Referencing in flextables

[heor-andreas]

Hi David,

I hope you are well.

We are working with flextable and officedown to generate Word reports of academic writing. As part of the writing we use references to academic papers. Some of these references will appear in tables. Using the markdown language these are entered as [@author].
When outputting such a table through kable, the references get interpreted correctly, however, when passing the same table with references through flextable these references do not get interpreted. See example below.
See picture below.

Is there a workaround to this or a quick way to get this feature implemented?

I have also tried to apply Word table styles to the markdown table (which doesn't change the formatting) and the printout of the data.frame (which adjusts the style, but doesn't interpret the references).

Looking forward to hearing your thoughts.

Thanks
Andreas

[davidgohel]

Hello,

Could you add a reproducible example?

Thanks

[davidgohel]

OK, I think the following code + 17-11-17-Zotero.bib.zip reproduces your issue.

This is a kable feature, I don't know yet if this can be implemented.

---
output: officedown::rdocx_document
bibliography: 17-11-17-Zotero.bib
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, fig.cap = TRUE)
library(officedown)
library(officer)
library(readxl)
library(knitr)
library(flextable)
mydat <- structure(list(Bibcode = c("[@brookmeyer_forecasting_2007]", 
"[@hellmann_new_1935]", "[@becke_densityfunctional_1993]", "[@stephens_ab-initio_1994]"
), Author = c("Brookmeyer", "Hellman", "Becke", "Stephens"), 
    Rate = c(100, 50, 100, 20), url = c("http://www.alzheimersanddementia.com/article/S1552-5260(07)00475-X/abstract", 
    "http://scitation.aip.org/content/aip/journal/jcp/3/1/10.1063/1.1749559", 
    "http://scitation.aip.org/content/aip/journal/jcp/98/7/10.1063/1.464913", 
    NA), doi = c("10.1016/j.jalz.2007.04.381", "10.1063/1.1749559", 
    "10.1063/1.464913", "10.1021/j100096a001"), number = c(1, 
    2, 3, 4)), row.names = c(NA, -4L), class = c("tbl_df", "tbl", 
"data.frame"))
```

This is a kable table.

```{r echo=FALSE}
kable(mydat)
```


This is a flextable

```{r echo=FALSE}
flextable(mydat)
```

17-11-17-Zotero.bib.zip

[heor-andreas]

Thanks David,

I've done some investigating and am wondering if this somehow has to do with the way the flextable object is getting passed through to pandoc. I think pandoc is where the reference conversion happens.

In the code above I have just added this html code as plain text

<table>
<thead>
  <tr>
    <th>reference here:</th>
    <th>[@hellmann_new_1935]</th>
  </tr>
</thead>
</table>

when I knit this using the knit button in RStudio I get the kable tables done right, the flextable not interpreting the references and the bit I added in the html code at the end is also correctly converted to a reference. So it seems that things within the html arguments get parsed through the bibilography building part. In other words the fact that flextable generates html code does not seem to be the problem.

I then tried kniting to .md file. I can then see that the references in both the flextable html and the raw html at the end of the code are the same thing.

when I now knit this .md document to .html (clicking the preview button in RStudio) I get the references correctly done in flextable as well.

makes me wonder if the flextable code bypasses some of the pandoc processes?

[davidgohel]

flextable doesn't require pandoc, it does not use pandoc to produce the code. To enable this feature, it has to be coded somewhere (maybe with ftExtra from @atusy).

[atusy]

@heor-andreas

https://github.com/atusy/ftExtra does support Pandoc's markdown partly, but not citations.
Although I'm not sure if it is even possible, feel free to open an issue as a feature request.

Thanks for mention @davidgohel

[atusy]

@heor-andreas

This feature is experimentally implemented by atusy/ftExtra#27

See example at https://ftextra.atusy.net/articles/format_columns.html#citations

[github-actions]

This old thread has been automatically locked. If you think you have found something related to this, please open a new issue and link to this old issue if necessary.