New script to compile Markdown manuscript

chemdtex.sh

@@ -0,0 +1,125 @@
+#!/usr/bin/env bash
+# GPL3 2023 solarchemist
+
+set -euo pipefail
+
+usage() {
+  cat <<EOF
+Usage: $(basename "${BASH_SOURCE[0]}") <flags>
+
+Compiles Markdown manuscript and SI to PDF, and optionally converts to HTML,
+ODT, or DOCX.
+
+Available flags:
++ --help            --> display this usage guide
++ --convert-all     --> convert to HTML, ODT, or other formats
++ --convert-html    --> convert to HTML
+EOF
+  exit
+}
+
+msg() {
+  echo >&2 -e "${1-}"
+}
+
+# Examples:
+# die "some message"
+# die "some message and wait 6 seconds before exiting" 6
+# die "some message and exit immediately" 0
+die() {
+  local msg=$1
+  msg "$msg"
+  # short delay to aid reading last message in case terminal closes on exit
+  # if $2 was provided, set it as delay
+  # if $2 is unset or null, use a 3 second delay
+  simpledelay.sh ${2:-3}
+  exit 0
+}
+
+parse_params() {
+   while :; do
+      case "${1-}" in
+      -h | --help) usage ;;
+      -v | --verbose) set -x ;;
+      --convert-all) convert_formats=true ;;
+      --convert-html) convert_html=true ;;
+      -?*) die "Unknown option: $1" ;;
+      *) break ;;
+      esac
+      shift
+   done
+   return 0
+}
+
+convert_formats=false
+convert_html=false
+parse_params "$@"
+
+
+# note the colons (same function as in $PATH), they are important
+export TEXINPUTS=.:${PWD}/templates/:
+
+# these are indexed bash arrays
+# https://www.cyberciti.biz/faq/bash-for-loop-array/
+declare -a MANUSCRIPT_SOURCES=("paper" "si")
+declare -a MANUSCRIPT_TEMPLATES=("templates/wiley.latex" "templates/wiley-si.latex")
+# these are just strings
+BIBLIOGRAPHIES="--bibliography ./refmngr/zotero.bib --bibliography ./references.bib"
+PANDOC=pandoc
+LATEXMK=latexmk
+
+bibtex_zotero="/media/bay/taha/chepec/literature/_refmngr/_bibrep/library.bib"
+if [ -f "$bibtex_zotero" ]; then
+   msg "Updating the local copy of Zotero library.bib"
+   cp $bibtex_zotero ./refmngr/zotero.bib
+fi
+
+# for some reason, *.aux file is messing up the subsequent run (but not the first run):
+# /paper.aux:224: Undefined control sequence. l.224 \BibFileName[0]{paper.html}
+rm -f *.aux
+
+
+# note the exclamation mark in the for loop variable, makes it the loop index
+# https://stackoverflow.com/a/51794732/1198249
+for M in "${!MANUSCRIPT_SOURCES[@]}"; do
+  # manuscript value in current loop
+  manuscript="${MANUSCRIPT_SOURCES[M]}"
+  # template value in current loop
+  template="${MANUSCRIPT_TEMPLATES[M]}"
+
+  # The --biblatex option is not for use with the --citeproc option or with PDF output
+  # https://pandoc.org/MANUAL.html#citation-rendering
+  msg "pandoc $manuscript.md to TeX"
+  # https://unix.stackexchange.com/questions/278502/accessing-array-index-variable-from-bash-shell-script-loop
+  $PANDOC --standalone --biblatex $manuscript.md -o $manuscript.tex --template $template $BIBLIOGRAPHIES
+
+  msg "latexmk $manuscript.tex to PDF"
+  $LATEXMK -bibtex -pdf $manuscript
+
+  if [[ $convert_formats == "true" || $convert_html == "true" ]]; then
+    # TeX to HTML
+    # successfully produces HTML but after lots of error-like warnings and non-zero exit
+    msg "Converting $manuscript.tex to HTML"
+    # this command creates a HTML file as expected, but returns a non-zero exit code
+    # which causes this script to exit immediately afterwards. Circumvented by "|| true"
+    # --output-dir works, but also produces all the auxiliary files in the root (very weird)
+    make4ht --format html5 $manuscript.tex || true
+  fi
+
+  if [[ $convert_formats == "true" ]]; then
+    # ODT is not too bad, better than html -> docx using pandoc
+    # pandoc cannot find figures or images, paths must be messed up...
+    msg "Converting $manuscript.html to ODT"
+    pandoc -s $manuscript.html -o $manuscript.odt
+    msg "Converting $manuscript.html to DOCX"
+    pandoc -s $manuscript.html -o $manuscript.docx
+    # to also convert to to DOC, consider abiword (works but produces broken equations and tables)
+    # https://stackoverflow.com/a/8384078/1198249
+    # if you want this, please run it on a workstation, not luxor
+    # (abiword is installed on workstations)
+    # abiword --to=doc $manuscript.odt
+  fi
+done
+
+
+exit 0

chertex.sh

@@ -124,6 +124,8 @@ TeXfiletype="tex"
 tikzfiles="*.tikz"
 Rnwfiles="*.Rnw"
 # aux files
+aux_4ct="*.4ct"  # TOC entries for HTML output
+aux_4tc="*.4tc"  # TOC entries for HTML output
 aux_acn="*.acn"
 aux_acr="*.acr"
 aux_alg="*.alg"  # glossaries
@@ -139,6 +141,7 @@ aux_glg="*.glg"
 aux_glo="*.glo"  # glossaries
 aux_gls="*.gls*" # glossaries and bib2gls
 aux_ist="*.ist"  # glossaries (makeindex style file)
+aux_lg="*.lg"    # tex4ht
 aux_lob="*.lob"
 aux_lof="*.lof"
 aux_log="*.log"
@@ -156,13 +159,15 @@ aux_run="*.run*"
 aux_slg="*.slg"
 aux_slo="*.slo"
 aux_sls="*.sls"
-aux_stc="*.stc*" # minitoc
+aux_stc="*.stc*"  # minitoc
 aux_tikz="*-tikzDictionary"
 aux_tdo="*.tdo"
+aux_tmp="*.tmp"   # tex4ht
 aux_toc="*.toc"
-aux_xdy="*.xdy"  # glossaries (xindy)
+aux_xdy="*.xdy"   # glossaries (xindy)
+aux_xref="*.xref" # pandoc or tex4ht cross-refs
 # all aux files in a single string
-auxfiles="${aux_acn} ${aux_acr} ${aux_alg} ${aux_aux} ${aux_bbl} ${aux_blg} ${aux_dep} ${aux_dpth} ${aux_fdb} ${aux_fig} ${aux_fls} ${aux_glg} ${aux_glo} ${aux_gls} ${aux_ist} ${aux_lob} ${aux_lof} ${aux_log} ${aux_lol} ${aux_lor} ${aux_los} ${aux_lot} ${aux_maf} ${aux_out} ${aux_lox} ${aux_make} ${aux_map} ${aux_mtc} ${aux_run} ${aux_slg} ${aux_slo} ${aux_sls} ${aux_stc} ${aux_tikz} ${aux_tdo} ${aux_toc} ${aux_xdy}"
+auxfiles="${aux_4ct} ${aux_4tc} ${aux_acn} ${aux_acr} ${aux_alg} ${aux_aux} ${aux_bbl} ${aux_blg} ${aux_dep} ${aux_dpth} ${aux_fdb} ${aux_fig} ${aux_fls} ${aux_glg} ${aux_glo} ${aux_gls} ${aux_ist} ${aux_lg} ${aux_lob} ${aux_lof} ${aux_log} ${aux_lol} ${aux_lor} ${aux_los} ${aux_lot} ${aux_maf} ${aux_out} ${aux_lox} ${aux_make} ${aux_map} ${aux_mtc} ${aux_run} ${aux_slg} ${aux_slo} ${aux_sls} ${aux_stc} ${aux_tikz} ${aux_tdo} ${aux_toc} ${aux_xdy} ${aux_xref} ${aux_tmp}"
 
 
 if [[ $show_menu == "false" ]]; then