添加XML回退解析机制支持复杂Excel文件
当pandas无法解析某些包含非标准元素的Excel文件时， 添加了XML直接解析功能来提取工作表名称和数据。 - 实现了`_extract_sheet_names_from_xml`方法从XML提取工作表名称 - 实现了`_read_excel_sheet_xml`方法直接解析Excel XML数据 - 添加多种命名空间支持以处理不同Excel格式 - 在pandas解析失败时自动回退到XML解析方式 fix(excel-storage-service): 修复XML解析中的命名空间问题 改进了XML解析逻辑，添加对多种命名空间的支持， 使用通配符查找元素以兼容不同Excel文件格式。 refactor(table-rag-service): 优化XML解析逻辑提高兼容性 统一了XML解析的命名空间处理方式， 改进了元素查找逻辑以更好地支持不同Excel格式。 feat(frontend): 添加RAG向量检索和索引重建功能 - 实现了RAG状态查看、搜索和索引重建接口 - 添加了前端RAG检索界面组件 - 增加了错误处理和加载状态提示
This commit is contained in:
@@ -45,8 +45,25 @@ class ExcelStorageService:
|
||||
return []
|
||||
content = z.read('xl/workbook.xml')
|
||||
root = ET.fromstring(content)
|
||||
ns = {'main': 'http://purl.oclc.org/ooxml/spreadsheetml/main'}
|
||||
sheets = root.findall('.//main:sheet', ns)
|
||||
|
||||
# 尝试多种命名空间
|
||||
namespaces = [
|
||||
'http://schemas.openxmlformats.org/spreadsheetml/2006/main',
|
||||
'http://purl.oclc.org/ooxml/spreadsheetml/main',
|
||||
]
|
||||
|
||||
for ns_uri in namespaces:
|
||||
ns = {'main': ns_uri}
|
||||
sheets = root.findall('.//main:sheet', ns)
|
||||
if sheets:
|
||||
names = [s.get('name') for s in sheets if s.get('name')]
|
||||
if names:
|
||||
return names
|
||||
|
||||
# 尝试通配符
|
||||
sheets = root.findall('.//{*}sheet')
|
||||
if not sheets:
|
||||
sheets = root.findall('.//sheet')
|
||||
return [s.get('name') for s in sheets if s.get('name')]
|
||||
except Exception:
|
||||
return []
|
||||
@@ -79,72 +96,77 @@ class ExcelStorageService:
|
||||
if 'xl/sharedStrings.xml' in z.namelist():
|
||||
ss_content = z.read('xl/sharedStrings.xml')
|
||||
ss_root = ET.fromstring(ss_content)
|
||||
ns = {'main': 'http://purl.oclc.org/ooxml/spreadsheetml/main'}
|
||||
for si in ss_root.findall('.//main:si', ns):
|
||||
t = si.find('.//main:t', ns)
|
||||
shared_strings.append(t.text if t is not None else '')
|
||||
for si in ss_root.iter():
|
||||
if si.tag.endswith('}si') or si.tag == 'si':
|
||||
t = si.find('.//{*}t')
|
||||
shared_strings.append(t.text if t is not None and t.text else '')
|
||||
|
||||
sheet_file = f'xl/worksheets/sheet{sheet_index}.xml'
|
||||
sheet_content = z.read(sheet_file)
|
||||
root = ET.fromstring(sheet_content)
|
||||
ns = {'main': 'http://purl.oclc.org/ooxml/spreadsheetml/main'}
|
||||
|
||||
rows_data = []
|
||||
for row in root.findall('.//main:row', ns):
|
||||
row_idx = int(row.get('r', 0))
|
||||
if row_idx <= header_row + 1:
|
||||
continue
|
||||
headers = {}
|
||||
|
||||
row_cells = {}
|
||||
for cell in row.findall('main:c', ns):
|
||||
cell_ref = cell.get('r', '')
|
||||
col_letters = ''.join(filter(str.isalpha, cell_ref))
|
||||
cell_type = cell.get('t', 'n')
|
||||
v = cell.find('main:v', ns)
|
||||
for row in root.iter():
|
||||
if row.tag.endswith('}row') or row.tag == 'row':
|
||||
row_idx = int(row.get('r', 0))
|
||||
|
||||
if v is not None and v.text:
|
||||
if cell_type == 's':
|
||||
try:
|
||||
val = shared_strings[int(v.text)]
|
||||
except (ValueError, IndexError):
|
||||
val = v.text
|
||||
elif cell_type == 'b':
|
||||
val = v.text == '1'
|
||||
else:
|
||||
val = v.text
|
||||
else:
|
||||
val = None
|
||||
row_cells[col_letters] = val
|
||||
# 收集表头行
|
||||
if row_idx == header_row + 1:
|
||||
for cell in row:
|
||||
if cell.tag.endswith('}c') or cell.tag == 'c':
|
||||
cell_ref = cell.get('r', '')
|
||||
col_letters = ''.join(filter(str.isalpha, cell_ref))
|
||||
cell_type = cell.get('t', 'n')
|
||||
v = cell.find('{*}v')
|
||||
if v is not None and v.text:
|
||||
if cell_type == 's':
|
||||
try:
|
||||
headers[col_letters] = shared_strings[int(v.text)]
|
||||
except (ValueError, IndexError):
|
||||
headers[col_letters] = v.text
|
||||
else:
|
||||
headers[col_letters] = v.text
|
||||
else:
|
||||
headers[col_letters] = col_letters
|
||||
continue
|
||||
|
||||
if row_cells:
|
||||
rows_data.append(row_cells)
|
||||
if row_idx <= header_row + 1:
|
||||
continue
|
||||
|
||||
row_cells = {}
|
||||
for cell in row:
|
||||
if cell.tag.endswith('}c') or cell.tag == 'c':
|
||||
cell_ref = cell.get('r', '')
|
||||
col_letters = ''.join(filter(str.isalpha, cell_ref))
|
||||
cell_type = cell.get('t', 'n')
|
||||
v = cell.find('{*}v')
|
||||
|
||||
if v is not None and v.text:
|
||||
if cell_type == 's':
|
||||
try:
|
||||
val = shared_strings[int(v.text)]
|
||||
except (ValueError, IndexError):
|
||||
val = v.text
|
||||
elif cell_type == 'b':
|
||||
val = v.text == '1'
|
||||
else:
|
||||
val = v.text
|
||||
else:
|
||||
val = None
|
||||
row_cells[col_letters] = val
|
||||
|
||||
if row_cells:
|
||||
rows_data.append(row_cells)
|
||||
|
||||
if not rows_data:
|
||||
return pd.DataFrame()
|
||||
|
||||
df = pd.DataFrame(rows_data)
|
||||
|
||||
if header_row >= 0:
|
||||
first_row_sheet = f'xl/worksheets/sheet{sheet_index}.xml'
|
||||
sheet_content = z.read(first_row_sheet)
|
||||
root = ET.fromstring(sheet_content)
|
||||
first_row = root.find(f'.//main:row[@r="{header_row + 1}"]', ns)
|
||||
if first_row is not None:
|
||||
headers = {}
|
||||
for cell in first_row.findall('main:c', ns):
|
||||
cell_ref = cell.get('r', '')
|
||||
col_letters = ''.join(filter(str.isalpha, cell_ref))
|
||||
cell_type = cell.get('t', 'n')
|
||||
v = cell.find('main:v', ns)
|
||||
if v is not None and v.text:
|
||||
if cell_type == 's':
|
||||
try:
|
||||
headers[col_letters] = shared_strings[int(v.text)]
|
||||
except (ValueError, IndexError):
|
||||
headers[col_letters] = v.text
|
||||
else:
|
||||
headers[col_letters] = v.text
|
||||
df.columns = [headers.get(col, col) for col in df.columns]
|
||||
if headers:
|
||||
df.columns = [headers.get(col, col) for col in df.columns]
|
||||
|
||||
return df
|
||||
except Exception as e:
|
||||
|
||||
Reference in New Issue
Block a user