Skip to content
This repository has been archived by the owner on Nov 8, 2023. It is now read-only.

分詞留起來合成 #22

Merged
merged 3 commits into from
Nov 16, 2018
Merged

分詞留起來合成 #22

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
04cca11
留分詞起來合成，較無輕聲問題
sih4sing5hong5 Nov 16, 2018
4fff459
修試驗
sih4sing5hong5 Nov 16, 2018
0c5085e
flake8
sih4sing5hong5 Nov 16, 2018
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 4 additions & 2 deletions siunn1ua2ah4/settings.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -132,13 +132,15 @@

KIUNN1_KHAU2_TSHAM1_SOO3 = {
'臺語': {
'服務腔口': '閩南語',
'標音欄位': '臺羅閏號調',
'資料來源':'多元書寫',
'服務腔口': '台語',
'標音欄位': '臺羅',
'拼音': 臺灣閩南語羅馬字拼音,
},
}
for 客話 in ['四縣腔', '海陸腔', '大埔腔', '饒平腔', '詔安腔', ]:
KIUNN1_KHAU2_TSHAM1_SOO3[客話.rstrip('腔')] = {
'資料來源': '綜合標音',
'服務腔口': 客話,
'標音欄位': '臺灣客話',
'拼音': 臺灣客家話拼音,
Expand Down
17 changes: 10 additions & 7 deletions 試驗/test收著資料試驗.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,14 +16,16 @@ def setUp(self):
self.目錄 = join(dirname(abspath(__file__)), '..', '圖')
self.圖 = join(self.目錄, 'ti1a2.jpg')
self.音 = join(self.目錄, 'huan1ing5.wav')
self.字 = join(self.目錄, 'huan1ing5.srt')
self.字幕 = join(self.目錄, 'huan1ing5.srt')
self.分詞 = '逐-家｜tak8-ke1 做-伙｜tso3-hue2 來｜lai5- 𨑨-迌｜tshit4-tho5 ！'

def test_無圖用預設圖(self, 敆做伙mock):
with TemporaryDirectory() as 目錄:
檔案的所在 = join(目錄, 'a.mp4')
做影片.收著資料('閩南語', [], [self.音], [self.字], 檔案的所在)
做影片.收著資料('閩南語', [], [self.音], [self.字幕], [self.分詞], 檔案的所在)
敆做伙mock.assert_called_once_with(
[I7SIAT4_TOO5], [self.音], [self.字], 檔案的所在)
[I7SIAT4_TOO5], [self.音], [self.字幕], 檔案的所在
)

def test_圖無夠愛循環用(self, 敆做伙mock):
with TemporaryDirectory() as 目錄:
Expand All @@ -32,13 +34,14 @@ def test_圖無夠愛循環用(self, 敆做伙mock):
'閩南語',
['a.jpg', 'b.jpg'],
[self.音, self.音, self.音],
[self.字, self.字, self.字],
[self.字幕, self.字幕, self.字幕],
[self.分詞, self.分詞, self.分詞],
檔案的所在
)
敆做伙mock.assert_called_once_with(
['a.jpg', 'b.jpg', 'a.jpg'],
[self.音, self.音, self.音],
[self.字, self.字, self.字],
[self.字幕, self.字幕, self.字幕],
檔案的所在
)

Expand All @@ -47,7 +50,7 @@ def test_無音用合成音(self, 揣聲音mock, 敆做伙mock):
揣聲音mock.return_value = self.音
with TemporaryDirectory() as 目錄:
檔案的所在 = join(目錄, 'a.mp4')
做影片.收著資料('閩南語', [self.圖], [], [self.字], 檔案的所在)
做影片.收著資料('閩南語', [self.圖], [], [self.字幕], [self.分詞], 檔案的所在)
敆做伙mock.assert_called_once_with(
[self.圖], [self.音], [self.字], 檔案的所在
[self.圖], [self.音], [self.字幕], 檔案的所在
)
16 changes: 10 additions & 6 deletions 試驗/test敆檔案試驗.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,30 +12,34 @@ def setUp(self):
self.目錄 = join(dirname(abspath(__file__)), '..', '圖')
self.圖 = join(self.目錄, 'ti1a2.jpg')
self.音 = join(self.目錄, 'huan1ing5.wav')
self.字 = join(self.目錄, 'huan1ing5.srt')
self.字幕 = join(self.目錄, 'huan1ing5.srt')

def test_一組(self):
with TemporaryDirectory() as 目錄:
檔案的所在 = join(目錄, 'a.mp4')
做影片.敆做伙([self.圖], [self.音], [self.字], 檔案的所在)
做影片.敆做伙([self.圖], [self.音], [self.字幕], 檔案的所在)
self.assertTrue(isfile(檔案的所在))

def test_兩組(self):
with TemporaryDirectory() as 目錄:
檔案的所在 = join(目錄, 'b.avi')
做影片.敆做伙([self.圖, self.圖], [self.音, self.音],
[self.字, self.字], 檔案的所在)
做影片.敆做伙(
[self.圖, self.圖],
[self.音, self.音],
[self.字幕, self.字幕],
檔案的所在
)
self.assertTrue(isfile(檔案的所在))

def test_兩組的檔案佮一組無仝(self):
with TemporaryDirectory() as 目錄:
一組檔案的所在 = join(目錄, 'c.avi')
做影片.敆做伙([self.圖], [self.音], [self.字], 一組檔案的所在)
做影片.敆做伙([self.圖], [self.音], [self.字幕], 一組檔案的所在)
兩組檔案的所在 = join(目錄, 'd.avi')
做影片.敆做伙(
[self.圖, self.圖],
[self.音, self.音],
[self.字, self.字],
[self.字幕, self.字幕],
兩組檔案的所在
)
with open(一組檔案的所在, 'rb') as 一組檔案:
Expand Down
54 changes: 24 additions & 30 deletions 鬥做伙/做影片.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,9 +9,6 @@
from siunn1ua2ah4.settings import I7SIAT4_TOO5


from 臺灣言語工具.解析整理.拆文分析器 import 拆文分析器
from 臺灣言語工具.解析整理.文章粗胚 import 文章粗胚

ssl.match_hostname = lambda cert, hostname: True


Expand All @@ -20,8 +17,8 @@ class 做影片(程式腳本):
@classmethod
def 使用者提供的資料(cls, 腔口參數, 圖陣列, 聲陣列, 文字陣列, 影片存檔所在, 縮圖存檔所在):
with TemporaryDirectory() as 目錄:
字陣列 = cls.轉文本資料(腔口參數, 文字陣列, 目錄)
cls.收著資料(腔口參數, 圖陣列, 聲陣列, 字陣列, 影片存檔所在)
字幕檔陣列, 分詞陣列 = cls.轉文本資料(腔口參數, 文字陣列, 目錄)
cls.收著資料(腔口參數, 圖陣列, 聲陣列, 字幕檔陣列, 分詞陣列, 影片存檔所在)
cls._走指令([
'avconv',
'-i', 影片存檔所在,
Expand All @@ -33,29 +30,30 @@ def 使用者提供的資料(cls, 腔口參數, 圖陣列, 聲陣列, 文字陣
])

@classmethod
def 收著資料(cls, 腔口參數, 圖陣列, 聲陣列, 字陣列, 存檔所在):
def 收著資料(cls, 腔口參數, 圖陣列, 聲陣列, 字幕檔陣列, 分詞陣列, 存檔所在):
with TemporaryDirectory() as 目錄:
新聲陣列 = []
for 第幾个, (聲, 字) in enumerate(zip_longest(聲陣列, 字陣列)):
for 第幾个, (聲, 分詞) in enumerate(zip_longest(聲陣列, 分詞陣列)):
if 聲 is not None:
新聲陣列.append(聲)
else:
新聲陣列.append(
cls.揣聲音(腔口參數, 字, join(目錄, '{}.wav'.format(第幾个))))
cls.揣聲音(腔口參數, 分詞, join(目錄, '{}.wav'.format(第幾个)))
)
if len(圖陣列) == 0:
新圖陣列 = [I7SIAT4_TOO5] * len(字陣列)
新圖陣列 = [I7SIAT4_TOO5] * len(字幕檔陣列)
else:
新圖陣列 = []
while len(新圖陣列) < len(字陣列):
while len(新圖陣列) < len(字幕檔陣列):
新圖陣列.append(圖陣列[len(新圖陣列) % len(圖陣列)])
cls.敆做伙(新圖陣列, 新聲陣列, 字陣列, 存檔所在)
cls.敆做伙(新圖陣列, 新聲陣列, 字幕檔陣列, 存檔所在)

@classmethod
def 敆做伙(cls, 圖陣列, 聲陣列, 字陣列, 存檔所在):
def 敆做伙(cls, 圖陣列, 聲陣列, 字幕檔陣列, 存檔所在):
with TemporaryDirectory() as 目錄:
全部結果表 = join(目錄, 'tuan.pio')
with open(全部結果表, 'w') as 表:
for 第幾个, (圖, 聲, 字) in enumerate(zip(圖陣列, 聲陣列, 字陣列)):
for 第幾个, (圖, 聲, 字幕檔) in enumerate(zip(圖陣列, 聲陣列, 字幕檔陣列)):
結果檔 = join(目錄, 'output{}.mp4'.format(第幾个))
暫時圖 = join(目錄, 'jpg{}.jpg'.format(第幾个))
cls._走指令([
Expand All @@ -65,7 +63,7 @@ def 敆做伙(cls, 圖陣列, 聲陣列, 字陣列, 存檔所在):
])
cls._走指令([
'avconv',
'-i', 暫時圖, '-i', 聲, '-vf', 'subtitles={}'.format(字),
'-i', 暫時圖, '-i', 聲, '-vf', 'subtitles={}'.format(字幕檔),
'-s', 'svga', '-y', 結果檔,
])
print("file '{}'".format(結果檔), file=表)
Expand Down Expand Up @@ -107,9 +105,10 @@ def 轉文本資料(cls, 腔口參數, 文字陣列, 目錄):
print(r1.status, r1.reason)
print(漢羅)
raise RuntimeError()
字陣列 = []
字幕檔陣列 = []
分詞陣列 = []
for 第幾筆, 資料 in enumerate(
json.loads(r1.read().decode('utf-8'))['綜合標音']
json.loads(r1.read().decode('utf-8'))[腔口參數['資料來源']]
):
檔名 = join(目錄, '{}.srt'.format(第幾筆))
cls._陣列寫入檔案(
Expand All @@ -121,30 +120,25 @@ def 轉文本資料(cls, 腔口參數, 文字陣列, 目錄):
資料[腔口參數['標音欄位']].strip()
]
)
字陣列.append(檔名)
return 字陣列
字幕檔陣列.append(檔名)
分詞陣列.append(資料['分詞'])
return 字幕檔陣列, 分詞陣列

@classmethod
def 揣聲音(cls, 腔口參數, 字, 存檔的所在):
with open(字) as 檔案:
*_, 漢字, 臺羅 = 檔案.read().strip().split('\n')
with open(存檔的所在, 'wb') as 存檔:
存檔.write(cls.掠聲音(腔口參數, 漢字, 臺羅))
return 存檔的所在
def 揣聲音(cls, 腔口參數, 分詞, 存檔的所在):
with open(存檔的所在, 'wb') as 存檔:
存檔.write(cls.掠聲音(腔口參數, 分詞))
return 存檔的所在

@classmethod
def 掠聲音(cls, 腔口參數, 漢字, 臺羅):
句物件 = 拆文分析器.對齊句物件(
文章粗胚.數字英文中央全加分字符號(漢字),
文章粗胚.建立物件語句前處理減號(腔口參數['拼音'], 臺羅)
)
def 掠聲音(cls, 腔口參數, 分詞):
domain = "xn--lhrz38b.xn--v0qr21b.xn--kpry57d"
網址 = "/{}?{}={}&{}={}".format(
quote('語音合成'),
quote('查詢腔口'),
quote(腔口參數['服務腔口']),
quote('查詢語句'),
quote(句物件.看分詞()),
quote(分詞),
)

conn = http.client.HTTPSConnection(domain)
Expand Down