Zotero database browser for iOn

[tjdean256]

If you didn't know already, the database created by the recently released 2.02 version of Zotero seems to be quite efficient. Zbi2 now filters a 11,000 record database fairly quickly.

td

[stsu]

That's news to me.
The version I am testing now creates its own database from zotero.sqlite, so that the search by any combination of title, author etc. is done within a reasonable time. I will check if it is possible without creating new database.
Thanks.

[stsu]

IMPORTANT!

Updated versions are available. See the first post of this thread.

zbi opens the database file zotero.sqlite in read-only mode, but I recently noticed that sqlite might write journal files to the file system even if the database is opened in read-only mode. Since SD cards have a limited number of writes, it can shorten the lifespan of the SD card. I am sorry for this bad behaviour.

In version 0.7_1 and later, the journal mode of sqlite is set to MEMORY, that is, the journal files are stored in memory and also the temp store is set to MEMORY, that is temporary files are stored in memory, hence the write access to the card would be minimal as long as I understand (but at the expence of database safety).

I tried to keep the write access to SD card as minimal as possible, and I hope that these new versions would achieve it. But I am not sure about the behaviour of sqlite. Any suggestion is welcome.

[stsu]

Version 0.9

The zip file is attached to the first post of this thread.

* Works for both zotero1 and zotero2.
* Query by any combination of collection, tag, creator etc is available.
* Optionally, you can add Publication column.
* Menu added.
* Appropriate tags only will be displayed.
* PC utilities to create zbi's database included.

INSTALL:
Install procedure is the same as before.
Unpacking the zip file, you will get two directories named "Programs" and "utilities". The "utilities" directory contains perl, python and ruby script files named "zotero2zbi" and a txt file "sql.txt", which are to be used on PC. See below.

DESCRIPTION:
Since this version, for performance reason, zbi creates its own database on memory or file "Programs/_zbi/share/zbi/zbi.sqlite".
When you use this version for the first time (and if you do not use one of zotero2zbi scripts, see below), a dialog would appear. In the dialog, there are 4 check boxes:
"On Memory Database" -- Choose "onmemory" database or creating database file. The "onmemory" database might be of use only when your zotero library is small.
"Create Index" -- If your zotero library is small, then unticking this could make search faster.
"Publication Column" -- Add a column and search buttons for publication column. Adding the column would make database creation and search slower.
"Create Metadb" -- Create metadata.db (if not exist) or not on opening an attachment file.

You can change these settings from menu.

Check your preference and click the "OK" button.
Then database will be created. It would take a while. In my case (about 1800 items), it took about 30 seconds. (Optionally, you can use "zotero2zbi" to create the database on your PC.)

Edit: In the case when the zotero library contains about 600 items, it took about 3 seconds.

If you choose onmemory database, then every time you start zbi, database will be created on memory and it will take as long as the first time. Otherwise, zbi starts much faster from next time.

(zbi will recreate database when
you changed the settings from menu or
the file "zotero.sqlite" is changed.
The later is checked by the file modified time, hence, be sure that the clocks of your PC and DR1000 are (almost) on time.)

PC UTILITIES:
The "utilities" directory contains:
sql.txt -- data file used by the following scripts.
zotero2zbi.pl -- perl script.
zotero2zbi.py -- python (2.5 or 2.6) script.
zotero2zbi.rb -- ruby script.

Every script will do the same work.
For example, in the case of python in unix like environment:

Usage: zotero2zbi.py [options] your_zotero.sqlite

Options:
-h, --help show this help message and exit
-i WITHIDX Create index if 1, drop index if 0, default 1
-o OPF Create Publication column if 1, do not if 0, default 0
-m WITH_METADB Create meta_db if 1, do not if 0, default 1
-d OUTDIR Output directory, default is current


Then 2 files

zbi.sqlite, zbi.conf

will be created in

OUTDIR/Programs/_zbi/share/zbi

and copy these two files to DR1000.

[stsu]

Version 0.9_1

* Some bugs fixed.

The zip file is attached to the first post of this thread.