longnow-for-markdown

commit 8257613d8c39b7e3677fc216b84c84b5d53c4021
parent 0a1b5f172733387ba5277e53b1408606f0af0c4d
Author: NunoSempere <nuno.sempere@gmail.com>
Date:   Fri, 30 Jul 2021 20:27:05 +0200

feat: Show how to use this to back up a Google doc

I also removed PPA fuckery.

Diffstat:
M
README.md
 | 
50
+++++++++++++++++++++++++++++++++++++++-----------

1 file changed, 39 insertions(+), 11 deletions(-)
diff --git a/README.md b/README.md
@@ -1,18 +1,8 @@
 This package takes a markdown file, and creates a new markdown file in which each link is accompanied by an archive.org link, in the format [...](original link) ([a](archive.org link)).
 
 ## How to install
-Several different ways: 
 - Add [this file](https://github.com/NunoSempere/longNowForMd/blob/master/longnowformd.sh) to your path, for instance by moving it to the `/usr/bin` folder and giving it execute permissions, or
-- copy its functions into your .bashrc file, or
-- for Ubuntu distributions 20.04 (Focal Fossa) and above:
-
-```
-$ sudo add-apt-repository ppa:nunosempere/longnowformd
-$ sudo apt-get update
-$ sudo apt install longnowformd
-```
-
-This uses the [nunosempere/longnowformd](https://launchpad.net/~nunosempere/+archive/ubuntu/longnowformd) PPA (Personal Package Archive)
+- copy its content (except the last line) into your .bashrc file, or
 
 This utility requires [archivenow](https://github.com/oduwsdl/archivenow) as a dependency, which itself requires a python installation. It can be installed with
 
@@ -32,3 +22,41 @@ For a reasonably sized file, the process will take a long time, so this is more 
 - Deal elegantly with images. Right now, they are also archived, and have to be removed manually afterwards.
 - Possibly: Throttle requests to the internet archive less. Right now, I'm sending a link roughly every 12 seconds, and then sleeping for a minute every 15 requests. This is probably too much throttling (the theoretical limit is 15 requests per minute), but I think that it does reduce the error rate. 
 - Pull requests are welcome.
+
+## How to use to back up Google Files
+
+You can download a .odt file from Google, and then convert it to a markdown file with 
+
+```
+function pandocodt(){
+  source="$1.odt"
+  output="$1.md"
+  pandoc -s "$source" -t markdown-raw_html-native_divs-native_spans-fenced_divs-bracketed_spans | awk ' /^$/ { print "\n"; } /./ { printf("%s ", $0); } END { print ""; } ' | sed -r 's/([0-9]+\.)/\n\1/g' | sed -r 's/\*\*(.*)\*\*/## \1/g'  | tr -s " " | sed -r 's/\\//g' | sed -r 's/\[\*/\[/g' | sed -r 's/\*\]/\]/g' > "$output"
+  ## Explanation: 
+  ## markdown-raw_html-native_divs-native_spans-fenced_divs-bracketed_spans: various flags to generate some markdown I like
+  ## sed -r 's/\*\*(.*)\*\*/## \1/g': transform **Header** into ## Header
+  ## sed -r 's/\\//g': Delete annoying "\"s
+  ## awk ' /^$/ { print "\n"; } /./ { printf("%s ", $0); } END { print ""; } ': compress paragraphs; see https://unix.stackexchange.com/questions/6910/there-must-be-a-better-way-to-replace-single-newlines-only
+  ## sed -r 's/([0-9]*\.)/\n\1/g': Makes lists nicer.
+  ## tr -s " ": Replaces multiple spaces
+}
+
+## Use: pandocodt FileNameWithoutExtension
+```
+
+Then, convert the file back to html with 
+
+```
+function pandocmd(){
+  source="$1.md"
+  output="$1.html"
+  pandoc -r gfm "$source" -o "$output"
+  ## sed -i 's|\[ \]\(([^\)]*)\)| |g' "$source" ## This removes links around spaces, which are very annoying. See https://unix.stackexchange.com/questions/297686/non-greedy-match-with-sed-regex-emulate-perls
+}
+
+## Use: pandocmd FileNameWithoutExtension
+```
+
+(this requires changing the name of the output file from `Source.md.longnow` to `Source.longnow.md` before running `$ pandocmd Source.longnow`)
+
+Then copy and paste the html into a Google doc and fix fomatting mistakes.