Allow for xlsx files to contain workbook2.xml instead of the expected…

[gowdy]

… workbook.xml. Add test for this.

[coveralls]

Coverage increased (+0.02%) to 57.949% when pulling 2b3342c on gowdy:master into 6eaa7c6 on python-excel:master.

[coveralls]

Coverage increased (+0.02%) to 57.949% when pulling 2b3342c on gowdy:master into 6eaa7c6 on python-excel:master.

[coveralls]

Coverage increased (+0.02%) to 57.949% when pulling 2b3342c on gowdy:master into 6eaa7c6 on python-excel:master.

[coveralls]

Coverage increased (+0.02%) to 57.949% when pulling 2b3342c on gowdy:master into 6eaa7c6 on python-excel:master.

[gowdy]

This is to fix Issue #241

[jdufresne]

I'm unfamiliar with the details of the issue. Under what conditions would workbook2.xml be used instead of workbook.xml? Is there any documentation to support this? I noticed openpyxl fails to open this document with KeyError: "There is no item named 'xl/workbook.xml' in the archive" as does LibreOffice.

[gowdy]

Hi Jon, I don't know which circumstance causes this, but it occurs for a file I downloaded from a data source. regards, Stephen.

…

On Sun, 8 Apr 2018, 17:54 Jon Dufresne, ***@***.***> wrote: I'm unfamiliar with the details of the issue. Under what conditions would workbook2.xml be used instead of workbook.xml? Is there any documentation to support this? I noticed openpyxl <https://openpyxl.readthedocs.io/en/stable/> fails to open this document with KeyError: "There is no item named 'xl/workbook.xml' in the archive" as does LibreOffice. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#242 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABsVf4V34WClry-ACZ-2QUb7DwomoR_Kks5tmkDcgaJpZM4RtPty> .

[jdufresne]

As both LibreOffice and openpyxl do not support workbook2.xml, I'm hesitant to merge this. If you can provide documentation that this is indeed part of the xlsx, I'll quickly change my mind and reopen the issue. At the moment, my first impression is this name may be nonstandard.

If so, might I suggest filing a bug with the data source explaining that they are delivering a nonstandard .xlsx file.

As a last resort, you could unzip and post process the file before passing it to xlrd.

Thanks for the PR and report.