'
tr_index = 0
tr_text_list = []
last_node_level = 0
num_pr_dict = {}
# 直接子节点用child表示，所有子节点用all表示
for table_child in table.childNodes:
if 'w:tr' in str(table_child):
tr = table_child
tr_child_nodes = tr.childNodes
tc_index = 0
tc_text_list = []
for tr_child in tr_child_nodes:
if 'w:tc' in str(tr_child).split(' '):
tc_text = ""
tc = tr_child
# 获取一格占多少列，相当于colspan
col_span = tc.getElementsByTagName("w:gridSpan")
if col_span:
col_span = int(col_span[0].getAttribute("w:val"))
else:
col_span = 1
# 获取是否是合并单元格的下一个空单元格，相当于rowspan
is_merge = tc.getElementsByTagName("w:vMerge")
if is_merge:
is_merge = is_merge[0].getAttribute("w:val")
if is_merge == "continue":
col_span_index = 0
real_tc_index = 0
if 0 <= tr_index - 1 < len(tr_text_list):
for tc_colspan in tr_text_list[tr_index - 1]:
if col_span_index < tc_index:
col_span_index += tc_colspan[1]
real_tc_index += 1
if real_tc_index < len(tr_text_list[tr_index - 1]):
tc_text = tr_text_list[tr_index - 1][real_tc_index][0]
# 设置colspan
table_text = table_text + "| "
# 放入文本
tc_child_nodes = tc.childNodes
for tc_child in tc_child_nodes:
if 'w:tbl' in str(tc_child).split(' '):
# 嵌套在tc中的表格
tc_text += recursion_read_table(tc_child)
if 'w:p' in str(tc_child).split(' '):
tc_p_all_nodes = tc_child.getElementsByTagName("*")
_t_list, _, num_pr_dict, last_node_level = read_p_text(unique_type_dir,
tc_child,
last_node_level,
num_pr_dict,
numbering_xml,
document_xml_rels)
# print('_t_list', _t_list)
tc_text += ''.join(_t_list)
# for tc_p_all in tc_p_all_nodes:
# if 'w:t' in str(tc_p_all).split(' '):
# # w:t必须加childNodes[0]才能读文本
# tc_text += tc_p_all.childNodes[0].nodeValue
# 结束该tc
table_text = table_text + tc_text + " | "
tc_index += 1
tc_text_list.append([tc_text, col_span])
# 结束该tr
table_text += ""
tr_index += 1
tr_text_list.append(tc_text_list)
# 结束该table
table_text += "
"
return table_text
log("into read_xml_table")
try:
body = document_xml.getElementsByTagName("w:body")[0]
table_text_list = []
body_nodes = body.childNodes
for node in body_nodes:
if 'w:tbl' in str(node).split(' '):
_table = node
_table_text = recursion_read_table(_table)
table_text_list.append(_table_text)
return table_text_list
except Exception as e:
log("read_xml_table error")
print("read_xml_table", traceback.print_exc())
return [-1]
@timeout(25, timeout_exception=TimeoutError)
def parse_xml(path):
# 解析xml
DOMTree = xml.dom.minidom.parse(path)
collection = DOMTree.documentElement
return collection
@timeout(25, timeout_exception=TimeoutError)
def parse_xml2(path):
# 解析xml
tree = xml.etree.ElementTree.parse(path)
root = tree.getroot()
return root
class DocxConvert:
def __init__(self, path, unique_type_dir):
self._doc = _Document(path)
self.path = path
self.unique_type_dir = unique_type_dir
# 解压docx
try:
f = zipfile.ZipFile(path)
for file in f.namelist():
if "word/" in str(file):
f.extract(file, self.unique_type_dir)
f.close()
except Exception as e:
log("docx format error!")
self._doc.error_code = [-3]
# 读取内容
try:
self.document_xml = parse_xml(self.unique_type_dir + "word/document.xml")
if os.path.exists(self.unique_type_dir + "word/numbering.xml"):
self.numbering_xml = parse_xml(self.unique_type_dir + "word/numbering.xml")
else:
self.numbering_xml = []
if os.path.exists(self.unique_type_dir + "word/_rels/document.xml.rels"):
self.document_xml_rels = parse_xml2(self.unique_type_dir + "word/_rels/document.xml.rels")
else:
self.document_xml_rels = []
except FileNotFoundError:
# 找不到解压文件，就用html格式读
log('FileNotFoundError')
self._doc.error_code = None
except TimeoutError:
log("parse_xml timeout")
self._doc.error_code = [-4]
@memory_decorator
def init_package(self):
# 各个包初始化
try:
self.docx = docx.Document(self.path)
self.zip = zipfile.ZipFile(self.path)
except:
log("cannot open docx!")
traceback.print_exc()
self._doc.error_code = [-3]
def convert(self):
self._page = _Page(None, 0)
# 先判断特殊doc文件，可能是html文本
is_html_doc = False
try:
with open(self.path, 'r') as f:
html_str = f.read()
if re.search('