| 00:35.39 | *** join/#utah gnord (~gog@unaffiliated/snowghost) |
| 00:50.14 | *** join/#utah gnord (~gog@unaffiliated/snowghost) |
| 05:48.22 | *** join/#utah fadein (~Erik@c-24-10-167-68.hsd1.ut.comcast.net) |
| 10:22.59 | *** join/#utah JPau1 (~JPaul@2604:a880:800:10::16e:c001) |
| 10:24.33 | *** join/#utah yukaia_ (sid110254@gateway/web/irccloud.com/x-dexykwsqxlxtzfdz) |
| 10:36.37 | *** join/#utah spangborn (~spangborn@216.51.232.249) |
| 15:31.48 | *** join/#utah aarobc (~aarobc@216.51.32.182) |
| 16:18.45 | *** join/#utah tiwula (~lane@70.89.246.241-Busname-UT.hfc.comcastbusiness.net) |
| 17:25.43 | jima | spangborn: ted? |
| 17:53.42 | jlp_znc | Perl: $s = "A"; $s++; => $s == "B" (and "Z" => "AA", etc.) |
| 17:53.47 | jlp_znc | Python: ?? |
| 17:57.28 | jlp_znc | so far, this is the best I have come up with: https://gist.github.com/jlp78/f306afc919dc06c8ce156475fc9320bf |
| 18:16.59 | jima | spangborn: thanks for enlightening me that i'm a terrible utahn in addition to the usual terrible jew. |
| 18:17.38 | jlp_znc | ? |
| 18:18.04 | jima | jlp_znc: https://xkcd.com/1254/ |
| 18:18.42 | jima | (the joke there being that i regularly respond to spangborn's tweets here) |
| 18:18.44 | jlp_znc | hmmm |
| 18:18.51 | jlp_znc | I see |
| 18:38.55 | *** join/#utah SunSparc (~jonathan@45.56.57.85) |
| 21:14.26 | TodPunk | jlp_znc: what about the python isn't working for you? |
| 21:15.31 | TodPunk | not that in a non-english ascii alphabet, everything is going to break in both worlds |
| 21:16.04 | TodPunk | (seriously, do not increment characters and still treat them as characters) |
| 22:25.36 | jlp_znc | TodPunk: it's not that it doesn't work for me... it's just that it seems clunky to do something that I would be a common task. |
| 22:26.03 | jlp_znc | think about the columnns of a spreadsheet. say you want to iterate over the, from A to ZZZ. |
| 22:26.21 | jlp_znc | (in fact, my application is using a python library to access a google sheets document) |
| 22:26.48 | jlp_znc | s/iterate over the/iterate over them/ |
| 22:27.26 | jlp_znc | in perl, thanks to the magical properaties of the autoincrement operator, it's easy to do $col++ and get something useful. |
| 22:27.42 | jlp_znc | in python, no so much. |
| 22:28.24 | romanovic | for a-z, you can do a quick chr(ord('a') + 1) |
| 22:28.47 | jlp_znc | which is fine until you hit Z and want to go to AA next |
| 22:28.58 | fungus | from string import uppercase as L |
| 22:29.02 | fungus | L[(l.find(c) + 1) % len(L)] |
| 22:30.00 | fungus | That won't get you double digits yet, but that's not that hard. |
| 22:30.16 | fungus | However, I get your point it's nowhere near as easy as $c++. |
| 22:31.36 | jlp_znc | I guess I'm bemoaning the fact that python doesn't have an autoincrement operator that is a cool as perl's... or just that python doesn't have an autoincrement operator. |
| 22:33.52 | romanovic | i think python correctly doesn't offer spreadsheet style indexing/incrementing as a special type or use case. PHP did this type of autoincrement as well. there's a reason we've moved away from it |
| 22:34.55 | romanovic | too many wrong assumptions about the programmer's intent, too many edge cases |
| 22:35.44 | TodPunk | jlp_znc: in all honesty, I'd prefer a lib that converted "AA-ZZ" to an index so I could just use the index to iterate through and the characters when I need to access a specific one by the human name |
| 22:35.59 | romanovic | TodPunk++ |
| 22:36.19 | TodPunk | the character columns are purely for UI |
| 22:36.28 | romanovic | wonders... (TodPunk++) == TodPunj ? |
| 22:36.45 | romanovic | s/TodPunj/TodPunl/ |
| 22:36.53 | TodPunk | TodPunk-- I guess =cP |
| 22:37.24 | jlp_znc | I'd be fine with that, too. the application where this was useful in Perl (back in the dark ages) has kind of gone away. think about splitting a file into chunks so you can transfer them over a shakey link more easily. you want the files to be named something like name.key, where key is nicely increasing alphanumerically so it's easy to cat name.* > name on the other side. |
| 22:38.31 | TodPunk | yeah, today that sounds stupid, even over a radio link. Compress JSON and use TCP or something |
| 22:38.35 | jlp_znc | now-a-days the links are more stable (no modems hanging up in the middle of your big file transfer) and the protocols handle resuming a partial transfer better. |
| 22:38.42 | TodPunk | I know back in the day we weren't sophisticated, I was there |
| 22:38.45 | TodPunk | just saying |
| 22:39.18 | TodPunk | (I also have to sometimes do some embedded work to implement low-bandwidth sketchy wireless devices updating data to wherever) |
| 22:39.29 | TodPunk | full disclosure: I also hate perl |
| 22:39.49 | TodPunk | granted, that's only because of the options I have today |
| 22:40.04 | TodPunk | back then perl was clearly better for a lot of things |
| 22:40.05 | jlp_znc | anyway, in my particular application, I'm updating a spreadsheet with some status info (don't get me started on how a spreadsheet is the wrong tool for the job), and I have to put some info in a cell and the one next to it. I want my script to be more robust in the face of the particular cell moving around as "someone" adds/removes columns. |
| 22:40.31 | TodPunk | but you're using google docs |
| 22:40.49 | jlp_znc | so I have a list of cells, and I want to put something in those cells and the cells immediately to their right |
| 22:41.02 | TodPunk | doesn't their API have an "append to row" where you pass in a string and it handles whatever? |
| 22:41.24 | jlp_znc | not the library I'm using, at any rate (gspread) |
| 22:41.35 | jlp_znc | I can update a cell given it's coordinate |
| 22:42.04 | jlp_znc | so I have a list of cells starting at a base column and row and I want to put something in them and their immediate neighbors to the right |
| 22:42.25 | jlp_znc | and I want to be robust in the case where the spot I'm putting this happens to cross the Z->AA boundary |
| 22:42.37 | jlp_znc | (it doesn't at the moment, but you never know) |
| 22:43.03 | jlp_znc | the code I found does the job, but just feels clunky. |
| 22:45.05 | TodPunk | it should be a little bit, you're doing string manipulation that absolutely should not translate to byte operations, you're just getting lucky in a niche (and arguably insane) use case |
| 22:45.46 | TodPunk | to be fair, the perl code wouldn't do that Z->AA either, so you're special casing either way |
| 22:46.03 | jlp_znc | actually, the perl code *would* do the Z -> AA transition. |
| 22:46.19 | TodPunk | you have to test for that case, right? |
| 22:46.31 | jlp_znc | nope |
| 22:46.42 | jlp_znc | part of the magical properties of autoincrement in perl |
| 22:46.55 | TodPunk | because it should be going to '[' |
| 22:46.56 | jlp_znc | perl -e '$a = 'Z'; $a++; print $a' prints AA |
| 22:47.08 | jlp_znc | it's not doing a literal ordinal increment |
| 22:47.39 | jlp_znc | it handles carry, too, so AZ -> BA |
| 22:48.03 | TodPunk | That... is stupid... I can't think of a reason why a language would want to do that UNLESS it was purely for spreadsheets |
| 22:48.31 | jlp_znc | and that case I described above where you're creating a bunch of files that need to be catted together in order later |
| 22:48.41 | TodPunk | Like, sincerely, I want to know what idiot was doing this and thought that was useful as the core way incrementing /characters/ should be handled |
| 22:48.57 | TodPunk | there's got to be some justification I wouldn't light him on fire for |
| 22:49.06 | jlp_znc | guess you'd have to ask Larry Wall |
| 22:51.15 | jlp_znc | he's on IRC... TimToady |
| 23:13.55 | *** join/#utah jholtom (~jholtom@pool-70-106-211-121.clppva.fios.verizon.net) |
| 23:42.27 | jlp_znc | so, did you ask him? :-) |
| 23:45.04 | TodPunk | no, I'm going the typical internet route of wishing for things without taking action. I might even tweet if I can muster some outrage for points |