Added support for windows and linux, removed unused function, added more logs

translate_docx.py

@@ -10,6 +10,7 @@ from docx import Document
 from docx.text.hyperlink import Hyperlink
 from docx.text.run import Run
 import nltk
+import platform
 
 nltk.download('punkt')
 nltk.download('punkt_tab')
@@ -22,21 +23,22 @@ from subprocess import Popen, PIPE
 from itertools import groupby
 import fileinput
 
-ip="192.168.20.216"
-port="8000"
+ip = "192.168.20.216"
+port = "8000"
 
-def translate(text, ip, port):
 
+def translate(text, ip, port):
     myobj = {
-                'id': '1', 
-                'src': text,
-            }
+        'id': '1',
+        'src': text,
+    }
     port = str(int(port))
     url = 'http://' + ip + ':' + port + '/translate'
-    x = requests.post(url, json = myobj)
+    x = requests.post(url, json=myobj)
     json_response = json.loads(x.text)
     return json_response['tgt']
 
+
 # Class to align original and translated sentences
 # based on https://github.com/mtuoc/MTUOC-server/blob/main/GetWordAlignments_fast_align.py
 class Aligner():
@@ -50,12 +52,19 @@ class Aligner():
         self.forward_alignment_file_path = os.path.join(temp_folder, "forward.align")
         self.reverse_alignment_file_path = os.path.join(temp_folder, "reverse.align")
 
+        if platform.system().lower() == "windows":
+            fastalign_bin = "fast_align.exe"
+            atools_bin = "atools.exe"
+        else:
+            fastalign_bin = "./fast_align"
+            atools_bin = "./atools"
+
         self.forward_command = lambda \
-                x: f'fast_align.exe -i {x} -d -T {fwd_T} -m {fwd_m} -f {forward_params_path} > {self.forward_alignment_file_path}'
+                x: f'{fastalign_bin} -i {x} -d -T {fwd_T} -m {fwd_m} -f {forward_params_path} > {self.forward_alignment_file_path}'
         self.reverse_command = lambda \
-                x: f'fast_align.exe -i {x} -d -T {rev_T} -m {rev_m} -f {reverse_params_path} -r > {self.reverse_alignment_file_path}'
+                x: f'{fastalign_bin} -i {x} -d -T {rev_T} -m {rev_m} -f {reverse_params_path} -r > {self.reverse_alignment_file_path}'
 
-        self.symmetric_command = f'atools.exe -i {self.forward_alignment_file_path} -j {self.reverse_alignment_file_path} -c grow-diag-final-and'
+        self.symmetric_command = f'{atools_bin} -i {self.forward_alignment_file_path} -j {self.reverse_alignment_file_path} -c grow-diag-final-and'
 
     def __simplify_alignment_file(self, file):
         with fileinput.FileInput(file, inplace=True, backup='.bak') as f:
@@ -131,39 +140,6 @@ def extract_paragraphs_with_runs(doc):
     return paragraphs_with_runs
 
 
-def tokenize_paragraph_with_runs2(runs_in_paragraph):
-    text_paragraph = " ".join(run["text"] for run in runs_in_paragraph)
-    sentences = sent_tokenize(text_paragraph)
-    tokenized_sentences = [word_tokenize(sentence) for sentence in sentences]
-
-    tokenized_sentences_with_style = []
-    for tokenized_sentence in tokenized_sentences:
-        tokenized_sentence_with_style = []
-        token_idx = 0
-        for run in runs_in_paragraph:
-            text_in_run = run["text"].strip()
-
-            if text_in_run == tokenized_sentence[token_idx]:
-                new_run = run.copy()
-                new_run["text"] = text_in_run
-                tokenized_sentence_with_style.append(new_run)
-                token_idx += 1
-                if token_idx >= len(tokenized_sentence):
-                    break
-            elif len(text_in_run) > len(tokenized_sentence[token_idx]):
-                if text_in_run.startswith(tokenized_sentence[token_idx]):
-                    for token in word_tokenize(text_in_run):
-                        if token == tokenized_sentence[token_idx]:
-                            new_run = run.copy()
-                            new_run["text"] = token
-                            tokenized_sentence_with_style.append(new_run)
-                        token_idx += 1
-            else:
-                raise "oops"
-        tokenized_sentences_with_style.append(tokenized_sentence_with_style)
-    return tokenized_sentences_with_style
-
-
 def tokenize_with_runs(runs, detokenizer):
     text_paragraph = detokenizer.detokenize([run["text"] for run in runs])
     sentences = sent_tokenize(text_paragraph)
@@ -194,7 +170,7 @@ def tokenize_with_runs(runs, detokenizer):
                         word_left = word_left.removeprefix(tokens_with_style[token_index]["text"])
                         token_index += 1
                 else:
-                    raise "oops"
+                    raise "Something unexpected happened I'm afraid"
         tokenized_sentences_with_style.append(sentence_with_style)
     return tokenized_sentences_with_style
 
@@ -311,8 +287,7 @@ def preprocess_runs(runs_in_paragraph):
     return new_runs
 
 
-
-def translate_document(input_file, 
+def translate_document(input_file,
                        aligner,
                        detokenizer,
                        ip="192.168.20.216",
@@ -322,7 +297,7 @@ def translate_document(input_file,
     # load original file, extract the paragraphs with their runs (which include style and formatting)
     doc = Document(input_file)
     paragraphs_with_runs = extract_paragraphs_with_runs(doc)
-   
+
     # translate each paragraph
     translated_paragraphs = []
     for paragraph in tqdm.tqdm(paragraphs_with_runs, desc="Translating paragraphs..."):
@@ -333,13 +308,17 @@ def translate_document(input_file,
 
     processed_original_paragraphs_with_runs = [preprocess_runs(runs) for runs in paragraphs_with_runs]
 
+    print("Generating alignments...")
     translated_sentences_with_style = generate_alignments(processed_original_paragraphs_with_runs,
                                                           translated_paragraphs, aligner,
                                                           temp_folder, detokenizer)
+    print("Finished alignments")
+
     # flatten the sentences into a list of tokens
     translated_tokens_with_style = [item for sublist in translated_sentences_with_style for item in sublist]
     # group the tokens by style/run
     translated_runs_with_style = group_by_style(translated_tokens_with_style, detokenizer)
+    print("Grouped by style")
 
     # group the runs by original paragraph
     translated_paragraphs_with_style = defaultdict(list)
@@ -365,4 +344,5 @@ def translate_document(input_file,
             run.font.color.rgb = item['font_color']
 
     out_doc.save("translated.docx")
+    print("Saved file")
     return "translated.docx"