buku/README.md

MarkIt

markit is a cmdline bookmark management utility written in Python3 and SQLite3. markit exists because of my monumental dependency on historious. I wanted the same database on my local system. However, I couldn't find an equally flexible cmdline solution. Hence, MarkIt!

The SQLite3 database file is stored in $HOME/.cache/markit/bookmarks.db for each user.

markit is GPLv3 licensed.

If you find markit useful, please consider donating via PayPal.

Features

• Add, update or remove a bookmark
• Add tags to bookmarks
• Optionally fetch page title data from the web (default: disabled)
• Use (partial) tags or keywords to search bookmarks
• Any or all search keyword match options
• Unique URLs to avoid duplicates, show index if URL already exists
• Open bookmark in browser using index
• Open search results in browser
• Browser (Chromium and Firefox based) errors and warnings suppression
• Show single bookmark by ID or all bookmarks in a go
• Delete all bookmarks
• Add a bookmark at N^th index, to fill deleted bookmark indices
• Secure parameterized SQLite3 queries to access database
• Handle first level of redirections (reports IP blocking)
• Unicode in URL works
• UTF-8 request and response, page character set detection
• Works with Python 3.x
• Coloured output for clarity
• Easily create batch add or update (refresh) scripts
• Unformatted selective output (for creating batch update scripts)
• Manpage for quick reference
• Optional debug information
• Fast and clean (no ads or clutter)
• Minimal dependencies
• Open source and free

Installation

markit requires Python 3.x to work.

1. If you have git installed (the steps are tested on Ubuntu 14.04.3 x64_64):

$ git clone https://github.com/jarun/markit/  
$ cd markit
$ sudo make install

To remove, run:

$ sudo make uninstall

2. If you do not have git installed:
   Download the latest stable release or development version source code. Extract, cd into the directory and run:

$ sudo make install

If you do not want to install, markit is standalone:

$ chmod +x markit
$ ./markit ...

Usage

Operational notes:

• It's advisable to copy URLs directly from the browser address bar, i.e., along with the leading http:// or https:// token. markit looks up title data (found within tags of HTML) from the web ONLY for fully-formed HTTP(S) URLs.
• If the URL contains characters like ;, & or brackets they may be interpreted specially by the shell. To avoid it, add the URL within single ' or double " qoutes.
• The same URL cannot be added twice. You can update tags and re-fetch title data. You can also delete it and insert at the same index.
• You can either add or update or delete record(s) in one instance. A combination of these operations is not supported in a single run.
• Search works in mysterious ways:
  • Substrings match (match matches rematched) for URL, tags and title.
  • All the keywords are treated together as a single tag in the same order. Bookmarks with partial or complete tag matches are shown in results.
  • -s : match any of the keywords in URL or title. Order is irrelevant.
  • -S : match all the keywords in URL or title. Order is irrelevant.
  • Search results are indexed serially. This index is different from actual database index of a bookmark reord which is shown within () after the URL.

Cmdline help:

Usage: markit [OPTIONS] KEYWORDS...
Bookmark manager. Your private Google.

Options
  -a URL tag 1, tag 2, ...   add URL as bookmark with comma separated tags
  -d N                       delete entry at DB index N (from -P output)
  -D                         delete ALL bookmarks
  -i N                       insert entry at DB index N, useful to fill deleted index
  -o N                       open URL at DB index N in browser
  -p N                       show details of bookmark record at DB index N
  -P                         show all bookmarks along with index from DB
  -s keyword(s)              search all bookmarks for a (partial) tag or any keyword
  -S keyword(s)              search all bookmarks for a (partial) tag or all keywords
  -u N                       update entry at DB index N
  -w                         fetch title info from web, works with -a, -i, -u
  -x N                       works with -P, N=1: show only URL, N=2: show URL and tag
  -z                         show debug information
                             you can either add or update or delete in one instance
                             any other option shows help and exits markit

Keys
  1-N                        open Nth search result in browser. Enter exits markit.

Examples

1. Add a new bookmark with tags linux news and open source:

$ markit -a http://tuxdiary.com linux news, open source
Added at index 15012014

The assigned automatic index 15012014 is unique, one greater than highest index already in use in database. 2. Add a bookmark, fetch page title information from web:

$ markit -a -w http://tuxdiary.com linux news, open source
Title: [TuxDiary | Linux, open source and a pinch of leisure.]
Added at index 15012014

3. Update existing bookmark at index 15012014 with a new tag:

$ markit -u 15012014 -w http://tuxdiary.com linux news, open source, magazine
Title: [TuxDiary | Linux, open source and a pinch of leisure.]
Updated

4. Delete bookmark at index 15012014:

$ markit -d 15012014

5. Delete all bookmarks:

$ markit -D

6. Insert a bookmark at index 15012014 (fails if index or URL exists in database):

$ markit -i 15012014 -w http://tuxdiary.com/about linux news, open source
Title: [A journey with WordPress | TuxDiary]
Added at index 15012014

This option is useful in filling deleted indices from database manually. 7. Show info on bookmark at index 15012014:

$ markit -p 15012014

8. Show all bookmarks with real index from database:

$ markit -P

9. Open URL at index 15012014 in browser:

$ markit -o 15012014

10. Search bookmarks for a tag matching *kernel debugging* or any of the keywords *kernel* and *debugging* in URL or title (separately):

$ markit -s kernel debugging

10. Search bookmarks for a tag matching *kernel debugging* or all the keywords *kernel* and *debugging* in URL or title (separately):

$ markit -S kernel debugging

12. Show debug info:

$ markit -z

13. Show help:

$ markit

14. Check manpage:

$ man markit

15. markit doesn't have any import feature of its own. To import URLs in bulk, create a script with URLs and tags like the following (check TIP below):

#!/bin/bash
markit -aw https://wireless.wiki.kernel.org/ networking, device drivers
markit -aw https://courses.engr.illinois.edu/ece390/books/artofasm/ArtofAsm.html assembly
markit -aw http://www.tittbit.in/
markit -aw http://www.mikroe.com/chapters/view/65/ electronics
markit -aw "http://msdn.microsoft.com/en-us/library/bb470206(v=vs.85).aspx" file systems
markit -aw http://www.ibm.com/developerworks/linux/library/l-linuxboot/index.html boot process

Make the script executbale and run to batch add bookmarks. 16. To update all URLs (refresh) along with your tags, first get the unformatted selective output with URL and tags:

$ markit -P -x 2 | tee myurls

Add markit -wu in front of all the lines (check TIP below). Should look like:

#!/bin/bash
markit -wu 50 https://wireless.wiki.kernel.org/ networking, device drivers
markit -wu 51 https://courses.engr.illinois.edu/ece390/books/artofasm/ArtofAsm.html assembly
markit -wu 52 http://www.tittbit.in/
markit -wu 53 http://www.mikroe.com/chapters/view/65/ electronics
markit -wu 54 "http://msdn.microsoft.com/en-us/library/bb470206(v=vs.85).aspx" file systems
markit -wu 55 http://www.ibm.com/developerworks/linux/library/l-linuxboot/index.html boot process

Run the script:

$ chmod +x myurls
$ ./myurls

TIP:
To add the same text at the beginning of multiple lines using vim editor:

• Press Ctrl-v to select the first column of text in the lines you want to change (visual mode).
• Press Shift-i and type the text you want to insert.
• Hit Esc, wait 1 second and the inserted text will appear on every line.

Using sed:

$ sed -i 's/^/markit -wu /' filename

#License GPL v3
Copyright (C) 2015 by Arun Prakash Jana <engineerarun@gmail.com>

Contributions

I would love to see pull requests with the following features:

• Exact word match (against substring in a word as it works currently. Hint: REGEXP)
• Parse full page data??? Might end up writing a search engine like Google. ;)
• Optional password protection

Developer(s)

Arun Prakash Jana <engineerarun@gmail.com>