def docx2pdf(doc_fn, pdf_fn):
log.info("Converting: %s --> %s" %(os.path.basename(doc_fn), os.path.basename(pdf_fn)))
if os.path.exists(pdf_fn):
log.info("File exist and skip.")
return
word = client.DispatchEx("Word.Application")
doc_trans = word.Documents.Open(doc_fn)
doc_trans.SaveAs(pdf_fn, FileFormat=17)
doc_trans.Close()
def translate_doc(doc_raw, doc_trans, trans_cache):
if os.path.exists(doc_trans):
log.info("File %s exists, translate_doc will be skipped." %
(doc_trans))
return
doc = docx.Document(doc_raw)
process_paragraphs = doc.paragraphs
log.info("Translating from caiyun...")
for i, batch_par in tqdm(enumerate(batch(process_paragraphs, size=100))):
source = []
for j, par in enumerate(batch_par):
en_text = para2text(par)
if len(en_text) == 0:
continue
if en_text in trans_cache:
continue
source.append(en_text)
if len(source) == 0:
continue
translated = translate_caiyun(source)
for src, target in zip(source, translated):
trans_cache[src] = target
log.info("Translate local...")
for par in tqdm(process_paragraphs):
en_text = para2text(par)
if len(en_text) == 0:
continue
if en_text in trans_cache:
par.text = trans_cache[en_text]
log.info("Translate done and save.")
doc.save(doc_trans)
def merge_pages(pdfTransFn1,
pdfTransFn2,
dst_fn,
page_nums1=None,
page_nums2=None):
infile1 = PdfFileReader(pdfTransFn1, 'rb')
infile2 = PdfFileReader(pdfTransFn2, 'rb')
if page_nums1 is None:
page_nums1 = [("first", [i]) for i in range(infile1.getNumPages())]
if page_nums2 is None:
page_nums2 = [("second", [i]) for i in range(infile2.getNumPages())]
pages_sides = cross_iter(page_nums1, page_nums2)
log.info("Merge Pages: %d+%d --> %s" %
(len(page_nums1), len(page_nums2), dst_fn))
# if os.path.exists(dst_fn):
# log.info("File Exist and skip.")
# return
output = PdfFileWriter()
for (side, page_nums) in pages_sides:
if side == "first":
pages = get_page_from_nums(infile1, page_nums)
elif side == "second":
pages = get_page_from_nums(infile2, page_nums)
else:
raise Exception("F**k!")
for p in pages:
output.addPage(p)
# 只考虑头个文件的书签
# bookmarks = infile1.getOutlines()
# if len(bookmarks) > 0:
# for bm in bookmarks:
# bm_page_num = infile1.getDestinationPageNumber(bm)
# 利用python处理pdf：奇数页pdf末尾添加一个空白页 - https://zhuanlan.zhihu.com/p/34246341
# 青梅煮马: 刚好遇到这个问题 把 PyPDF2\utils.py 第238行的'latin-1'编码修改为'uft-8'即可
# 上面的方法会影响兼容性，改下面的方法
# PyPDF2 编码问题’latin-1′ codec can’t encode characters in position 8-11: ordinal not in range(256) https://www.codenong.com/cs105218309/
with open(dst_fn, 'wb') as f:
output.write(f)
def translate_doc(doc_raw, doc_trans, trans_cache):
doc = docx.Document(doc_raw)
process_paragraphs = doc.paragraphs
log.info("Translating from caiyun...")
for batch_par in tqdm(batch(process_paragraphs, size=100)):
source = []
for par in batch_par:
en_text = para2text(par)
if len(en_text) == 0:
continue
if en_text in trans_cache:
continue
source.append(en_text)
if len(source) == 0:
continue
translated = translate_caiyun(source)
for src,target in zip(source, translated):
trans_cache[src] = target
log.info("Translate local...")
for par in tqdm(process_paragraphs):
en_text = para2text(par)
if len(en_text) == 0:
continue
if en_text in trans_cache:
par.text = trans_cache[en_text]
log.info("Translate done and save.")
doc.save(doc_trans)
def convert_pdf_to_docx_v2(inputFileName, outputFileName):
log.info("Converting: %s --> %s" % (os.path.basename(inputFileName), os.path.basename(outputFileName)))
if os.path.exists(outputFileName):
log.info("File Exist, skip...")
return
pdf2word = PDF2Word.PDF2Word()
try:
pdf2word.setConversionMethod(PDF2Word.optConversionMethod.CNV_METHOD_USE_TEXTBOXES)
pdf2word.setOutputDocumentFormat(PDF2Word.optOutputDocumentFormat.OPT_OUTPUT_DOCX_VIA_OFFICE)
pdf2word.setDocumentType(PDF2Word.optDocumentType.DOCTYPE_MULTI_COLUMN)
pdf2word.setAdjustSpacing(True)
pdf2word.ConvertToWord(inputFileName, outputFileName, "", 0, -1)
except PDF2Word.PDF2WordException as ex:
log.info(ex)
sys.exit()
def convert_pdf_to_docx_v1(pdf_fn, word_fn):
log.info("Converting: %s --> %s" % (os.path.basename(pdf_fn), os.path.basename(word_fn)))
if os.path.exists(word_fn):
log.info("File Exist, skip...")
return
pdf2word = PDF2Word.PDF2Word()
try:
pdf2word.setOutputDocumentFormat(PDF2Word.optOutputDocumentFormat.OPT_OUTPUT_DOCX)
pdf2word.setConnectHyphens(True)
pdf2word.setShrinkCharacterSpacingToPreventWrap(True)
pdf2word.setFileConversionTimeout(600000)
pdf2word.ConvertToWord(pdf_fn, word_fn, "", 0, -1)
except PDF2Word.PDF2WordException as ex:
log.info(ex)
sys.exit()
def remove_first_page(pdf_fn, dst_fn):
log.info("Converting: %s --> %s" %(os.path.basename(pdf_fn), os.path.basename(dst_fn)))
if os.path.exists(dst_fn):
log.info("File Exist, skip...")
return dst_fn
infile = PdfFileReader(pdf_fn, 'rb')
output = PdfFileWriter()
skip_pages = [0]
for i in range(infile.getNumPages()):
if i in skip_pages:
continue
p = infile.getPage(i)
output.addPage(p)
with open(dst_fn, 'wb') as f:
output.write(f)
log.info("Done.")
return dst_fn
check_file_exists(inputFileName)
check_dir_exists(outputDirName)
cache_trans = dc.Cache(outputDirName)
if mode not in SUPPORTED_MODES:
print("Supported Modes: %s" %(",".join(SUPPORTED_MODES)))
print("Check help for how to choose modes.")
sys.exit()
# 第0步：将原始PDF复制到目标目录
inputBaseNameWithoutExt = os.path.basename(inputFileName).replace(".pdf", "")
if mode.lower().startswith("single"):
# 第1步：将PDF分为2个，以应对该共享软件每次间隔1页留空白，方法是将PDF的第1页删掉
pdfFileName1 = os.path.join(outputDirName, "%s_1_RawPart1.pdf" %(inputBaseNameWithoutExt))
pdfFileName2 = os.path.join(outputDirName, "%s_1_RawPart2.pdf" % (inputBaseNameWithoutExt))
log.info("%s, %s" %(pdfFileName1, pdfFileName2))
if not os.path.exists(pdfFileName1):
copyfile(inputFileName, pdfFileName1)
remove_first_page(pdfFileName1, pdfFileName2)
# 第2步：分别将这2个PDF转为DOCX
docFileName1 = os.path.join(outputDirName, "%s_2_RawPart1.docx" %(inputBaseNameWithoutExt))
docFileName2 = os.path.join(outputDirName, "%s_2_RawPart2.docx" %(inputBaseNameWithoutExt))
log.info("%s, %s" %(docFileName1, docFileName2))
convert_pdf_to_docx(pdfFileName1, docFileName1)
convert_pdf_to_docx(pdfFileName2, docFileName2)
# 第3步：将这2个DOCX翻译，仍然保存为DOCX
docTransFn1 = os.path.join(outputDirName, "%s_3_TranslatedPart1.docx" %(inputBaseNameWithoutExt))
docTransFn2 = os.path.join(outputDirName, "%s_3_TranslatedPart2.docx" %(inputBaseNameWithoutExt))
log.info("%s, %s" %(docTransFn1, docTransFn2))
