Creating a corpus from different sentences to minimize the number of Translations using DeepL

deepl_corpus_creation.py

# Importing our previous translation function
from run_translation import translate_sentence

# Sentences input
sentence1 = 'I want to translate a first sentence without any link to the second one.'
sentence2 = 'The starlings ate all the cherries in one afternoon, there won\'t be any more for us.''

# Creation of the corpus as a list of strings
corpus  = [sentence1, sentence2] 

# Defining a joiner to separate the sentences
joiner = '\n____\n'

# Create a single "extended" string to load DeepL once
full_text = corpus.join(joiner)

# Running the translation process
translated_text = translate_sentence(full_text)

# Splitting the translation
corpus_translated = translated_text.split(joiner)

# Checking that we have the  correct number of sentences after translation and split
assert len(corpus)==len(corpus_translated)
print("Translation of the corpus completed.")