10.idb_db、sqlalchemy在nuitka打包文件中下如何加载

---

一、 前言

在使用nuitka在32位环境下打包ibm_db+sqlachemy遇到的一些问题记录！

二、ibm_db与sqlachemy结合问题

1.需要手动加载ibm_db模块给sqlachemy

报“Can’t load plugin: sqlalchemy.dialects:db2.ibm_db”错误，如下:

  File "<string>", line 2, in create_engine
  File "C:\Users\Administrator\Desktop\py\DataToolsExe\testDtsRun.dist\sqlalchemy\util\deprecations.py", line 309, in warned
    return fn(*args, **kwargs)
  File "C:\Users\Administrator\Desktop\py\DataToolsExe\testDtsRun.dist\sqlalchemy\engine\create.py", line 534, in create_engine
    entrypoint = u._get_entrypoint()
  File "C:\Users\Administrator\Desktop\py\DataToolsExe\testDtsRun.dist\sqlalchemy\engine\url.py", line 661, in _get_entrypoint
    cls = registry.load(name)
  File "C:\Users\Administrator\Desktop\py\DataToolsExe\testDtsRun.dist\sqlalchemy\util\langhelpers.py", line 343, in load
    raise exc.NoSuchModuleError(
sqlalchemy.exc.NoSuchModuleError: Can't load plugin: sqlalchemy.dialects:db2.ibm_db

解决办法：
修改sqlalchemy\util\langhelpers.py文件，加入最后2行

    def load(self, name):
        if name in self.impls:
            return self.impls[name]()

        if self.auto_fn:
            loader = self.auto_fn(name)
            if loader:
                self.impls[name] = loader
                return loader()

		# 在后面报错之前返回idb_db模块，以下修改不适用其他情况，例如mysql，可能MySQL在前面就找到模块了。
        if name == 'db2.ibm_db':
            from ibm_db_sa.ibm_db import DB2Dialect_ibm_db as r
            return r
  
        raise exc.NoSuchModuleError(
            "Can't load plugin: %s:%s" % (self.group, name)
        )

2.sqlachemy版本问题

由于sqlachemy版本更新比ibm_db、ibm_db_sa更新要快。调用过程出错。

  File "<string>", line 2, in create_engine
  File "C:\Users\Administrator\Desktop\py\DataToolsExe\testDtsRun.dist\sqlalchemy\util\deprecations.py", line 309, in warned
    return fn(*args, **kwargs)
  File "C:\Users\Administrator\Desktop\py\DataToolsExe\testDtsRun.dist\sqlalchemy\engine\create.py", line 534, in create_engine
    entrypoint = u._get_entrypoint()
  File "C:\Users\Administrator\Desktop\py\DataToolsExe\testDtsRun.dist\sqlalchemy\engine\url.py", line 661, in _get_entrypoint
    cls = registry.load(name)
  File "C:\Users\Administrator\Desktop\py\DataToolsExe\testDtsRun.dist\sqlalchemy\util\langhelpers.py", line 343, in load
    from ibm_db_sa.ibm_db import DB2Dialect_ibm_db as r
  File "C:\Users\Administrator\Desktop\py\DataToolsExe\testDtsRun.dist\ibm_db_sa\__init__.py", line 22, in <module>
    from . import ibm_db, pyodbc, base, zxjdbc
  File "C:\Users\Administrator\Desktop\py\DataToolsExe\testDtsRun.dist\ibm_db_sa\zxjdbc.py", line 32, in <module>
    from sqlalchemy.connectors.zxJDBC import ZxJDBCConnector
ModuleNotFoundError: No module named 'sqlalchemy.connectors.zxJDBC'

版本适配方法：
ibm_db/ibm_db_sa用最新版本，默认即可；
sqlachemy降版本，例如1.3.19

3.nuitka提示编码错误

使用nuitka进行打包时提示错误（32位，非调试模式）；
注：调试模式没有下面错误提示。

  File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38-32\lib\site-packages\nuitka\__main__.py", line 142, in <module>
    main()
  File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38-32\lib\site-packages\nuitka\__main__.py", line 128, in main
    MainControl.main()
  File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38-32\lib\site-packages\nuitka\MainControl.py", line 916, in main
    result, options = compileTree()
  File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38-32\lib\site-packages\nuitka\MainControl.py", line 793, in compileTree
    loader_code = LoaderCodes.getMetapathLoaderBodyCode(bytecode_accessor)
  File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38-32\lib\site-packages\nuitka\codegen\LoaderCodes.py", line 120, in getMetapathLoaderBodyCode
    getModuleMetapathLoaderEntryCode(
  File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38-32\lib\site-packages\nuitka\codegen\LoaderCodes.py", line 55, in getModuleMetapathLoaderEntryCode
    Plugins.encodeDataComposerName(module.getFullName().asString())
  File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38-32\lib\site-packages\nuitka\plugins\Plugins.py", line 872, in encodeDataComposerName
    name = name.encode("latin1")
UnicodeEncodeError: 'latin-1' codec can't encode characters in position 17-20: ordinal not in range(256)

解决办法：
打开错误提示文件看看里面什么鬼。原文如下:

nuitka\plugins\Plugins.py

    def encodeDataComposerName(cls, name):
        if str is not bytes:
            # Encoding needs to match generated source code output.
            name = name.encode("latin1")

修改如下：

    def encodeDataComposerName(cls, name):
        #这一句逻辑错了，本来是想判断name如果不是bytes类似，就转换为bytes
        #if str is not bytes:         
        if type(name) != bytes:
            # Encoding needs to match generated source code output.
            # name = name.encode("latin1")
            try:
                name = name.encode("latin1")
            except:
                name = name.encode("utf8")