Use a lua filter rather than a massive python script

.circleci/config.yml

@@ -6,12 +6,6 @@ jobs:
         working_directory: ~/tex-template
         steps:
             - checkout
-            - run:
-                name: Install Dependencies
-                command: apt-get install -y python-flake8
-            - run:
-                name: Lint
-                command: flake8 build.py --ignore=E128,E501
             - run:
                 name: Test build
-                command: python3 build.py test/main.md
+                command: bash build.sh test/main.md

build.py

@@ -1,68 +0,0 @@
-#!/usr/bin/env python3
-import argparse
-import pathlib
-import os
-import subprocess
-import re
-import tempfile
-
-
-CWD = pathlib.Path(os.getcwd())
-BASE_DIR = pathlib.Path(os.path.dirname(os.path.abspath(__file__)))
-TEMPLATE_FILE = BASE_DIR / 'main.tex'
-OUTPUT_FILE = CWD / 'output.pdf'
-WORDCOUNT_TEMPLATE = BASE_DIR / 'wordcount.yaml'
-
-
-def build(input_file, additional_args=None):
-    build_args = ['pandoc', str(input_file)]
-
-    if additional_args is not None:
-        build_args.extend(additional_args)
-
-    build_args.extend([
-        '--template',
-        str(TEMPLATE_FILE),
-        '-o',
-        str(OUTPUT_FILE)
-    ])
-    subprocess.check_output(build_args, cwd=str(BASE_DIR))
-
-
-def get_word_count():
-    words = subprocess.check_output([
-        'pdftotext', str(OUTPUT_FILE), '-'
-    ], cwd=str(BASE_DIR))
-    return len(re.findall(r'\w+', words.decode()))
-
-
-def write_wordcount():
-    _, out_file_path = tempfile.mkstemp(suffix='.yaml')
-    wordcount = get_word_count()
-    with open(out_file_path, 'w') as out_file:
-        with WORDCOUNT_TEMPLATE.open() as template:
-            out_file.write(template.read().format(wordcount=wordcount))
-    return pathlib.Path(out_file_path)
-
-
-def parse_args():
-    parser = argparse.ArgumentParser()
-    parser.add_argument("-v", "--verbose", help="Set verbosity level (repeat argument)", action="count", default=0)
-    parser.add_argument("input", help="Input File", action='store', type=pathlib.Path,)
-    parser.add_help = True
-    return parser.parse_args()
-
-
-if __name__ == '__main__':
-    args = parse_args()
-    input_file = args.input.resolve()
-    operating_dir = input_file.parents[0]
-    context_file = operating_dir.joinpath('context.yaml')
-    if not input_file.is_file():
-        raise ValueError("Failed to find {}".format(input_file))
-    additional_args = []
-    if context_file.exists():
-        additional_args.append(str(context_file))
-    build(input_file, additional_args)
-    additional_args.append(write_wordcount())
-    build(input_file, additional_args)

build.sh

@@ -0,0 +1,15 @@
+#!/usr/bin/env bash
+
+SCRIPT=`realpath $0`
+SCRIPTPATH=`dirname $SCRIPT`
+
+INPUT_FILE=`realpath $1`
+OUTPUT_FILE=output.pdf
+
+PANDOC_ARGS="$INPUT_FILE --template $SCRIPTPATH/main.tex -o $OUTPUT_FILE"
+
+echo "Building document..."
+pandoc $PANDOC_ARGS
+
+echo "Running second pass..."
+pandoc $PANDOC_ARGS --lua-filter=second-pass.lua

second-pass.lua

@@ -0,0 +1,23 @@
+function file_exists(name)
+    local f=io.open(name,"r")
+    if f~=nil then io.close(f) return true else return false end
+ end
+
+
+local function countWords()
+    local rawText = pandoc.pipe("pdftotext", {"output.pdf", "-"}, "")
+    return pandoc.pipe("wc", {"-w"}, rawText)
+end
+
+
+return {
+    {
+      Meta = function(meta)
+        if file_exists("output.pdf") then
+            meta["wordcount"] = countWords()
+            meta["secondpass"] = true
+        end
+        return meta
+      end,
+    }
+}

wordcount.yaml

@@ -1,4 +0,0 @@
----
-wordcount: {wordcount}
-secondpass: true
-...