From Koha Test Wiki MediaWiki Postgres

This RFC should be the basis for writing an endpoint for CRUD operations on subscriptions following the guidelines.

Contents 1 Actions and routes 2 subscription object definition 3 Comment 4 Comments

Actions and routes

TODO

subscription object definition

DB schema	Proposed API	Katrin	Details
biblionumber	biblio_id	biblio_id	foreign key for biblio.biblionumber that this subscription is attached to
subscriptionid	subscription_id	subscription_id	unique key for this subscription
librarian	user_id	user_id	librarian's username from borrowers.userid
startdate	start_date	start_date	` date default NULL, -- start date for this subscription
aqbooksellerid	??	vendor_id	int(11) default 0, -- foreign key for aqbooksellers.id to link to the vendor
cost	??	price	int(11) default 0,
aqbudgetid	??	fund_id	int(11) default 0,
weeklength	??	length_in_weeks	int(11) default 0, -- subscription length in weeks (will not be filled in if monthlength or numberlength is set)
monthlength	??	length_in_months	int(11) default 0, -- subscription length in weeks (will not be filled in if weeklength or numberlength is set)
numberlength	??	length_in_issues	int(11) default 0, -- subscription length in weeks (will not be filled in if monthlength or weeklength is set)
periodicity	??	frequency_id	integer default null, -- frequency type links to subscription_frequencies.id
countissuesperunit	??	issues_per_unit	INTEGER NOT NULL DEFAULT 1,
notes	??	notes	LONGTEXT, -- notes
status	??	status	varchar(100) NOT NULL default , -- status of this subscription
lastvalue1	??	last_value_1	int(11) default NULL,
innerloop1	??	inner_counter_1	int(11) default 0,
lastvalue2	??	last_value_2	int(11) default NULL,
innerloop2	??	inner_counter_2	int(11) default 0,
lastvalue3	??	last_value_3	int(11) default NULL,
innerloop3	??	inner_counter_3	int(11) default 0,
firstacquidate	??	first_issue_date	date default NULL, -- first issue received date
manualhistory	??	manual_history (BOOLEAN?)	tinyint(1) NOT NULL default 0, -- yes or no to managing the history manually
irregularity	??	irregularities	MEDIUMTEXT, -- any irregularities in the subscription
skip_serialseq	??	skip_issue_numbers	BOOLEAN NOT NULL DEFAULT 0,
letter	??	notice_code	varchar(20) default NULL,
numberpattern	??	numbering_pattern_id	integer default null, -- the numbering pattern used links to subscription_numberpatterns.id
locale	??	locale	VARCHAR(80) DEFAULT NULL, -- for foreign language subscriptions to display months, seasons, etc correctly
distributedto	??	UNUSED?	MEDIUMTEXT,
internalnotes	??	internal_notes	LONGTEXT,
callnumber	??	callnumber	MEDIUMTEXT, -- default call number
location	??	location	varchar(80) NULL default , -- default shelving location (items.location)
branchcode	??	library_id	varchar(10) NOT NULL default , -- default branches (items.homebranch)
lastbranch	??	UNUSED?	varchar(10),
serialsadditems	??	add_items	tinyint(1) NOT NULL default '0', -- does receiving this serial create an item record
staffdisplaycount	??	staff_display_count	VARCHAR(10) NULL, -- how many issues to show to the staff
opacdisplaycount	??	opac_display_count	VARCHAR(10) NULL, -- how many issues to show to the public
graceperiod	??	grace_period	int(11) NOT NULL default '0', -- grace period in days
enddate	??	end_date	date default NULL, -- subscription end date
closed	??	ended / cancelled (BOOLEAN)	INT(1) NOT NULL DEFAULT 0, -- yes / no if the subscription is closed
reneweddate	??	renewal_date	date default NULL, -- date of last renewal for the subscription
itemtype	??	item_type	VARCHAR( 10 ) NULL,
previousitemtype	??	previous_item_type	VARCHAR( 10 ) NULL,
mana_id	??	mana_id	int(11) NULL DEFAULT NULL,

Comment

Not sure if this is ready for voting yet. As there were a lot of blanks to fill, I've added a new column above with my suggestions.

• user_id = We should really fix Bug 13542, but until we do, the API should still accept a valid patron_id hat is translated to a userid internally.
• letter = We haven't done letter yet. Letter has a combined primary key of module, code and banchcode. Here the code is used, so I suggest: notice_code.

--Kfischer 18:26, 12 January 2020 (EST)

Comments