(save-buffer) (let ((org-confirm-babel-evaluate (lambda (lang body) (declare (ignore lang body)) nil))) (org-latex-export-to-pdf) (org-latex-export-to-pdf))
(save-buffer) (let ((org-babel-tangle-comment-format-beg "%file [[%link][%source-name]]") (org-babel-tangle-comment-format-end "End %file [[%link][%source-name]]")) (org-babel-tangle))
CLOSED: [2016-03-06 Sun 13:53]
Genealogy in general is a fairly interesting subject, and can, in the study of history be important. While there are many libraries and even applications to deal with genealogical data, there has yet to be an adequate genealogy library for Lisp. The goal is to build a library upon which to build a genealogical research application, and also to possibly integrate into a later application for the management of museum collections and meta-data.
This library is designed from the perspective of a Computer Scientist, rather than that of a genealogist, and as such is not meant to be used directly, but instead is meant to be the basis of a much better genealogy application.
One of the key components of genealogical software is the ability to effectively store information related to members of a family. This will include things such as people, birth records, death records, marriage records, divorce records and notes.
To accomplish the storage of this information, I'm using a schema-less relational database management system called LambdaLite, which is written entirely in lisp, and uses the local file system to store data effectively.
To accomplish the goal of storing data within the database system that is LambdaLite, I need to be able to constrain various attributes to be certain values. To do this, instead of having to manually write code to constrain them each time, I use a few macros to do this. These macros include:
Constrains to be one of several values. This macro takes one or more arguments, and generates a function that checks to see if the argument passed to it is in any of the given values.
Checks to see if the value is one of the values in a specific column. This macro takes two arguments, a table name and a column name, then generates a function which will calculate all possible values, iterating for each row in the table, getting the values in the specified column, and then testing the value passed to the function to see if it is a member of the possible values.
Checks to see if this is one of the possible IDs in a table column.
This is done through the same mechanism as in-table-column
, but simply ensuring that the passed object is a) a list, and b) the elements in the list are mapped through the in-table-column
result, with the values of a and be being reduced using and
to produce a single boolean.
This checks to see whether or not the passed value is within the specified column, ensuring that it is not within it. This function takes three things, a table, a column and a type. It first ensures that the passed object is of the specified type, and if it is, calculates all values currently in the specified table and column, and then checks to see that the passed object is not within the list.
(defmacro constrain-values (&rest values) `(lambda (object) (the boolean (member object (list ,@values))))) (defmacro in-table-column (table column) `(lambda (object) (let ((possible-values (iter (for row in (select ,table)) (collect (,column row))))) (the boolean (member object possible-values))))) (defmacro ids-in-table-column (table column) `(lambda (object) (let ((possible-values (iter (for row in (select ,table)) (collect (,column row))))) (and (listp object) (reduce #'and (map 'list ,(in-table-column table column) object)))))) (defmacro unique-in-column (table column type) `(lambda (object) (if (typep object ',type) (let ((possible-values (iter (for row in (select ,table)) (collect (,column row))))) (not (the boolean (member object possible-values))))))) (defun generate-table-id (table) (check-type table keyword) (1+ (length (select table))))
This is the People table, used to store the most bare information about a person. This includes the following:
The ID used to identify the person within the database.
The name of the person.
The gender of the person. Can be one of:
M
F
T
The person's father. This is either an person ID, or 0 if we don't know who the father is.
The person's mother. As with father, this can be either a person ID or 0.
(defattributes :/person-id (unique-in-column :people :/person-id integer) :/person-name #'stringp :/gender (constrain-values "M" "F" "T") :/father (lambda (object) (or (= 0 object) (funcall (in-table-column :people :/person-id) object))) :/mother (lambda (object) (or (= 0 object) (funcall (in-table-column :people :/person-id) object))))
Another important thing is to be able to log births, or possible birth dates. To do this, you need four pieces of information:
The ID used to reference the person's birth.
The ID of the person born.
When the person was born.
Where the person was born.
(defattributes :/birth-id (unique-in-column :births :/birth-id integer) :/birth-person (in-table-column :people :/person-id) :/birth-date #'stringp :/birth-location #'stringp)
Furthermore, to be as complete as possible, you need to be able to store and query death information. This includes things such as:
The ID used to track this death record.
The ID of the person who died.
When the person died.
Where the person died.
(defattributes :/death-id (unique-in-column :deaths :/death-id integer) :/death-person (in-table-column :people :/person-id) :/death-date #'stringp :/death-location #'stringp)
Further, to be able to keep track of relationships (and thus families), you need to be able to track marriages. This entails keeping track of the following information:
ID used to track the marriage within this system.
ID of the husband in the marriage.
ID of the wife in the marriage.
Date the marriage was considered to have started.
Date the marriage ended (Divorce, death, annulment).
(defattributes :/marriage-id (unique-in-column :marriages :/marriage-id integer) :/husband (in-table-column :people :/person-id) :/wife (in-table-column :people :/person-id) :/wedding-date #'stringp :/end-date #'stringp)
To keep track of the dissolution of marriages, and to enable correct report generation, we must keep track of divorces. To do this, we store the following information:
How the divorce is referred to within the database.
The ID of the marriage the divorce terminates
The date the Divorce is effective.
(defattributes :/divorce-id (unique-in-column :divorces :/divorce-id integer) :/marriage (in-table-column :marriages :/marriage-id) :/divorce-date #'stringp)
Keeping notes within the database is a good idea, it allows the notes to be linked directly to the relevant data, and can help to keep organized. To store a note, you need the following pieces of data:
The ID used to reference the note.
The title of the note.
The text of the note, formatted using markdown.
An optional link to a media file, such as an image or oral history.
(defattributes :/note-id (unique-in-column :notes :/note-id integer) :/note-title #'stringp :/note-text #'stringp :/media-link #'stringp)
As LambdaLite is schemaless, the following attributes can be mixed in to other tables, and can be used to help link records quickly and easily.
The ID of a relevant Person.
The ID of a relevant Birth.
The ID of a relevant Death.
The ID of a relevant Marriage.
The ID of a relevant Divorce.
(defattributes :/person (in-table-column :people :/person-id) :/birth (in-table-column :births :/birth-id) :/death (in-table-column :deaths :/death-id) :/marriage (in-table-column :marriages :/marriage-id) :/divorce (in-table-column :divorces :/divorce-id))
[0/10]
[0/5]
[0/3]
(defun generate-ahnentafel-numbers (starting-person number-of-generations) (let ((ahnentafel-list (cons (cons 1 starting-person) nil))) (labels ((generate-number (current gender) (if (string= gender "M") (* 2 current) (1+ (* 2 current)))) (recurse (person number generation gender) (if (not (= generation 0)) (let ((new-number (generate-number number gender)) (father (:/father person)) (mother (:/mother person))) (push (cons new-number person) ahnentafel-list) (if (not (= 0 father)) (recurse father new-number (1- generation) "M")) (if (not (= 0 mother)) (recurse mother new-number (1- generation) "F")))))) (recurse (:/father starting-person) 1 (1- number-of-generations) "M") (recurse (:/mother starting-person) 1 (1- number-of-generations) "F")) ahnentafel-list))
(defun format-ahnentafel-record (record) (destructuring-bind (number . person) record (let ((name (:/person-name (get-person person))) (birthdate (:/birth-date (get-birth person))) (death (let ((death-record (get-death person))) (if (null death-record) "" (format nil " -- ~a" (:/death-date death-record)))))) (format nil "~10,5R: ~A, ~A~A" number name birthdate death)))))
(defun output-ahnentafel (file start-person total-generations) (let ((ahnentafel-text (map 'list #'format-ahnentafel-record (sort (generate-ahnentafel-numbers start-person total-generations) #'< :key #'car)))) (with-open-file (output file :direction :output :if-exists :output :if-does-not-exist :create) (map 'list #'(lambda (line) (format output "~A~&" line)) ahnentafel-text))))
[0/5]
[0/6]
(defpackage #:config-parser (:use :esrap :cl) (:import-from #:parse-number #:parse-number) (:export open-configuration-file)) (defpackage #:cl-genealogy (:use #:cl #:lambdalite #:iterate #:archive) (:export generate-graph print-ahnentafel database new-person new-death new-marriage new-divorce))
(in-package #:cl-genealogy) <<ahnentafel-numbering>> <<format-ahnentafel-record>> <<output-ahnentafel>>
(asdf:defsystem #:cl-genealogy :description "Describe genie here" :author "Samuel Flint <swflint@lisp.technology>" :license "GNU GPLv3 or Later" :depends-on (#:cl-utilities #:esrap #:parse-number #:lambdalite #:iterate #:archive) :serial t :components ((:file "package") (:file "database") (:file "family-tree") (:file "ahnentafel")))
This document, and the code, forming the cl-genealogy
package is version src_sh{git describe –always –long –dirty –abbrev=10 –tags}