Python连接SAP BW Cube调用RFC_READ_TABLE报错的原因排查

我有一个SAP BW Cube，可通过Power BI的“SAP Business Warehouse Message Server”连接器及以下Power M代码访问：

let
    Source = SapBusinessWarehouse.Cubes("My BW Ip", "00", "001", [Implementation="2.0"]),
    ZCOP = Source{[Name="ZCOP"]}[Data],
    #"ZCOP/ZQ_COP_2" = ZCOP{[Id="ZCOP/ZQ_COP_2"]}[Data],
    #"Added Items" = Cube.Transform(#"ZCOP/ZQ_COP_2",
        {
            {Cube.AddAndExpandDimensionColumn, "[0CALMONTH]", {"[0CALMONTH].[LEVEL01]"}, {"Año/Mes natural.Año/Mes natural Nivel 01"}},
            {Cube.AddAndExpandDimensionColumn, "[0PLANT]", {"[0PLANT].[LEVEL01]"}, {"Centro.Centro Nivel 01"}},
            {Cube.AddAndExpandDimensionColumn, "[0COSTDATE]", {"[0COSTDATE].[LEVEL01]"}, {"Fecha cálculo coste.Fecha cálculo coste Nivel 01"}},
            {Cube.AddAndExpandDimensionColumn, "[0MATERIAL]", {"[0MATERIAL].[LEVEL01]"}, {"Material.Material Nivel 01"}},
            {Cube.AddAndExpandDimensionColumn, "[0COMP_CODE]", {"[0COMP_CODE].[LEVEL01]"}, {"Sociedad.Sociedad Nivel 01"}},
            {Cube.AddAndExpandDimensionColumn, "[0COSTSTATUS]", {"[0COSTSTATUS].[LEVEL01]"}, {"Status cálculo coste.Status cálculo coste Nivel 01"}},
            {Cube.AddAndExpandDimensionColumn, "[0CURTYPE]", {"[0CURTYPE].[LEVEL01]"}, {"Tipo de moneda.Tipo de moneda Nivel 01"}},
            {Cube.AddAndExpandDimensionColumn, "[0VALUATION]", {"[0VALUATION].[LEVEL01]"}, {"Vista valoración.Vista valoración Nivel 01"}},
            {Cube.AddMeasureColumn, " Análisis Calidad", "[Measures].[00O2TS3ZGQX4VVAT463GW2O53]"},
            {Cube.AddMeasureColumn, " Costos Ind. Fabricación", "[Measures].[00O2TS3ZGQX4VVAT463GW2HTJ]"},
            {Cube.AddMeasureColumn, " Gastos Admin", "[Measures].[00O2TS3ZGQX4VVAT8TY9H77HZ]"},
            {Cube.AddMeasureColumn, " Mano de Obra Directa", "[Measures].[00O2TS3ZGQX4VVAT0KHEJIEUV]"},
            {Cube.AddMeasureColumn, "Materia Prima", "[Measures].[00O2TS3ZGQX4VVASYNHA84OTZ]"},
            {Cube.AddMeasureColumn, "Material de Empaque", "[Measures].[00O2TS3ZGQX4VVASYNHA84V5J]"},
            {Cube.AddMeasureColumn, "Subcontratación", "[Measures].[00O2TS3ZGQX4VVAT8TY9H716F]"}
        })
in
    #"Added Items"

尝试用Python的pyrfc库实现相同操作，代码如下：

from pyrfc import Connection
import pandas as pd

# Define connection parameters
conn_params = {
    'user': 'user',
    'passwd': 'pass',
    'ashost': 'My IP', 
    'sysnr': '00',
    'client': '001',
    'lang': 'EN'
}

# Establish connection
conn = Connection(**conn_params)

# Read data from SAP BW cube
cube_name = 'ZCOP/ZQ_COP_2'
result = conn.call('RFC_READ_TABLE', QUERY_TABLE=cube_name)

# Close connection
conn.close()

# Convert result to DataFrame
df = pd.DataFrame(result['DATA'])

# Extract column names from result
columns = [col['FIELDNAME'] for col in result['FIELDS']]

# Set DataFrame column names
df.columns = columns

# Perform transformations similar to Power Query
# Add dimension columns
df['Año/Mes natural.Año/Mes natural Nivel 01'] = df['[0CALMONTH].[LEVEL01]']
df['Centro.Centro Nivel 01'] = df['[0PLANT].[LEVEL01]']
df['Fecha cálculo coste.Fecha cálculo coste Nivel 01'] = df['[0COSTDATE].[LEVEL01]']
df['Material.Material Nivel 01'] = df['[0MATERIAL].[LEVEL01]']
df['Sociedad.Sociedad Nivel 01'] = df['[0COMP_CODE].[LEVEL01]']
df['Status cálculo coste.Status cálculo coste Nivel 01'] = df['[0COSTSTATUS].[LEVEL01]']
df['Tipo de moneda.Tipo de moneda Nivel 01'] = df['[0CURTYPE].[LEVEL01]']
df['Vista valoración.Vista valoración Nivel 01'] = df['[0VALUATION].[LEVEL01]']

# Add measure columns
df['Análisis Calidad'] = df['[Measures].[00O2TS3ZGQX4VVAT463GW2O53]']
df['Costos Ind. Fabricación'] = df['[Measures].[00O2TS3ZGQX4VVAT463GW2HTJ]']
df['Gastos Admin'] = df['[Measures].[00O2TS3ZGQX4VVAT8TY9H77HZ]']
df['Mano de Obra Directa'] = df['[Measures].[00O2TS3ZGQX4VVAT0KHEJIEUV]']
df['Materia Prima'] = df['[Measures].[00O2TS3ZGQX4VVASYNHA84OTZ]']
df['Material de Empaque'] = df['[Measures].[00O2TS3ZGQX4VVASYNHA84V5J]']
df['Subcontratación'] = df['[Measures].[00O2TS3ZGQX4VVAT8TY9H716F]']

# Drop unnecessary columns
df.drop(columns=['[0CALMONTH].[LEVEL01]', '[0PLANT].[LEVEL01]', '[0COSTDATE].[LEVEL01]', '[0MATERIAL].[LEVEL01]',
                  '[0COMP_CODE].[LEVEL01]', '[0COSTSTATUS].[LEVEL01]', '[0CURTYPE].[LEVEL01]', '[0VALUATION].[LEVEL01]',
                  '[Measures].[00O2TS3ZGQX4VVAT463GW2O53]', '[Measures].[00O2TS3ZGQX4VVAT463GW2HTJ]',
                  '[Measures].[00O2TS3ZGQX4VVAT8TY9H77HZ]', '[Measures].[00O2TS3ZGQX4VVAT0KHEJIEUV]',
                  '[Measures].[00O2TS3ZGQX4VVASYNHA84OTZ]', '[Measures].[00O2TS3ZGQX4VVASYNHA84V5J]',
                  '[Measures].[00O2TS3ZGQX4VVAT8TY9H716F]'], inplace=True)

# Display DataFrame
print(df)

执行result = conn.call('RFC_READ_TABLE', QUERY_TABLE=cube_name)时，出现错误：

ABAPApplicationError: 5 (rc=5): key=TABLE_NOT_AVAILABLE, message=ID:SV Type:E Number:029 ZCOP/ZQ_COP_2 [MSG: class=SV, type=E, number=029, v1-4:=ZCOP/ZQ_COP_2;;;]

错误原因 #

• RFC_READ_TABLE是SAP官方标准RFC，仅支持读取ABAP字典中的透明表、簇表等数据库表对象，无法识别SAP BW的Cube或Query。
• 你要访问的ZCOP/ZQ_COP_2是BW查询对象，并非传统ABAP数据库表，因此RFC抛出"表不存在"的错误。
• Power BI的BW连接器通过BW专用的BICS（Business Intelligence Consumer Service）接口访问Cube和查询，与RFC_READ_TABLE的底层逻辑完全不同。

解决方案 #

方案1：使用专用BW访问库pysapbw #

pysapbw是Python专属的SAP BW访问库，直接支持BICS接口，可直接读取Cube或Query数据：

1. 安装库：

pip install pysapbw

2. 适配需求的代码示例：

from pysapbw import SAPBW
import pandas as pd

# 建立BW连接
bw = SAPBW(
    host="My BW Ip",
    system_number="00",
    client="001",
    username="user",
    password="pass",
    language="EN"
)

# 加载目标查询并转换为DataFrame
query = bw.get_query("ZCOP/ZQ_COP_2")
df = query.to_dataframe()

# 关闭连接
bw.close()

# 调整列名以匹配Power M的结果
column_mapping = {
    "[0CALMONTH].[LEVEL01]": "Año/Mes natural.Año/Mes natural Nivel 01",
    "[0PLANT].[LEVEL01]": "Centro.Centro Nivel 01",
    "[0COSTDATE].[LEVEL01]": "Fecha cálculo coste.Fecha cálculo coste Nivel 01",
    "[0MATERIAL].[LEVEL01]": "Material.Material Nivel 01",
    "[0COMP_CODE].[LEVEL01]": "Sociedad.Sociedad Nivel 01",
    "[0COSTSTATUS].[LEVEL01]": "Status cálculo coste.Status cálculo coste Nivel 01",
    "[0CURTYPE].[LEVEL01]": "Tipo de moneda.Tipo de moneda Nivel 01",
    "[0VALUATION].[LEVEL01]": "Vista valoración.Vista valoración Nivel 01",
    "[Measures].[00O2TS3ZGQX4VVAT463GW2O53]": "Análisis Calidad",
    "[Measures].[00O2TS3ZGQX4VVAT463GW2HTJ]": "Costos Ind. Fabricación",
    "[Measures].[00O2TS3ZGQX4VVAT8TY9H77HZ]": "Gastos Admin",
    "[Measures].[00O2TS3ZGQX4VVAT0KHEJIEUV]": "Mano de Obra Directa",
    "[Measures].[00O2TS3ZGQX4VVASYNHA84OTZ]": "Materia Prima",
    "[Measures].[00O2TS3ZGQX4VVASYNHA84V5J]": "Material de Empaque",
    "[Measures].[00O2TS3ZGQX4VVAT8TY9H716F]": "Subcontratación"
}

df = df.rename(columns=column_mapping)
# 移除原始列（可选）
df = df.drop(columns=list(column_mapping.keys()), errors="ignore")

print(df)

方案2：自定义ABAP RFC函数 #

如果无法使用第三方库，可联系SAP ABAP开发人员创建自定义RFC函数，内部通过BW API（如RSDRI_INFOPROV_READ）读取查询数据：

1. 开发自定义RFC：函数输入参数为查询名称，输出为包含所需维度和指标的结构化数据表。
2. Python端调用示例：

from pyrfc import Connection
import pandas as pd

conn_params = {
    'user': 'user',
    'passwd': 'pass',
    'ashost': 'My BW Ip', 
    'sysnr': '00',
    'client': '001',
    'lang': 'EN'
}

conn = Connection(**conn_params)
# 调用自定义RFC，假设函数名为Z_BW_QUERY_READ
result = conn.call('Z_BW_QUERY_READ', QUERY_NAME='ZCOP/ZQ_COP_2')
conn.close()

# 转换为DataFrame
df = pd.DataFrame(result['QUERY_DATA'])
# 根据自定义RFC的输出结构调整列名
print(df)

内容的提问来源于stack exchange，提问作者Gustavo A. Marín Acevedo