chrrrisw / pybooksearch Goto Github PK

Fill mode should query as many searchers as possible to resolve all UNKNOWN fields in the book.

Now that we can search on LCCN, perhaps it's best to include this in the book and CSV as a separate field. In fact, we often get back additional ISBN information in the form of both ISBN_10 and ISBN_13 numbers, so perhaps we should store these separately.

Includes:

• ISBN/LCCN mode
• Fill mode
• Multiple searchers

I would like to be able to export the CSV file in a format that can be imported by the Book Catalogue android app.

Currently the ISBNSearchOrg searcher replaces '&' with '&' to make the resultant strings HTML/XML safe. Should this be the case? Apart from the fact that this is an incomplete solution to the problem, should this really be done in the GTK interface where it's required - rather than corrupt the raw data?

For those of us with many bookcases, and for those packing books for transport, the ability to specify a location would be useful. In text mode the command could look something like:

l:box1

followed by all the books in the 'box1' location. To change location, simply type:

l:box2

or:

l:

to go back to not specifying a location.

With more than two searchers, conflict resolution may get tedious. A future enhancement may leave conflict resolution until all options are in.

Eventually we may want to able to package the code. The reorg branch has been pushed with current functionality, but with the code reorganised to be more package like.

As it says on the tin.

Currently, if the ISBN lookup fails, the user is given a "could not contact server" message - this is misleading.

I've found that the csv.Sniffer occasionally fails to detect the delimiter and aborts the program.

If the delimiter is set on the command line, don't do the sniff test.

In fill mode, if there is a conflict between two sources, the user choice is put into the book field in lower case. This is a result of the lower case comparison which should be left as-is, the user choices should be the original data.

Re-query mode should read in an existing library file and query the searchers again for any updated information. In combination with "fill mode" this will allow a library file to fill in any UNKNOWN fields that may have been added in the interim.

I know you haven't worked on this in a long time, but hopefully you can still help me out. The script seems to work fine, but it doesn't appear to actually update the auto_library.csv file, despite saying "Saved ./auto_library.csv".

Any thoughts on what might be happening?

Currently the website used to perform the search is hard coded, this should be selectable.

I might be doing something stupid, but after doing a search for an ISBN, saving the results, and then trying to execute the script again, I get this error message immediately.

The full Traceback shows that it's originating from the new_book function originating in crwLibrary.py

It first fails on the self._binding = value.strip()

Any thoughts?

The current delimiter is now '|', but the test data uses ','. Either transform the test data, or use the csv.Sniffer class to determine the CSV dialect on read. Perhaps also allow selection of delimiter for writing via command line parameter.

Text mode only allows adding new entries and provides little feedback to the user. Improve feedback, and allow editing of entries if the search fails.

As auto_library.csv is the default name for the output file, having the test data in this file means that users have to remove this file to start their own library. This will result in a local change to the repository that will potentially clash with future updates to the test data.

WARNING
When this is closed, all users of auto_library.csv should save a copy of their data elsewhere and then copy it back after pulling the change.