将变量传递给函数时的Python溢出错误

Question

I have a query that gets an arbitary number of rows from the database, but usually no more than two.

I call the query with

ret = self.workaround.delete_last_version(pdf_file, self.cust_id, self.part_id)

If i hardcode the values and feed them into the query, it works fine, but when I run the whole program normally, and input the data in through a gui, I get when the query function is called

OverflowError: long too big to convert

However, before the call is made, I convert part_id and cust_id to strings with str()

Here are the values that I've been testing with.

part_id = "168440901713431956015724879141"
cust_id = "15424322074155018763160235136213"
pdf_file = "1971-48.pdf"

The cust_id is 32 chars long, and part_id is 30

And here is the query.

query = """
    SELECT Id, Name, DocDateStamp, Dir_Id
    FROM SdmDocumentList
    WHERE Owner_Id = ? AND Name = ?
    ORDER BY DocDateStamp desc
    """

    self.cursor.execute(query, (part_id, pdf_file))
    result = self.cursor.fetchall()

And the cursor is defined in __init__

def __init__(self):

    self.conn =pyodbc.connect('DRIVER=SQL Server;SERVER=mssql;PORT=CORRECTPORT;DATABASE=DATABASENAME;UID=USER;PWD=SUPERSECRETPASSWORD;TDS_Version=8.0;Trusted_Connection=no') #windows
    self.cursor = self.conn.cursor()

The expected output, which I get when I call the function from command line

[('70D8B606A5BD421F97467F2C8D1D8F04', '1971-48.pdf', datetime.datetime(2017, 5, 12, 9, 28, 58, 743333), 9), 
 ('F049665629814B83B63B0536F985090B', '1971-48.pdf', datetime.datetime(2017, 5, 12, 9, 28, 22, 86666), 9), 
 ('39EAB6B173B745E19BA8A0598AD8F015', '1971-48.pdf', datetime.datetime(2017, 5, 12, 9, 27, 58, 933333), 9), 
 ('6309915CEA504839A8D7340F9A4FD601', '1971-48.pdf', datetime.datetime(2017, 5, 12, 9, 12, 26, 36666), 9), 
 ('5D3E3AA218CA4FF59E8EC2DE2A6AB217', '1971-48.pdf', datetime.datetime(2017, 5, 12, 9, 11, 37, 760000), 9), 
 ('1C2AC1073E754A41A998DF99BD0F4F48', '1971-48.pdf', datetime.datetime(2017, 5, 12, 9, 10, 49, 986666), 9), 
 ('05EF8020EA354E669D1D930650FBCB02', '1971-48.pdf', datetime.datetime(2017, 5, 12, 8, 54, 59, 80000), 9), 
 ('834979EFB639466ABC73E76F88DF6750', '1971-48.pdf', datetime.datetime(2017, 5, 12, 8, 54, 5, 46666), 9), 
 ('F3EF2C75856E4926A52204EBA59072CB', '1971-48.pdf', datetime.datetime(2017, 5, 12, 8, 50, 38, 406666), 9),
 ('6F7FC2652E114162AF6B3C4C30818582', '1971-48.pdf', datetime.datetime(2017, 5, 12, 8, 37, 59, 610000), 9)]

I'm running Python2.7 on Windows7 64bit and 6gb of RAM, with mssql

I've Googled and looked at other issues with OverflowErrors and longs being too long to convert, but none of the solutions posted seemed to fix my issue, and all the issues I found on the internet had long being converted to another numeric type. At the line I get the error, I'm not even trying to convert anything, just pass the variables to a function.

Is there some memory limit to how much data can be sent to a function in Python, or am I missing something else entirely?

More than happy to provide more information if needed, but I think I've written down everything that would be needed for this.

Thanks in advance for any help

EDIT Query that fails after updating pyodbc:

The new query that's causing problems after updating pyodbc

 Dir_Id = 9                              # Manufactering Data folder (from SdmDirSructure)
        DocumentType_RecNo = 3                  # PDF
        OwnerType_RecNo = 2                     # parent_id for Product (from SdmDirSructure)
        FromTo = 14872094126171117726173141183125     # CNCDrill user ID
        DocumentSubject_RecNo = 109             # Xcheck data (from SdmDocumentSubjects)
        Owner_Id = self.part_id                 # It is the part id
        DocDateStamp = file_datetime            # datetime of the document
        RegisteredDate = file_datetime          # when the document was added to the doc manager
        LastReadDate = file_datetime            # date that the document last readed (set as the RegisteredDate)
        ArchiveLocation_RecNo = 0               # Not in use
        IsReceived = True                       # Not in use
        IsArchived = False                      # Not in use
        IsOwnedByUser = 0                       # No-one ownes the file

        query = """
        INSERT INTO SdmDocumentList (Id, Name, Dir_Id, DocumentType_RecNo, OwnerType_RecNo, FromTo, DocumentSubject_RecNo, Owner_Id, DocDateStamp, RegisteredDate, LastReadDate, ArchiveLocation_RecNo, IsReceived, IsArchived, IsOwnedByUser, IsClassified) 
        VALUES(?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?)
        """
        self.cursor.execute(query, (uid,pdf_file, Dir_Id, DocumentType_RecNo, OwnerType_RecNo, FromTo, DocumentSubject_RecNo,Owner_Id, DocDateStamp, RegisteredDate, LastReadDate, ArchiveLocation_RecNo, IsReceived, IsArchived, IsOwnedByUser, IsClassified))
        self.cursor.commit()

Answer 1

 FromTo = 14872094126171117726173141183125     # CNCDrill user ID

The cuplrit. Eventually I started printing every variable and its type. It turns out that where I'd hardcoded the 'FromTo' variable, which wasn't going to change, ever, for this script, I had forgotten to put " " around it, making it a long, which was crashing the query.

I'd realised that when I was getting the part and customer id's out of the database earlier in the script, it was returning them as Unicode. I did some research into Unicode, and tried alternatives, such as defining the encoding at the top of the script, but they didn't work. However, this led me down the correct path of checking that every variable was the type I was expecting.

Thanks for all the help on Friday, @Gord Thompson May