Trying to use Excel::Reader::XLSX

[Ian Daysh]

Hi,

Probably not the place to ask this, but here goes ayway.

I know this package has not had much development, and I'm trying to get my head around the internals and see if I can use it in my application. I might be stretching your memory as it seems the last checkin was over a year ago!

Is it possible to set a cell object to a particular row/column as in Spreadsheet::ParseExcel?

If not, I can see my only option is to create an array of $row->values() and use $cells[row][col] instead. That would be currently up to 120x5 cells.

I see that if a next_cell() call hits an empty cell (the first cells in the first few rows are empty in my spreadsheet) then it doesn't carry on to the next cell in that row.

I understand you have other priorities, and I would like to contribute to this (beyond help with documentation for Spreadsheet::WriteExcel that I have done in the past). If you can point me to a place to start understanding the code so I can contribute I'll see what I can do.

Thanks

Ian.

[jmcnamara]

On Wednesday, 21 August 2013 18:17:56 UTC+1, Ian Daysh wrote:

Hi,

Probably not the place to ask this, but here goes ayway.

I know this package has not had much development, and I'm trying to get my head around the internals and see if I can use it in my application. I might be stretching your memory as it seems the last checkin was over a year ago!

Hi,

One of my goals when I started Excel::Reader::XLSX was to have a fast and memory efficient solution and avoid the flaws in Spreadsheet::ParseExcel. As such the basic design was to read a row at a time and to only store that row in memory. I figured that once the basic framework was in place I could create a ParseExcel style wrapper around it (but only as an option, the main interface was going to be different).

My first pass was a "sparse" reader when next_row() got the next row that had data (and skipped any non-existent rows in between). And the same for cells within the row.

While this was useful as a back-end parser I realised after it was finished that this wasn't a good interface to expose to the end-user and started writing a "sequential" reader where next_row() returned the next sequential row whether it had data or not (up until the last row). I got some of the way through that when one or more real life activities intervened to soak up my free time and I had to drop it.

So that is where it is at the moment. A deconstructed engine in pieces on the floor waiting to be rebuilt.

However, I've lost interest. I have about 10 repos on Github (as well as some forks) of which  2 or 3 of them are reasonably active and take up any free time I have.

All I can suggest is to take any useful parts and start your own module. The efficient use of XML::LibXML took a while to figure out and might be helpful. I'd suggest using a similar test suite but to avoid the JSON step (that was to make the tests shareable if I wanted to rewrite it in another language but they were a pain to generate and maintain and I as going to revert to something simpler).

Apart from that the test suite is probably the best place to start to understand what is going on.

Regards,

John

[Ian Daysh]

Hi John,

Thanks for the info. I can understand the losing interest part, and I'm not going to rant on about how you can leave something in this kind of state! ;-)

I've managed to get my application working as per my array of arrays idea, and don't expect too much of a problem as the datasets are quite small.

I might look at taking what you've done as a starting point and enhancing from there, but then again as I have a working system it's not likely in the short term :-D. If things start getting slow, then enhancement may be necessary.

Thanks for all your work on all these modules.

Regards

Ian.

--
You received this message because you are subscribed to the Google Groups "Spreadsheet::WriteExcel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to spreadsheet-write...@googlegroups.com.
To post to this group, send email to spreadsheet...@googlegroups.com.
Visit this group at http://groups.google.com/group/spreadsheet-writeexcel.
For more options, visit https://groups.google.com/groups/opt_out.