github/bilingual_book_maker
1
1
Fork 0
mirror of https://github.com/yihong0618/bilingual_book_maker.git synced 2025-06-06 11:35:49 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat: support single_translate for txt & srt (#265)

Browse Source 
* feat: support single_translate for txt & srt

* style: black format

* chore: change func name
This commit is contained in:
zstone12 2023-05-07 21:22:46 +08:00 committed by GitHub
parent 60df5f2723
commit b5060b8746
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 24 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
book_maker/loader/srt_loader.py
View File

@ -43,6 +43,7 @@ class SRTBookLoader(BaseBookLoader):
self.test_num = test_num self.test_num = test_num
self.accumulated_num = 1 self.accumulated_num = 1
self.blocks = [] self.blocks = []
self.single_translate = single_translate
self.resume = resume self.resume = resume
self.bin_path = f"{Path(srt_name).parent}/.{Path(srt_name).stem}.temp.bin" self.bin_path = f"{Path(srt_name).parent}/.{Path(srt_name).stem}.temp.bin"
@ -76,6 +77,9 @@ class SRTBookLoader(BaseBookLoader):
def _get_block_text(self, block): def _get_block_text(self, block):
return f"{block['number']}\n{block['time']}\n{block['text']}" return f"{block['number']}\n{block['time']}\n{block['text']}"
def _get_block_except_text(self, block):
return f"{block['number']}\n{block['time']}"
def _concat_blocks(self, sliced_text: str, text: str): def _concat_blocks(self, sliced_text: str, text: str):
return f"{sliced_text}\n\n{text}" if sliced_text else text return f"{sliced_text}\n\n{text}" if sliced_text else text
@ -186,7 +190,7 @@ class SRTBookLoader(BaseBookLoader):
translated_blocks = [] translated_blocks = []
# try to translate one by one, so don't accumulate too much # try to translate one by one, so don't accumulate too much
print( print(
f"retry it one by one: {self.blocks[begin]['number']} - {self.blocks[end-1]['number']}" f"retry it one by one: {self.blocks[begin]['number']} - {self.blocks[end - 1]['number']}"
) )
for block in self.blocks[begin:end]: for block in self.blocks[begin:end]:
try: try:
@ -210,15 +214,25 @@ class SRTBookLoader(BaseBookLoader):
for i, block in enumerate(translated_blocks): for i, block in enumerate(translated_blocks):
text = block.get("text", "") text = block.get("text", "")
self.p_to_save.append(text) self.p_to_save.append(text)
self.bilingual_result.append( if self.single_translate:
f"{self._get_block_text(self.blocks[begin + i])}\n{text}" self.bilingual_result.append(
) f"{self._get_block_except_text(self.blocks[begin + i])}\n{text}"
)
else:
self.bilingual_result.append(
f"{self._get_block_text(self.blocks[begin + i])}\n{text}"
)
else: else:
for i, block in enumerate(self.blocks[begin:end]): for i, block in enumerate(self.blocks[begin:end]):
text = self.p_to_save[begin + i] text = self.p_to_save[begin + i]
self.bilingual_result.append( if self.single_translate:
f"{self._get_block_text(self.blocks[begin + i])}\n{text}" self.bilingual_result.append(
) f"{self._get_block_except_text(self.blocks[begin + i])}\n{text}"
)
else:
self.bilingual_result.append(
f"{self._get_block_text(self.blocks[begin + i])}\n{text}"
)
index += end - begin index += end - begin
if self.is_test and index > self.test_num: if self.is_test and index > self.test_num:

4
book_maker/loader/txt_loader.py
View File

@ -33,6 +33,7 @@ class TXTBookLoader(BaseBookLoader):
self.bilingual_temp_result = [] self.bilingual_temp_result = []
self.test_num = test_num self.test_num = test_num
self.batch_size = 10 self.batch_size = 10
self.single_translate = single_translate
try: try:
with open(f"{txt_name}", encoding="utf-8") as f: with open(f"{txt_name}", encoding="utf-8") as f:
@ -73,7 +74,8 @@ class TXTBookLoader(BaseBookLoader):
print(e) print(e)
raise Exception("Something is wrong when translate") from e raise Exception("Something is wrong when translate") from e
self.p_to_save.append(temp) self.p_to_save.append(temp)
self.bilingual_result.append(batch_text) if not self.single_translate:
self.bilingual_result.append(batch_text)
self.bilingual_result.append(temp) self.bilingual_result.append(temp)
index += self.batch_size index += self.batch_size
if self.is_test and index > self.test_num: if self.is_test and index > self.test_num: