In most cases, it is simpler to (1) read a PDF file directly in WORD, or (2) use save as function in Acrobat Reader to save as a DOC file.
Pandoc can be installed in anaconda by
conda install pandoc
or in brew.
brew install pandoc
I strongly recommand to use brew version. For brew, switch to the brew environment first.
swt_brew # custom command in .bash_profile
Then go to the folder where you have .tex files to convert. Run the following command:
pandoc --citeproc -f latex -s FeSH-JGR.tex -o FeSH-JGR.docx --bibliography=FeSH.bib
Change the file names appropriately.
- The command works only with
brewversion ofpandocfor me. - Note that the order betweeen
pandoc-crossrefandciteprocshould not change. - Run multiples of time (unconfirmed).
Pandocknows how to deal withbibunitsand\input{}.- Note that
line numberreferencing and citing do not work withpandoc. - Do not place both
-F pandoc-crossrefand--citeprocat the same time, it will repeat the figure and table caption titles. - Note that renumbering for figures and tables (for example S1 for supplementary contents) are all ignored. So be careful if this is important.
pandoc --citeproc -f latex -s Nature-main\ text.tex --csl=../nature.csl -o main.docx --bibliography=B30.bib
*.csl can be donwloaded from: https://github.com/citation-style-language/styles. A range of journal format is available there.
pandoc -t latex -f docx in.docx -o out.tex
pandoc -F pandoc-crossref --citeproc --mathjax -f latex -t html -s FeSH-JGR.tex -o FeSH-JGR.html --bibliography=FeSH.bib
For better outcome to html:
- convert all pdf figures to png figures
\multicolumnis not compatible withpandoc- Figure labelings are properly done, but not for table labeling. I try everything but so far the line above is the best. If you remove any of
-F pandoc-crossrefand--citeprocfrom the line above, it will mess up the citations.