!HEfOZuCWbFqQODrTRb:matrix.org

Xapian

32 Members
The Xapian search engine library https://xapian.org/ | GSoC https://trac.xapian.org/wiki/GSoCProjectIdeas | FAQ https://trac.xapian.org/wiki/FAQ | doc https://getting-started-with-xapian.readthedocs.io/ | If you have a question, just ask it and WAIT for a reply (there's often a delay due to timezones)16 Servers

Load older messages


SenderMessageTime
3 Jul 2020
@dipanshu:matrix.orgdipanshu
In reply to @freenode_jaylett:matrix.org
dipanshu what do you mean duplicated in include/weight.h? You mean for each class? Maybe give me line numbers so I can check what you mean.
I am talking of lines from 523-564 that describe the constructor when only normalizations string is passed . Then lines 567-609 describe the constructor when normalisations string , parameter slope ans delta are passed . Since now we call the second constructor from the first one, I think lines 523-564 should be avoided. Otherwise, when I add documentation for the 2 new constructors that use enums, there also this duplication will exist which I think won't look very neat.
10:42:13
@dipanshu:matrix.orgdipanshuhttps://github.com/dipanshu124/xapian/blob/enum_parameters/xapian-core/include/xapian/weight.h10:42:17
@freenode_jaylett:matrix.orgjaylett (Incidentally you can click on the line number and it'll add a fragment ID to the URL that will highlight that line) 10:43:59
@freenode_jaylett:matrix.orgjaylett You should look at the doxygen documentation and see if there's an easy way of referencing another constructor / overloaded method from a docstring. Then you could say "normalisation string as described <here>" or similar. 10:45:13
@freenode_bremner:matrix.orgbremner hah. you used "doxygen" and "easy" in the same sentence :P. 10:46:13
@freenode_bremner:matrix.orgbremnerI spent several hours yesterday documenting one function, my snark is earned ;)10:46:32
@freenode_jaylett:matrix.orgjaylettWell, easy for doxygen ;)10:46:33
@freenode_jaylett:matrix.orgjaylett If you don't have to rebuild it, it's easy 10:46:49
@freenode_jaylett:matrix.orgjaylett dipanshu note that for enums, the API docs will have a link to the enum definition which is where most of the documentation will be — so there won't actually be much duplication 10:47:19
@freenode_jaylett:matrix.orgjaylett This isn't true of the normalisation string however, so it's worth tidying up if it's possible while still being useful 10:47:33
@dipanshu:matrix.orgdipanshuhttps://github.com/dipanshu124/xapian/blob/193ededc18510c3a76579d1085dec03f150add0a/xapian-core/include/xapian/weight.h#L476 This extends more that 80 and hence xapian-check-patch is failing .But it is an algebraic expression and forcing it to 2 lines will make it look awkward . What should I do ?12:03:43
@freenode_jaylett:matrix.orgjaylett I wouldn't give the formula at all. Instead, reference the paper (particularly if there's an open online URL) and describe what it actually does. The formula doesn't help me understand when I'd use it, and "pivoted" is pretty terse. 12:08:02
@freenode_jaylett:matrix.orgjaylettIf you want to include the formula (and there are arguments in favour), then either split outside braces at operators, or split like a C++ expression12:08:32
@freenode_jaylett:matrix.orgjaylettThe former is a bit more maths like12:08:38
@dipanshu:matrix.orgdipanshuBut all the formulas are already present in documentation (that used the normalisation string) . So may be for enums we can avoid the formulas as you suggest.12:10:11
@freenode_jaylett:matrix.orgjaylett It's much better to have all of them explained in one place. The existing documentation however has never been in a release series, so we can rewrite it any way we think makes it better. (I mean, that's generally true of release series as well, for documentation.) 12:22:22
@freenode_dantti:matrix.orgdantti joined the room.14:59:05
@dipanshu:matrix.orgdipanshu Running test: closedb5....glass\dbw\docdata.glass - The process cannot access the file because it is being used by another process. .glass\dbw\position.glass - The process cannot access the file because it is being used by another process. .glass\dbw\postlist.glass - The process cannot access the file because it is being used by another process. .glass\dbw\spelling.glass - The process cannot access the file because it is being used by another process. .glass\dbw\synonym.glass - The process cannot access the file because it is being used by another process. .glass\dbw\termlist.glass - The process cannot access the file because it is being used by another process. FAIL: DatabaseCreateError: Can't create new database at '.glass/dbw': a database already exists and I was told not to overwrite it 16:55:07
@dipanshu:matrix.orgdipanshuAppveyor build failed on this test16:55:32
@freenode_sec^nd:matrix.orgsec^nd left the room.17:12:52
@freenode_jaylett:matrix.orgjaylett That sounds like a transient problem with Appveyor, since you haven't touched that test or anything that it relies on that I can think of. 17:18:28
@dipanshu:matrix.orgdipanshuSo re running Appveyor should fix it ?17:20:14
@freenode_jaylett:matrix.orgjaylett Should do, but you need to do another pass over the enum docstring. For a start it isn't a docstring (doesn't start /**), but also you can use more of doxygen to make a nice output. Look at https://github.com/xapian/xapian/blob/master/xapian-core/include/xapian/query.h#L75 17:22:55
@freenode_jaylett:matrix.orgjaylettWhich results in https://xapian.org/docs/apidoc/html/classXapian_1_1Query.html#a7e7b6b8ad0c915c2364578dfaaf6100b17:23:00
@freenode_jaylett:matrix.orgjaylett Also, don't get rid of @param docs, for instance at https://github.com/xapian/xapian/pull/302/files#diff-8a2d77838da5b7da44ddb21e3d97acb1L493 17:24:16
@freenode_sec^nd:matrix.orgsec^nd joined the room.18:03:07
@exter-dg:matrix.orgexter-dgI have merged my branch with the remote master, and have added testcases and files for libarchive in omindexcheck19:12:17
@exter-dg:matrix.orgexter-dghttps://github.com/xapian/xapian/pull/30019:12:19
@olly.nz:matrix.orgolly Appveyor fails like that sometimes - I've not managed to work out why yet though 20:01:50
@olly.nz:matrix.orgolly Don't worry about such failures (or feel free to debug!) 20:03:00

There are no newer messages yet.


Back to Room List