From 85ddba33d79aefe569a0c8bae63eb481058e5f10 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Maciej=20Wilczy=C5=84ski?= <maciej@wmi.amu.edu.pl>
Date: Fri, 7 Apr 2017 11:43:35 +0200
Subject: [PATCH] Added markdown cleanup, gitignore, and makefile

---
 .gitignore          |  3 +++
 Makefile            | 36 ++++++++++++++++++++++++++++++++++++
 convert_markdown.py |  7 ++++---
 md_cleanup.py       | 30 ++++++++++++++++++++++++++++++
 wiki_cleanup.py     |  1 -
 5 files changed, 73 insertions(+), 4 deletions(-)
 create mode 100644 .gitignore
 create mode 100644 Makefile
 create mode 100644 md_cleanup.py

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..7380cf6
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,3 @@
+md/
+wiki/
+oldwiki/
diff --git a/Makefile b/Makefile
new file mode 100644
index 0000000..283749a
--- /dev/null
+++ b/Makefile
@@ -0,0 +1,36 @@
+PYTHON = python3
+DIR = $(CURDIR)
+WIKIOUTDIR = $(DIR)/wiki
+MDOUTDIR = $(DIR)/md
+OLDDIR = $(DIR)/oldwiki
+
+.PHONY: default clean clean_md clean_wiki \
+	export initial_cleanup final_cleanup \
+	convert all
+
+default:
+	echo "Moin export and convert -- read Makefile for more details"
+
+clean: clean_md clean_wiki
+
+clean_md:
+	rm -rf $(MDOUTDIR)
+
+clean_wiki:
+	rm -rf $(WIKIOUTDIR)
+
+export: moin_export.py
+	mkdir $(WIKIOUTDIR) &> /dev/null || true
+	$(PYTHON) moin_export.py $(OLDDIR) $(WIKIOUTDIR)
+
+initial_cleanup: wiki_cleanup.py
+	$(PYTHON) wiki_cleanup.py $(WIKIOUTDIR)
+
+final_cleanup: md_cleanup.py
+	$(PYTHON) md_cleanup.py $(MDOUTDIR)
+
+convert: convert_markdown.py
+	mkdir $(MDOUTDIR) &> /dev/null || true
+	$(PYTHON) convert_markdown.py $(WIKIOUTDIR) $(MDOUTDIR)
+
+all: clean export initial_cleanup convert final_cleanup
diff --git a/convert_markdown.py b/convert_markdown.py
index 957435a..0afc962 100644
--- a/convert_markdown.py
+++ b/convert_markdown.py
@@ -6,17 +6,18 @@ import subprocess
 
 # argv[1] - wiki files directory
 
-if len(sys.argv) < 2:
-	print("Argument missing. Usage: python3 convert_markdown.py <wiki content dir>")
+if len(sys.argv) < 3:
+	print("Argument missing. Usage: python3 convert_markdown.py <wiki content dir> <output dir>")
 else:
 	path = sys.argv[1]
+	output_path = sys.argv[2]
 	if os.path.exists(path):
 		files = [ f for f in listdir(path) if isfile(join(path, f)) ]
 		for filename in files:
 			name, file_extension = os.path.splitext(filename)
 			if file_extension == ".txt":
 				print("Converting %s/%s" % (path, filename))
-				options = ['pandoc','-f','mediawiki','-t','markdown','-s',join(path,filename),'-o',join(path,name + ".md")]
+				options = ['pandoc','-f','mediawiki','-t','markdown','-s',join(path,filename),'-o',join(output_path,name + ".md")]
 				subprocess.call(options)
 	else:
 		print("Error: the path specified is invalid")
diff --git a/md_cleanup.py b/md_cleanup.py
new file mode 100644
index 0000000..babacb5
--- /dev/null
+++ b/md_cleanup.py
@@ -0,0 +1,30 @@
+from os import listdir
+from os.path import isfile, join
+import os
+import sys
+import re
+import subprocess
+
+# argv[1] - markdown files directory
+
+if len(sys.argv) < 2:
+	print("Argument missing. Usage: python3 md_cleanup.py <md content dir>")
+else:
+	path = sys.argv[1]
+	if os.path.exists(path):
+		files = [ f for f in listdir(path) if isfile(join(path, f)) ]
+		for filename in files:
+			name, file_extension = os.path.splitext(filename)
+			if file_extension == ".md":
+				print("Cleaning up %s/%s" % (path, filename))
+				with open(join(path,filename), "r+", encoding="utf-8") as file:
+					content = file.read()
+					content = re.sub(r"`\*(.*)`[\\]?", r"*\g<1>", content) 
+					content = re.sub(r"!(\w+)", r"\g<1>", content) 
+					file.seek(0)
+					file.truncate()
+					file.write(content)
+					file.close()
+	else:
+		print("Error: the path specified is invalid")
+
diff --git a/wiki_cleanup.py b/wiki_cleanup.py
index 2f7c322..64186c8 100644
--- a/wiki_cleanup.py
+++ b/wiki_cleanup.py
@@ -17,7 +17,6 @@ else:
 			name, file_extension = os.path.splitext(filename)
 			if file_extension == ".txt":
 				print("Cleaning up %s/%s" % (path, filename))
-				#options = ['pandoc','-f','mediawiki', '-t', 'markdown', '-s', join(path,filename), '-o', join(path,name + ".md")]
 				with open(join(path,filename), "r+", encoding="utf-8") as file:
 					content = file.read()
 					content = re.sub("(\-{3,}\n(.*)[\n]*)$", "", content) # remove trailing categories list