
|
|
134
|
+
## Adding Custom Abbreviations
|
|
135
|
+
|
|
136
|
+
Edit `src/bg_text_normalizer/bg_abbreviations.py` and add entries to the appropriate dictionary:
|
|
137
|
+
|
|
138
|
+
```python
|
|
139
|
+
# In ADDRESS_ABBREVS, TITLE_ABBREVS, etc.
|
|
140
|
+
CUSTOM_ABBREVS = {
|
|
141
|
+
'your_abbrev.': 'пълна форма',
|
|
142
|
+
}
|
|
143
|
+
```
|
|
144
|
+
|
|
145
|
+
## Dependencies
|
|
146
|
+
|
|
147
|
+
**None** — pure Python, no external dependencies required.
|
|
@@ -0,0 +1,118 @@
|
|
|
1
|
+
# Bulgarian Text Normalizer for TTS
|
|
2
|
+
|
|
3
|
+
A comprehensive text normalization package that converts written Bulgarian text into its spoken form, designed as a preprocessing step for Text-to-Speech (TTS) systems.
|
|
4
|
+
|
|
5
|
+
## Features
|
|
6
|
+
|
|
7
|
+
| Category | Examples |
|
|
8
|
+
|----------|----------|
|
|
9
|
+
| **Numbers** | `1500` → `хиляда и петстотин` |
|
|
10
|
+
| **Dates** | `15.02.2026 г.` → `петнадесети февруари две хиляди двадесет и шеста година` |
|
|
11
|
+
| **Time** | `14:30 ч.` → `четиринадесет и тридесет часа` |
|
|
12
|
+
| **Currency** | `99.99 лв.` → `деветдесет и девет лева и деветдесет и девет стотинки` |
|
|
13
|
+
| **Percentages** | `15.5%` → `петнадесет цяло и пет десети процента` |
|
|
14
|
+
| **Ordinals** | `21-ви` → `двадесет и първи` |
|
|
15
|
+
| **Abbreviations** | `бул. Витоша, гр. София` → `булевард Витоша, град София` |
|
|
16
|
+
| **Phone numbers** | `+359 888 123 456` → digit-by-digit reading |
|
|
17
|
+
| **Roman numerals** | `век XXI` → `век двадесет и първи` |
|
|
18
|
+
| **Symbols** | `№10` → `номер десет` |
|
|
19
|
+
|
|
20
|
+
## Grammatical Correctness
|
|
21
|
+
|
|
22
|
+
- **Gender agreement**: Handles masculine/feminine/neuter (`един`/`една`/`едно`, `два`/`две`)
|
|
23
|
+
- **Ordinal forms**: Full gender-aware ordinals (`първи`/`първа`/`първо`)
|
|
24
|
+
- **Year reading**: Ordinal feminine form matching "година" (`две хиляди двадесет и шеста`)
|
|
25
|
+
- **Space-separated thousands**: `7 000 000` → `седем милиона`
|
|
26
|
+
|
|
27
|
+
## Usage
|
|
28
|
+
|
|
29
|
+
### Quick usage
|
|
30
|
+
```python
|
|
31
|
+
from bg_text_normalizer import normalize_text
|
|
32
|
+
|
|
33
|
+
result = normalize_text("На 15.02.2026 г. в 14:30 ч. цената е 1500.50 лв.")
|
|
34
|
+
# "На петнадесети февруари две хиляди двадесет и шеста година в четиринадесет
|
|
35
|
+
# и тридесет часа цената е хиляда и петстотин лева и петдесет стотинки."
|
|
36
|
+
```
|
|
37
|
+
|
|
38
|
+
### Class-based usage
|
|
39
|
+
```python
|
|
40
|
+
from bg_text_normalizer import BulgarianTextNormalizer
|
|
41
|
+
|
|
42
|
+
normalizer = BulgarianTextNormalizer(expand_abbrevs=True, verbose=False)
|
|
43
|
+
result = normalizer.normalize("бул. Витоша №10, гр. София")
|
|
44
|
+
# "булевард Витоша номер десет, град София"
|
|
45
|
+
```
|
|
46
|
+
|
|
47
|
+
### Individual modules
|
|
48
|
+
```python
|
|
49
|
+
from bg_text_normalizer.bg_numbers import number_to_words_cardinal, number_to_words_ordinal
|
|
50
|
+
from bg_text_normalizer.bg_dates import normalize_date
|
|
51
|
+
from bg_text_normalizer.bg_currency import normalize_currency
|
|
52
|
+
|
|
53
|
+
number_to_words_cardinal(2500, gender='m') # "две хиляди и петстотин"
|
|
54
|
+
number_to_words_ordinal(15, gender='m') # "петнадесети"
|
|
55
|
+
normalize_date(15, 2, 2026) # "петнадесети февруари две хиляди двадесет и шеста"
|
|
56
|
+
normalize_currency("99.99", "BGN") # "деветдесет и девет лева и деветдесет и девет стотинки"
|
|
57
|
+
```
|
|
58
|
+
|
|
59
|
+
## Integration with TTS Training (Qwen3-TTS)
|
|
60
|
+
|
|
61
|
+
Use this normalizer as a preprocessing step when preparing your training data:
|
|
62
|
+
|
|
63
|
+
```python
|
|
64
|
+
import json
|
|
65
|
+
from bg_text_normalizer import normalize_text
|
|
66
|
+
|
|
67
|
+
# Process your JSONL training data
|
|
68
|
+
with open('raw_data.jsonl', 'r') as f_in, open('normalized_data.jsonl', 'w') as f_out:
|
|
69
|
+
for line in f_in:
|
|
70
|
+
entry = json.loads(line)
|
|
71
|
+
entry['text'] = normalize_text(entry['text'])
|
|
72
|
+
f_out.write(json.dumps(entry, ensure_ascii=False) + '\n')
|
|
73
|
+
```
|
|
74
|
+
|
|
75
|
+
For inference (runtime TTS), add normalization before synthesis:
|
|
76
|
+
|
|
77
|
+
```python
|
|
78
|
+
from bg_text_normalizer import normalize_text
|
|
79
|
+
|
|
80
|
+
def synthesize(text: str):
|
|
81
|
+
normalized = normalize_text(text)
|
|
82
|
+
# ... pass normalized text to TTS model
|
|
83
|
+
```
|
|
84
|
+
|
|
85
|
+
## File Structure
|
|
86
|
+
|
|
87
|
+
```
|
|
88
|
+
bg-text-normalizer/
|
|
89
|
+
├── src/
|
|
90
|
+
│ └── bg_text_normalizer/
|
|
91
|
+
│ ├── __init__.py # Package entry point
|
|
92
|
+
│ ├── bg_normalizer.py # Main orchestrator
|
|
93
|
+
│ ├── bg_numbers.py # Cardinal, ordinal, decimal numbers
|
|
94
|
+
│ ├── bg_dates.py # Date normalization
|
|
95
|
+
│ ├── bg_time.py # Time normalization
|
|
96
|
+
│ ├── bg_currency.py # Currency (BGN, EUR, USD, GBP)
|
|
97
|
+
│ ├── bg_abbreviations.py # 100+ Bulgarian abbreviations
|
|
98
|
+
│ ├── bg_phone.py # Phone number reading
|
|
99
|
+
│ └── bg_roman.py # Roman numeral conversion
|
|
100
|
+
├── test_normalizer.py # Test suite
|
|
101
|
+
├── pyproject.toml
|
|
102
|
+
└── README.md
|
|
103
|
+
```
|
|
104
|
+
|
|
105
|
+
## Adding Custom Abbreviations
|
|
106
|
+
|
|
107
|
+
Edit `src/bg_text_normalizer/bg_abbreviations.py` and add entries to the appropriate dictionary:
|
|
108
|
+
|
|
109
|
+
```python
|
|
110
|
+
# In ADDRESS_ABBREVS, TITLE_ABBREVS, etc.
|
|
111
|
+
CUSTOM_ABBREVS = {
|
|
112
|
+
'your_abbrev.': 'пълна форма',
|
|
113
|
+
}
|
|
114
|
+
```
|
|
115
|
+
|
|
116
|
+
## Dependencies
|
|
117
|
+
|
|
118
|
+
**None** — pure Python, no external dependencies required.
|
|
@@ -0,0 +1,42 @@
|
|
|
1
|
+
[build-system]
|
|
2
|
+
requires = ["setuptools>=68.0", "wheel"]
|
|
3
|
+
build-backend = "setuptools.build_meta"
|
|
4
|
+
|
|
5
|
+
[project]
|
|
6
|
+
name = "bg-text-normalizer"
|
|
7
|
+
version = "1.0.0"
|
|
8
|
+
description = "Bulgarian text normalization for TTS — converts numbers, dates, currency, abbreviations to spoken form"
|
|
9
|
+
readme = "README.md"
|
|
10
|
+
license = "MIT"
|
|
11
|
+
requires-python = ">=3.8"
|
|
12
|
+
authors = [
|
|
13
|
+
{name = "Radi Totev"},
|
|
14
|
+
]
|
|
15
|
+
keywords = ["bulgarian", "tts", "text-normalization", "nlp", "speech-synthesis", "bulgarian-language"]
|
|
16
|
+
classifiers = [
|
|
17
|
+
"Development Status :: 4 - Beta",
|
|
18
|
+
"Intended Audience :: Developers",
|
|
19
|
+
"Intended Audience :: Science/Research",
|
|
20
|
+
"Programming Language :: Python :: 3",
|
|
21
|
+
"Programming Language :: Python :: 3.8",
|
|
22
|
+
"Programming Language :: Python :: 3.9",
|
|
23
|
+
"Programming Language :: Python :: 3.10",
|
|
24
|
+
"Programming Language :: Python :: 3.11",
|
|
25
|
+
"Programming Language :: Python :: 3.12",
|
|
26
|
+
"Programming Language :: Python :: 3.13",
|
|
27
|
+
"Topic :: Multimedia :: Sound/Audio :: Speech",
|
|
28
|
+
"Topic :: Scientific/Engineering :: Artificial Intelligence",
|
|
29
|
+
"Topic :: Text Processing :: Linguistic",
|
|
30
|
+
"Natural Language :: Bulgarian",
|
|
31
|
+
]
|
|
32
|
+
|
|
33
|
+
[project.urls]
|
|
34
|
+
Homepage = "https://github.com/raditotev/bg-text-normalizer"
|
|
35
|
+
Repository = "https://github.com/raditotev/bg-text-normalizer"
|
|
36
|
+
Issues = "https://github.com/raditotev/bg-text-normalizer/issues"
|
|
37
|
+
|
|
38
|
+
[tool.setuptools]
|
|
39
|
+
license-files = ["LICENSE"]
|
|
40
|
+
|
|
41
|
+
[tool.setuptools.packages.find]
|
|
42
|
+
where = ["src"]
|
|
@@ -0,0 +1,40 @@
|
|
|
1
|
+
"""
|
|
2
|
+
bg-text-norm: Bulgarian Text Normalizer for TTS
|
|
3
|
+
=================================================
|
|
4
|
+
Converts written Bulgarian text (numbers, dates, times, currency,
|
|
5
|
+
abbreviations) into spoken word forms for TTS preprocessing.
|
|
6
|
+
|
|
7
|
+
Usage:
|
|
8
|
+
from bg_text_normalizer import normalize_text
|
|
9
|
+
result = normalize_text("На 15.02.2026 г. в 14:30 ч. цената е 1500.50 лв.")
|
|
10
|
+
|
|
11
|
+
from bg_text_normalizer import BulgarianTextNormalizer
|
|
12
|
+
normalizer = BulgarianTextNormalizer()
|
|
13
|
+
result = normalizer.normalize("...")
|
|
14
|
+
"""
|
|
15
|
+
|
|
16
|
+
from .bg_normalizer import BulgarianTextNormalizer, normalize_text
|
|
17
|
+
from .bg_numbers import number_to_words_cardinal, number_to_words_ordinal, float_to_words
|
|
18
|
+
from .bg_dates import normalize_date, normalize_year
|
|
19
|
+
from .bg_time import normalize_time
|
|
20
|
+
from .bg_currency import normalize_currency
|
|
21
|
+
from .bg_abbreviations import normalize_abbreviations, expand_abbreviation
|
|
22
|
+
from .bg_phone import normalize_phone_number
|
|
23
|
+
from .bg_roman import roman_to_arabic
|
|
24
|
+
|
|
25
|
+
__version__ = "1.0.0"
|
|
26
|
+
__all__ = [
|
|
27
|
+
'BulgarianTextNormalizer',
|
|
28
|
+
'normalize_text',
|
|
29
|
+
'number_to_words_cardinal',
|
|
30
|
+
'number_to_words_ordinal',
|
|
31
|
+
'float_to_words',
|
|
32
|
+
'normalize_date',
|
|
33
|
+
'normalize_year',
|
|
34
|
+
'normalize_time',
|
|
35
|
+
'normalize_currency',
|
|
36
|
+
'normalize_abbreviations',
|
|
37
|
+
'expand_abbreviation',
|
|
38
|
+
'normalize_phone_number',
|
|
39
|
+
'roman_to_arabic',
|
|
40
|
+
]
|
|
@@ -0,0 +1,238 @@
|
|
|
1
|
+
"""
|
|
2
|
+
Bulgarian Abbreviation Expansion
|
|
3
|
+
===================================
|
|
4
|
+
Expands common Bulgarian abbreviations to their full spoken forms.
|
|
5
|
+
|
|
6
|
+
Categories:
|
|
7
|
+
- Address abbreviations (бул., ул., пл., ж.к., кв.)
|
|
8
|
+
- Title abbreviations (г-н, г-жа, д-р, проф., инж.)
|
|
9
|
+
- Geographic (гр., с., обл.)
|
|
10
|
+
- Measurement units (км, м, кг, л, мм, см)
|
|
11
|
+
- Administrative (ЕТ, ООД, ЕООД, АД, ЕАД)
|
|
12
|
+
- Common abbreviations (т.е., т.н., напр., и т.н., и др.)
|
|
13
|
+
- Institution abbreviations (НАП, НОИ, МОН, МВР, НЗОК)
|
|
14
|
+
"""
|
|
15
|
+
|
|
16
|
+
import re
|
|
17
|
+
from typing import Dict, Tuple
|
|
18
|
+
|
|
19
|
+
# Abbreviation → (full form, is_case_sensitive)
|
|
20
|
+
# Order matters: longer abbreviations should be checked first
|
|
21
|
+
|
|
22
|
+
# Address abbreviations
|
|
23
|
+
ADDRESS_ABBREVS: Dict[str, str] = {
|
|
24
|
+
'бул.': 'булевард',
|
|
25
|
+
'ул.': 'улица',
|
|
26
|
+
'пл.': 'площад',
|
|
27
|
+
'ж.к.': 'жилищен комплекс',
|
|
28
|
+
'ж. к.': 'жилищен комплекс',
|
|
29
|
+
'кв.': 'квартал',
|
|
30
|
+
'бл.': 'блок',
|
|
31
|
+
'вх.': 'вход',
|
|
32
|
+
'ет.': 'етаж',
|
|
33
|
+
'ап.': 'апартамент',
|
|
34
|
+
'п.к.': 'пощенски код',
|
|
35
|
+
}
|
|
36
|
+
|
|
37
|
+
# Geographic abbreviations
|
|
38
|
+
GEO_ABBREVS: Dict[str, str] = {
|
|
39
|
+
'гр.': 'град',
|
|
40
|
+
'с.': 'село',
|
|
41
|
+
'обл.': 'област',
|
|
42
|
+
'общ.': 'община',
|
|
43
|
+
'р-н': 'район',
|
|
44
|
+
}
|
|
45
|
+
|
|
46
|
+
# Title abbreviations
|
|
47
|
+
TITLE_ABBREVS: Dict[str, str] = {
|
|
48
|
+
'г-н': 'господин',
|
|
49
|
+
'г-жа': 'госпожа',
|
|
50
|
+
'г-ца': 'госпожица',
|
|
51
|
+
'д-р': 'доктор',
|
|
52
|
+
'проф.': 'професор',
|
|
53
|
+
'доц.': 'доцент',
|
|
54
|
+
'акад.': 'академик',
|
|
55
|
+
'инж.': 'инженер',
|
|
56
|
+
'арх.': 'архитект',
|
|
57
|
+
'адв.': 'адвокат',
|
|
58
|
+
'св.': 'свети',
|
|
59
|
+
}
|
|
60
|
+
|
|
61
|
+
# Measurement unit abbreviations
|
|
62
|
+
MEASUREMENT_ABBREVS: Dict[str, str] = {
|
|
63
|
+
'км/ч': 'километра в час',
|
|
64
|
+
'м/с': 'метра в секунда',
|
|
65
|
+
'кв.м': 'квадратни метра',
|
|
66
|
+
'кв. м': 'квадратни метра',
|
|
67
|
+
'кв.м.': 'квадратни метра',
|
|
68
|
+
'куб.м': 'кубически метра',
|
|
69
|
+
'куб. м': 'кубически метра',
|
|
70
|
+
'км': 'километра',
|
|
71
|
+
'м': 'метра',
|
|
72
|
+
'см': 'сантиметра',
|
|
73
|
+
'мм': 'милиметра',
|
|
74
|
+
'кг': 'килограма',
|
|
75
|
+
'гр': 'грама', # Note: context-dependent (could also be "град")
|
|
76
|
+
'мг': 'милиграма',
|
|
77
|
+
'т': 'тона',
|
|
78
|
+
'л': 'литра',
|
|
79
|
+
'мл': 'милилитра',
|
|
80
|
+
'дка': 'декара',
|
|
81
|
+
'ха': 'хектара',
|
|
82
|
+
}
|
|
83
|
+
|
|
84
|
+
# Administrative/business abbreviations
|
|
85
|
+
BUSINESS_ABBREVS: Dict[str, str] = {
|
|
86
|
+
'ЕООД': 'еоод', # Spelled out as abbreviation
|
|
87
|
+
'ООД': 'оод',
|
|
88
|
+
'ЕТ': 'едноличен търговец',
|
|
89
|
+
'АД': 'акционерно дружество',
|
|
90
|
+
'ЕАД': 'еад',
|
|
91
|
+
'КД': 'командитно дружество',
|
|
92
|
+
'СД': 'събирателно дружество',
|
|
93
|
+
}
|
|
94
|
+
|
|
95
|
+
# Common text abbreviations
|
|
96
|
+
COMMON_ABBREVS: Dict[str, str] = {
|
|
97
|
+
'т.е.': 'тоест',
|
|
98
|
+
'т. е.': 'тоест',
|
|
99
|
+
'т.н.': 'така наречен',
|
|
100
|
+
'т. н.': 'така наречен',
|
|
101
|
+
'т.нар.': 'така наречен',
|
|
102
|
+
'напр.': 'например',
|
|
103
|
+
'и т.н.': 'и така нататък',
|
|
104
|
+
'и т. н.': 'и така нататък',
|
|
105
|
+
'и др.': 'и други',
|
|
106
|
+
'вж.': 'виж',
|
|
107
|
+
'ср.': 'сравни',
|
|
108
|
+
'вкл.': 'включително',
|
|
109
|
+
'изд.': 'издание',
|
|
110
|
+
'стр.': 'страница',
|
|
111
|
+
'гл.': 'глава',
|
|
112
|
+
'чл.': 'член',
|
|
113
|
+
'ал.': 'алинея',
|
|
114
|
+
'б.р.': 'бележка на редактора',
|
|
115
|
+
'б. р.': 'бележка на редактора',
|
|
116
|
+
'б.а.': 'бележка на автора',
|
|
117
|
+
'б. а.': 'бележка на автора',
|
|
118
|
+
'пр.н.е.': 'преди новата ера',
|
|
119
|
+
'пр. н. е.': 'преди новата ера',
|
|
120
|
+
'сл.н.е.': 'след новата ера',
|
|
121
|
+
'сл. н. е.': 'след новата ера',
|
|
122
|
+
'пр.Хр.': 'преди Христа',
|
|
123
|
+
'сл.Хр.': 'след Христа',
|
|
124
|
+
}
|
|
125
|
+
|
|
126
|
+
# Institution abbreviations (spelled out letter by letter or as words)
|
|
127
|
+
INSTITUTION_ABBREVS: Dict[str, str] = {
|
|
128
|
+
'НАП': 'национална агенция по приходите',
|
|
129
|
+
'НОИ': 'национален осигурителен институт',
|
|
130
|
+
'НЗОК': 'национална здравноосигурителна каса',
|
|
131
|
+
'МОН': 'министерство на образованието и науката',
|
|
132
|
+
'МВР': 'министерство на вътрешните работи',
|
|
133
|
+
'МЗ': 'министерство на здравеопазването',
|
|
134
|
+
'МФ': 'министерство на финансите',
|
|
135
|
+
'МРРБ': 'министерство на регионалното развитие и благоустройството',
|
|
136
|
+
'БНБ': 'българска народна банка',
|
|
137
|
+
'БНТ': 'бнт',
|
|
138
|
+
'БТВ': 'бтв',
|
|
139
|
+
'НС': 'народно събрание',
|
|
140
|
+
'ЕС': 'европейски съюз',
|
|
141
|
+
'ООН': 'организация на обединените нации',
|
|
142
|
+
'НАТО': 'нато',
|
|
143
|
+
'ДДС': 'данък добавена стойност',
|
|
144
|
+
'ДАНС': 'данс',
|
|
145
|
+
'ЕГН': 'единен граждански номер',
|
|
146
|
+
'БУЛСТАТ': 'булстат',
|
|
147
|
+
'ПИН': 'пин',
|
|
148
|
+
'ПДФ': 'пдф',
|
|
149
|
+
'СМС': 'есемес',
|
|
150
|
+
}
|
|
151
|
+
|
|
152
|
+
# Number sign / hash
|
|
153
|
+
SYMBOL_ABBREVS: Dict[str, str] = {
|
|
154
|
+
'№': 'номер',
|
|
155
|
+
'&': 'и',
|
|
156
|
+
'@': 'кльомба',
|
|
157
|
+
}
|
|
158
|
+
|
|
159
|
+
|
|
160
|
+
def expand_abbreviation(abbrev: str) -> str:
|
|
161
|
+
"""
|
|
162
|
+
Expand a single abbreviation to its full form.
|
|
163
|
+
|
|
164
|
+
Args:
|
|
165
|
+
abbrev: The abbreviation to expand
|
|
166
|
+
|
|
167
|
+
Returns:
|
|
168
|
+
Full form or original if not found
|
|
169
|
+
"""
|
|
170
|
+
# Check all dictionaries
|
|
171
|
+
for d in [ADDRESS_ABBREVS, GEO_ABBREVS, TITLE_ABBREVS,
|
|
172
|
+
MEASUREMENT_ABBREVS, BUSINESS_ABBREVS, COMMON_ABBREVS,
|
|
173
|
+
INSTITUTION_ABBREVS, SYMBOL_ABBREVS]:
|
|
174
|
+
if abbrev in d:
|
|
175
|
+
return d[abbrev]
|
|
176
|
+
# Case-insensitive check
|
|
177
|
+
for key, val in d.items():
|
|
178
|
+
if key.lower() == abbrev.lower():
|
|
179
|
+
return val
|
|
180
|
+
return abbrev
|
|
181
|
+
|
|
182
|
+
|
|
183
|
+
def normalize_abbreviations(text: str) -> str:
|
|
184
|
+
"""
|
|
185
|
+
Expand all recognized abbreviations in the text.
|
|
186
|
+
|
|
187
|
+
Processes abbreviations in order of specificity (longer first)
|
|
188
|
+
to avoid partial matches.
|
|
189
|
+
"""
|
|
190
|
+
# Collect all abbreviations, sorted by length (longest first)
|
|
191
|
+
all_abbrevs = {}
|
|
192
|
+
for d in [COMMON_ABBREVS, ADDRESS_ABBREVS, GEO_ABBREVS, TITLE_ABBREVS,
|
|
193
|
+
BUSINESS_ABBREVS, INSTITUTION_ABBREVS, SYMBOL_ABBREVS]:
|
|
194
|
+
all_abbrevs.update(d)
|
|
195
|
+
|
|
196
|
+
# Sort by length descending to match longer abbreviations first
|
|
197
|
+
sorted_abbrevs = sorted(all_abbrevs.items(), key=lambda x: len(x[0]), reverse=True)
|
|
198
|
+
|
|
199
|
+
for abbrev, full_form in sorted_abbrevs:
|
|
200
|
+
# Escape special regex characters in the abbreviation
|
|
201
|
+
escaped = re.escape(abbrev)
|
|
202
|
+
# Use word boundary or space/start/end boundary
|
|
203
|
+
# Be careful with dots - they're part of some abbreviations
|
|
204
|
+
pattern = r'(?<!\w)' + escaped + r'(?!\w)'
|
|
205
|
+
text = re.sub(pattern, full_form, text, flags=re.IGNORECASE)
|
|
206
|
+
|
|
207
|
+
# Handle measurement units (these need number context)
|
|
208
|
+
# Match: digit + space? + unit
|
|
209
|
+
for unit, full_form in sorted(MEASUREMENT_ABBREVS.items(),
|
|
210
|
+
key=lambda x: len(x[0]), reverse=True):
|
|
211
|
+
escaped_unit = re.escape(unit)
|
|
212
|
+
pattern = r'(\d)\s*' + escaped_unit + r'\b'
|
|
213
|
+
text = re.sub(pattern, r'\1 ' + full_form, text)
|
|
214
|
+
|
|
215
|
+
return text
|
|
216
|
+
|
|
217
|
+
|
|
218
|
+
if __name__ == '__main__':
|
|
219
|
+
test_cases = [
|
|
220
|
+
"бул. Витоша №10, гр. София",
|
|
221
|
+
"г-н Иванов е д-р по медицина",
|
|
222
|
+
"Офис на 5 км от центъра",
|
|
223
|
+
"Площ: 120 кв.м",
|
|
224
|
+
"т.е. не може да се направи",
|
|
225
|
+
"Фирма Тест ЕООД",
|
|
226
|
+
"Скорост: 60 км/ч",
|
|
227
|
+
"Заплата: 2500 лв. (вкл. ДДС)",
|
|
228
|
+
"ж.к. Люлин, бл. 305, вх. А, ет. 5, ап. 20",
|
|
229
|
+
"проф. д-р Петров",
|
|
230
|
+
"и т.н. и т.н.",
|
|
231
|
+
]
|
|
232
|
+
|
|
233
|
+
print("=== Abbreviation Expansion ===")
|
|
234
|
+
for test in test_cases:
|
|
235
|
+
result = normalize_abbreviations(test)
|
|
236
|
+
print(f" Input: {test}")
|
|
237
|
+
print(f" Output: {result}")
|
|
238
|
+
print()
|