Intra-query Special Characters

[jjc54]

Hi,

I have a query for a name that contains an apostrophe. I am trying to escape the apostrophe by using the backslash but doing so seems to halt the query and it never fetches any information from PubMed. What is the best way to handle this?

Example:

Query <- 'Christopher O\'Grady[AU] OR Christopher P O\'Grady[AU] OR CP O\'Grady[AU]'

out.Query <- batch_pubmed_download(pubmed_query_string = Query, 
                                   format = "xml", 
                                   batch_size = 1000,
                                   encoding = "UTF8")

[dami82]

This code looks appropriate to me. I tried to run it as is on my Ubuntu and it executed correctly in 4.39 seconds and generated a file of 6.8 Mb. I see no issues with this code and couldn't reproduce any error. Sorry.
For reference, I used easyPubMed_2.22; R version 4.2.1 (2022-06-23). Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 20.04.4 LTS. Thanks.

[jjc54]

Thanks for testing. I just updated R/easyPubMed. It seems the issue is in the next two lines of my code:

Query <- 'Christopher O\'Grady[AU] OR Christopher P O\'Grady[AU] OR CP O\'Grady[AU]'

out.Query <- batch_pubmed_download(pubmed_query_string = Query,
                                   format = "xml",
                                   batch_size = 1000,
                                   encoding = "UTF8")

QueryList <- articles_to_list(pubmed_data = out.Query)
Columns <- lapply(QueryList, article_to_df, autofill = TRUE)

When I execute the lapply, RStudio hangs and must be force quit.

[dami82]

Thanks for reporting this issue. In the end, the problem has nothing to do with the apostrophe, but with the number of authors listed for certain records.

I looked into this. Your query is returning some records with very long author lists. We are talking about 2,000 or more authors for a single record (e.g., PMID 28316497 has 2,872 authors listed). This is not very common (at least, I don't see many articles like this being published in journals in my area of interest), and this exception is breaking the code execution.
There are definitely ways to handle this in an efficient way. I am just wondering whether it would make more sense (for easyPubMed) to still parse all authors from these records, or just skip the records... In your specific case, do you really want to get author information from all authors from each PubMed record? If your answer is NO, you can try:

QueryList <- easyPubMed::articles_to_list(out.Query)
Columns1 <- lapply(QueryList, article_to_df, 
                   autofill = FALSE, getAuthors = FALSE) 

This code should complete in less than 5 seconds. In the meanwhile, I'll keep thinking about what could be the most reasonable way to fix this issue in the next release of the library. Thanks.

[jjc54]

Thank you for looking into this! I think we stumbled upon a publication from the Large Hadron Collider/CERN, which is unfortunate because my investigator is an orthopedic surgeon. There is an F. O'Grady listed as an author that is contributing to the query result.

Your fix worked perfectly as a workaround so I appreciate that. I'll just use that for now, thanks again!