Comments on: E-mail push update: Corrected CSV headers, fixed daily dumps

By: Anonymous

Anonymous — Fri, 20 Nov 2009 06:55:49 +0000

Date: Wed, 18 Nov 2009 01:35:01 +0000 (UTC)
Subject: [evec-upload] Upload

price,volRemaining,typeID,range,orderID,volEntered,minVolume,bid,issued,duration,stationID,regionID,solarSystemID,jumps,source
…
45.4099,48765.0,37,-1,1345142586,100000,1,True,47.99,45506.0,37,32767,1363864325,45506,1,False,2009-11-18 00:45:01.000,90,60012001,10000037,30003016,5,
…

note how in the middle of that row, it suddenly starts a new one.

this just an artefact of adding the trailing comma for no-source rows?
because parts of that mail do have a trailing comma, other parts dont.

By: Yann

Yann — Wed, 18 Nov 2009 23:00:59 +0000

The lack of a trailing , was due to the use of Contribtastic/libevecache sourced uploads (which doesn’t emit the trailing ,).

The header had a small snafu, but does correctly describe the fields, with the only caveat is the optional source field (at this time).

By: Msquare

Msquare — Wed, 18 Nov 2009 22:26:13 +0000

… Wow, you’re changing the layout tonight, too? Now suddenly the lines do not contain “,evetrace” (or something like that, those mails are gone already) at the end of each line. The original eve dump contains a trailing “,” on each line, now that is gone.

What is your intention on how much you will “massage” the original EVE dump files?

Will the header line be a true CSV header (gasp, even more parsing for me) or is it still just a line of constant content to (mostly) disregard?

By: Msquare

Msquare — Wed, 18 Nov 2009 21:12:43 +0000

Surprise !
Took me a while to notice things had changed. I only have the mail feed open for active playing periods, so when the mailfeed didnt work I assumed other things had broken, not the format. Still a good improvement, I’ve played a little with themail-identifier-subject thingy to get some indication of the original contributor (mainly myself )

By: N.N.

N.N. — Tue, 17 Nov 2009 10:47:42 +0000

please make sure to at least include the source field in all rows, even when it is empty.
number of fields in row not matching up with number of fields in header == unhappy CSV parser.

and thanks for all the work you are putting into this!