User talk:Bibcode Bot/Archives/2017
From Wikipedia, the free encyclopedia
 | This is an archive of past discussions with User:Bibcode Bot. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
doi number issue
After some CS1 maint, I did a self revert [here] because I noticed that some citation lead names weren’t in the actual papers provided by the citation doi numbers. Further history checking suggests that a problem may have occurred [here] I’m not sure how the bot captures the doi numbers but I think they are wrong. I suspect that this [1] is the correct Kushwahal ref, likewise correct cite for Pasquini[2] I think there may be others, hence it was easiest to self revert for now.
References
- Kushwaha, Pallavi; Lakhani, Archana; Rawat, R.; Chaddah, P. (2009). "Low-temperature study of field-induced antiferromagnetic-ferromagnetic transition in Pd-doped Fe-Rh". Physical Review B. 80 (17). doi:10.1103/PhysRevB.80.174413. ISSN 1098-0121.
- Pasquini, G.; Daroca, D. Pérez; Chiliotte, C.; Lozano, G. S.; Bekeris, V. (2008). "Ordered, Disordered, and Coexistent Stable Vortex Lattices inNbSe2Single Crystals". Physical Review Letters. 100 (24). doi:10.1103/PhysRevLett.100.247003. ISSN 0031-9007.
Edit broke math formulas
https://en.wikipedia.org/w/index.php?title=Brownian_motion&diff=665168290&oldid=664752787 Glrx (talk) 23:43, 6 June 2015 (UTC)
- Ugh, that stupid bug again. Though I bypassed all of those instances, but I guess not. I'd have eventually found the mistake, but good that you caught it before I got that. Headbomb {talk / contribs / physics / books} 03:14, 9 June 2015 (UTC)
- I assume this is the same bug: https://en.wikipedia.org/w/index.php?title=Cosmic_age_problem&type=revision&diff=671908444&oldid=670707271 (bottom part of the edit). (I've already manually reversed that.) --DanielPharos (talk) 09:19, 18 July 2015 (UTC)
- This seems to be the place to report this flawed edit at the article on Earth, where the bot changed two instances each of
\fracinto�rac, and of\rightintoightwith newlines prepended. Dhtwiki (talk) 23:23, 7 September 2017 (UTC)
Annoying disruption of template formatting
There is an aspect of Bibcode Bot that is quite annoying: the way it simply inserts the bibcode before the template's closing "}}". When the template is strung out horizontally this isn't a problem. But when a template is formatted vertically (i.e., on multiple lines) the placement of the closing braces at the beginning of a line is important for ready recognition of the end of the template. When the bibcode (or any other data) gets shoved in front of the braces it's harder to read, and confusing. So I would like to request that the bot check for newline (CR/LF) characters preceding the "}}", and preserve them in that position. If the bot could also check whether the vertical bars are preceded with a newline and possibly white space, and add its vertical bar in a conformable manner, that would be so much the better. ~ J. Johnson (JJ) (talk) 18:23, 25 February 2016 (UTC)
- I agree it would be an improvement, I've tried many times to implement that, but all my attemps have caused more headaches than it solved. It's something I've been meaning to look into again, but my time is fairly limited at the moment. Should free up in May or so however. Headbomb {talk / contribs / physics / books} 03:27, 4 March 2016 (UTC)
- I'd also like to ask for an improved formatting (example post-fix format). Maybe coding would become simpler if only a limited number of formats had to be recognized (my example uses one of the most used standard format for segregated refs). I also think that it is not a good idea to (manually) remove redundant url/arxiv params. Having a link on the citation's title is generally beneficial. Thx, Rfassbind – talk 00:38, 26 April 2016 (UTC)
- Links on the citation titles should only be used when the source is freely accessible. Those links I removed were paywalled and redundant with the arxiv/doi/etc. Arxiv preprints also tend to differ from published versions, so cannot be used instead of the actual publication, so arxiv links for a published article are inappropriate. Headbomb {talk / contribs / physics / books} 01:41, 26 April 2016 (UTC)
- My regex work-around for this is to completely grab the last parameter and all surrounding whitespace using
(\s*\|\s*)([\w_]+)(\s*=\s*)([^\}\|]*?)\|?(\s*\}\})and replace it with$1$2$3$4$1bibcode${3}BibcodeString$5, where$1,$3, and$5store the last parameter's whitespace around|,=, and}}, respectively, and duplicate$1&$3around a newly-place|bibcode=parameter. Curly brackets surround${3}so the engine doesn't accidentally use the first character in BibcodeString if it's a digit (which it probably will be). Unfortunately, all I know about python is that it can be an unintuitive pain in the ass and idk how well it plays with regex, but hopefully this helps. ~ Tom.Reding (talk ⋅dgaf) 03:13, 26 April 2016 (UTC)
- My regex work-around for this is to completely grab the last parameter and all surrounding whitespace using
Destroyed formula
Bibcode Bot destroyed a display formula in TeX at General Relativity. See here. Apparently it deleted "\n" and "\t" where they appeared in the formula. "\n" was replaced by a new-line; and "\t" by a tab or some blanks. JRSpriggs (talk) 20:51, 8 September 2017 (UTC)
Invalid arXiv identifier
This edit added an identifier I can't find in arXiv. Choess (talk) 02:43, 9 March 2016 (UTC)
- It did the same thing here. The common factor seems to be that they both have ASCL ID numbers, which the bot seems to be misidentifying as arXiv IDs. Modest Genius talk 09:57, 5 May 2016 (UTC)
- Once again, in this edit to the same article, List of cosmological computation software. Exactly the same error - it misidentified the ASCL for the article as an arXiv. Dairy{talk} 02:52, 11 September 2017 (UTC)
- This edit too Dairy{talk} 02:56, 11 September 2017 (UTC)