submitex 0.0.1

Tools and CLIs to make your tex-files submission-ready.

Python tools and CLIs to automatically convert your LaTeX-project into a structure in which it can be easily submitted to a journal.

For instance for manuscript.tex with bibtex-generated manuscript.bbl:

replacebib manuscript | replacefigs | collectfloats | resolveinputs | resolvepipes > newmanuscript.tex

Install

pip install submitex

Examples

Check out the cookbook section or further below.

Workflow for the arXiv

Modules/CLIs

Functionality is given by functions in the respective submitex.modulename, e.g. submitex.replacefigs.

The same modulename can be used to evoke the functionality from the command line.

module/CLI	description
replacefigs	Rename all figure files in the \includegraphics-environment to generic enumerated names and copy the respective files with the new names to top-level.
collectfloats	Remove all figure- and table-environments from their respective place in the document body and put them on their own page at the end of the document, first figures, then tables.
replacebib	Put the content of the bbl-file into the section where previously laid \bibliographystyle{...}\bibliography{filename}
resolveinputs	Replace \input{filename} with the content of filename
resolvepipes	Replace \input{\|command} with the output of the process command
removecomments	Remove comments (beginning with unescaped %) on non-whitespace lines, squash lines comments to a single line on white-space lines. Squash white-space lines to a single line.

Workflow for the arXiv

For the arxiv, it’s nice to copy all figures to top-level, resolve all inputs, replace the bibliography with the bbl file, and remove comments. Then in the end, you want a zip that you can upload. What follows is a Makefile that takes care of that.

Make sure to have diff-pdf installed! On MacOS:

brew install diff-pdf

Here is the Makefile:

# ============================
# Configurable parameters
# ============================

# Input directory (copied from Overleaf)
INPUT_DIR := /path/to/manuscript/directory

# Local working directories
PAPER_DIR := paper
ARXIV_DIR := arxiv
ARXIV_TEST := arxiv_test

# Manuscript base name (without .tex)
MAIN := main

# Tool to open a file (on Mac, just `open`)
OPEN := open

# ============================
# Targets
# ============================

default: init arxiv

init: clean copy latex

arxiv: processtex test zip

latex:
    cd $(PAPER_DIR); pdflatex $(MAIN); bibtex $(MAIN); pdflatex $(MAIN); pdflatex $(MAIN)

processtex:
    cd $(PAPER_DIR); replacebib $(MAIN) | resolveinputs | replacefigs | removecomments > $(MAIN)_for_arxiv.tex
    mkdir -p $(ARXIV_DIR)
    mkdir -p $(ARXIV_TEST)
    cp ./$(PAPER_DIR)/$(MAIN)_for_arxiv.tex ./$(ARXIV_DIR)
    cp ./$(PAPER_DIR)/Fig* ./$(ARXIV_DIR)
    cp ./$(PAPER_DIR)/$(MAIN)_for_arxiv.tex ./$(ARXIV_TEST)
    cp ./$(PAPER_DIR)/Fig* ./$(ARXIV_TEST)

copy:
    cp -r $(INPUT_DIR)/ ./$(PAPER_DIR)

clean:
    rm -rf ./$(ARXIV_DIR)
    rm -rf ./$(ARXIV_TEST)
    rm -rf ./$(PAPER_DIR)
    rm -f texput.log arxiv.zip diff.pdf

softclean:
    rm -rf ./$(ARXIV_TEST)
    rm -rf ./$(PAPER_DIR)
    rm -f texput.log diff.pdf

zip:
    zip -r arxiv.zip ./$(ARXIV_DIR)

test:
    cd $(ARXIV_TEST) &&\
            rm -f $(MAIN)_for_arxiv.pdf &&\
            pdflatex $(MAIN)_for_arxiv.tex &&\
            pdflatex $(MAIN)_for_arxiv.tex
    rm -f $(ARXIV_TEST)/$(MAIN)_for_arxiv
    diff-pdf $(ARXIV_TEST)/$(MAIN)_for_arxiv.pdf $(PAPER_DIR)/$(MAIN).pdf --output-diff=diff.pdf --skip-identical
    $(OPEN) diff.pdf

Python examples

replacefigs

import submitex.replacefigs as rf

tex = r"""
\begin{figure} \includegraphics[width=1in]{foo/bar} \includegraphics{result.png} \end{figure}

\begin{figure} \includegraphics[width=1in]{foo/bong.jpg} \end{figure}
"""

newtex, figpaths = rf.convert_and_get_figure_paths(tex, figure_prefix='figure_')

print(newtex)
for src, trg in figpaths:
    print(src, trg)

Output:

\begin{figure} \includegraphics[width=1in]{figure_01a} % original file: foo/bar
 \includegraphics{figure_01b.png} % original file: result.png
 \end{figure}

\begin{figure} \includegraphics[width=1in]{figure_02.jpg} % original file: foo/bong.jpg
 \end{figure}

foo/bar figure_01a
result.png figure_01b.png
foo/bong.jpg figure_02.jpg

collectfloats

import submitex.collectfloats as cf

tex = r"""\begin{document}
\begin{figure} \end{figure}
Test
\begin{figure} \end{figure} \begin{ table} \end{table }
This is another paragraph
\end{document}
"""

print(cf.convert(tex))

Output:

\begin{document}

Test

This is another paragraph
\afterpage{%
\begin{figure} \end{figure}
\clearpage}

\afterpage{%
\begin{figure} \end{figure}
\clearpage}

\afterpage{%
\begin{ table} \end{table }
\clearpage}

\end{document}

replacebib

import submitex.replacebib as rb

tex = r"""
\bibliographystyle{vancouver}
%\bibliographystyle{chicago}
%\bibliography{main.bib}
\bibliography {main.bib}
"""
bib = r"""
\begin{thebibliography}
\end{thebibliography}
"""
print(rb.convert(tex,bib))

Output:

%\bibliographystyle{chicago}
%\bibliography{main.bib}

\begin{thebibliography}
\end{thebibliography}

resolveinputs

File section1.tex:

\section{Section 1}
This is Section 1.

import submitex.resolveinputs as ri

tex = r"""
\input{ section01.tex}
%\input{ section01.tex}
"""

print(ri.convert(tex))

Output:

\section{Section 1}
This is Section 1.

%\input{ section01.tex}

resolvepipes

import submitex.resolvepipes as rp

tex = r"There's \inp{|python -c 'print(int(24*60*60*365.25))'} seconds in a year."
print("source:", tex)
print("out   :", rp.convert(tex), '\n')

tex = r"There's $\input { | ls -al ~ | wc -l }$ files/directories in your user directory."
print("source:", tex)
print("out   :", rp.convert(tex))

Output:

source: There's \inp{|python -c 'print(int(24*60*60*365.25))'} seconds in a year.
out   : There's 31557600 seconds in a year.

source: There's $\input { | ls -al ~ | wc -l }$ files/directories in your user directory.
out   : There's $      62$ files/directories in your user directory.

removecomments

import submitex.removecomments as rc
sample = r"""
code % comment
code with escaped \% percent stays
path with backslashes \\% not a comment
line with no comment



next line % another comment

\includegraphics[width=0.5\textwidth]{Fig35.pdf} % original file: example.pdf
   %
%
%
"""
print(rc.clean_text(sample))

Output:

code
code with escaped \% percent stays
path with backslashes \\
line with no comment

next line

\includegraphics[width=0.5\textwidth]{Fig35.pdf}
%

CLI usage

Almost all of the CLIs work like this:

resolvepipes oldmanuscript.tex > newmanuscript.tex
cat oldmanuscript.tex | resolvepipes > newmanuscript.tex

An exception is replacebib which needs another file to work. Typically, the file is called the same as the input file, so for

replacebib oldmanuscript > newmanuscript.tex

the procedure assumes that both oldmanuscript.tex and oldmanuscript.bbl exist in the cwd. Alternatively, provided it explicitly with the --bib flag. Then you can pipe. For instance

cat oldmanuscript.tex | replacebib -b otherbibfile.bbl > newmanuscript.tex

Note that that means you can pipe several or all of the commands together, for instance like so:

replacebib manuscript | replacefigs | collectfloats | resolveinputs | resolvepipes | removecomments > newmanuscript.tex

replacefigs

usage: replacefigs [-h] [-e ENC] [-d] [-F FIGPREFIX] [filename]

Rename all figure files in the `\includegraphics`-environment to generic
enumerated names and copy the respective files with the new names to top-
level.

positional arguments:
  filename              Files to convert

options:
  -h, --help            show this help message and exit
  -e ENC, --enc ENC     encoding
  -d, --dontcopyfigs    Per default, the figures that are found will be copied
                        to the current working directory, but you can turn
                        that off with this flag.
  -F FIGPREFIX, --figprefix FIGPREFIX
                        The prefix for the renamed figures (default: "Fig",
                        such that Fig01, Fig02, ...)

collectfloats

usage: collectfloats [-h] [-e ENC] [filename]

Remove all figure- and table-environments from their respective place in the
document body and put them on their own page at the end of the document, first
figures, then tables.

positional arguments:
  filename           Files to convert

options:
  -h, --help         show this help message and exit
  -e ENC, --enc ENC  encoding

replacebib

usage: replacebib [-h] [-e ENC] [-b BIB] [filename]

Put the content of the bbl-file into the section where previouly laid
`\biblipgraphystyle{...}\bibliography{filename}`

positional arguments:
  filename           Files to convert

options:
  -h, --help         show this help message and exit
  -e ENC, --enc ENC  encoding
  -b BIB, --bib BIB  We'll try to deduce a bib-file from the passed filename
                     of the TeX-source, but in case the bib-file is named
                     differently, you can provided it here

resolveinputs

usage: resolveinputs [-h] [-e ENC] [filename]

Replace `\input{filename}` with the content of `filename`

positional arguments:
  filename           Files to convert

options:
  -h, --help         show this help message and exit
  -e ENC, --enc ENC  encoding

resolvepipes

usage: resolvepipes [-h] [-e ENC] [filename]

Convert \input{|command} to the output of `command`.

positional arguments:
  filename           Files to convert

options:
  -h, --help         show this help message and exit
  -e ENC, --enc ENC  encoding

Example:

resolvepipes manuscript.tex > manuscript_with_executed_commands.tex
cat manuscript.tex | resolvepipes > manuscript_cmds.tex

removecomments

usage: removecomments [-h] [-e ENC] [filename]

Replace all comments with an empty string and squash all empty lines to one empty line

positional arguments:
  filename           Files to convert

options:
  -h, --help         show this help message and exit
  -e ENC, --enc ENC  encoding

Example:

removecomments manuscript.tex > manuscript_with_removed_comments.tex
cat manuscript.tex | removecomments > manuscript_clean.tex

Dependencies

submitex only uses the Python standard library.

License

This project is licensed under the MIT License.