In my eyes, besides using Evolution there are two ways to import the contacts. There is no automatic tool to do this without the Evolution.
Make a backup of your contacts.db before making any changes!
The first way is the easy way using folks-import to import your contacts. I have checked the source and it appears to be a importer from pidgin's blist.xml file.
blist.xml - A local copy of your buddy lists, for use in keeping locally applied aliases and group and buddy ordering between. I have tried it, out of curiosity, on my blist.xml and it worked for some contacts.
For you to create a blist.xml (here is a blist.h and the source for it) you could use somebody's work like vcftoxml.py and adjust it.
Edit - trying to create a dummy xml to be imported
I have tried to create simple xml (blist.xml) to be imported via the the importer.
The xml inspired by blist.xml example:
<?xml version="1.0" encoding="UTF-8" ?>
<purple version="1.0">
<blist>
<group name='GroupName'>
<contact>
<buddy account='test' proto='prpl-hangouts'>
<name>1131313213446</name>
<alias>Jakto Tomos</alias>
<phone>+6568309312</phone>
<email>testing@yahoo.com</email>
<address>Berlin</address>
</buddy>
</contact>
</group>
</blist>
<privacy>
</privacy>
</purple>
When I checked the created contact I have found out that only the alias name was imported.
So I have create blist.xml which mimics the real-life one:
<?xml version="1.0" encoding="UTF-8" ?>
<purple version="1.0">
<blist>
<group name='GroupName'>
<contact>
<buddy account='test' proto='prpl-hangouts'>
<name>1131313213446</name>
<alias>Jakto Tomos</alias>
<setting name='phone' type='string'>+6568309312</setting>
<setting name='email' type='string'>testing@yahoo.com</setting>
<setting name='address' type='string'>Lesser strasse, Berlin</setting>
</buddy>
</contact>
</group>
</blist>
<privacy>
</privacy>
</purple>
When I imported it again I got again only the alias name. So I went to the import-pidgin.vala source code and I have found out that it indeed imports only the name (contact ID) and alias as name:
...
/* Parse the <name> and <alias> elements beneath <buddy> */
for (Xml.Node *subiter = iter->children; subiter != null;
subiter = subiter->next)
{
if (subiter->type != ElementType.ELEMENT_NODE)
continue;
if (subiter->name == "alias")
alias = subiter->get_content ();
else if (subiter->name == "name")
{
/* The <name> element seems to give the contact ID, which
* we need to insert into the Persona's im-addresses property
* for the linking to work. */
string im_address = subiter->get_content ();
im_addresses.set (tp_protocol,
new ImFieldDetails (im_address));
im_address_string += " %s\n".printf (im_address);
}
}
...
So this way is feasible but you have to add parsing of the nodes you want to import. You could make a PR for such patch so others after you could use it (I recommend using the <setting name... notation as this is how the blist.xml file is formatted.
- The second way is way harder and you have to correctly populate the tables in the sqlite. This is the DB structure:
BEGIN TRANSACTION;
CREATE TABLE IF NOT EXISTS "folder_id_phone_list" (
"uid" TEXT NOT NULL,
"value" TEXT,
FOREIGN KEY("uid") REFERENCES "folder_id"("uid")
);
CREATE TABLE IF NOT EXISTS "folder_id_email_list" (
"uid" TEXT NOT NULL,
"value" TEXT,
FOREIGN KEY("uid") REFERENCES "folder_id"("uid")
);
CREATE TABLE IF NOT EXISTS "folder_id" (
"uid" TEXT,
"Rev" TEXT,
"file_as" TEXT,
"file_as_localized" TEXT,
"nickname" TEXT,
"full_name" TEXT,
"given_name" TEXT,
"given_name_localized" TEXT,
"family_name" TEXT,
"family_name_localized" TEXT,
"is_list" INTEGER,
"list_show_addresses" INTEGER,
"wants_html" INTEGER,
"x509Cert" INTEGER,
"vcard" TEXT,
"bdata" TEXT,
PRIMARY KEY("uid")
);
CREATE TABLE IF NOT EXISTS "keys" (
"key" TEXT,
"value" TEXT,
"folder_id" TEXT,
PRIMARY KEY("key"),
FOREIGN KEY("folder_id") REFERENCES "folders"
);
CREATE TABLE IF NOT EXISTS "folders" (
"folder_id" TEXT,
"version" INTEGER,
"multivalues" TEXT,
"lc_collate" TEXT,
"countrycode" VARCHAR(2),
PRIMARY KEY("folder_id")
);
INSERT INTO "keys" VALUES ('eds-reserved-namespace-is-populated','1','folder_id');
INSERT INTO "keys" VALUES ('revision','2019-11-04T07:28:25Z(0)','folder_id');
INSERT INTO "folders" VALUES ('folder_id',11,'email;prefix:phone','en_US.UTF-8','US');
CREATE INDEX IF NOT EXISTS "UID_INDEX_phone_folder_id" ON "folder_id_phone_list" (
"uid"
);
CREATE INDEX IF NOT EXISTS "INDEX_email_folder_id" ON "folder_id_email_list" (
"value"
);
CREATE INDEX IF NOT EXISTS "UID_INDEX_email_folder_id" ON "folder_id_email_list" (
"uid"
);
CREATE INDEX IF NOT EXISTS "SINDEX_family_name_folder_id" ON "folder_id" (
"family_name_localized"
);
CREATE INDEX IF NOT EXISTS "INDEX_family_name_folder_id" ON "folder_id" (
"family_name"
);
CREATE INDEX IF NOT EXISTS "SINDEX_given_name_folder_id" ON "folder_id" (
"given_name_localized"
);
CREATE INDEX IF NOT EXISTS "INDEX_given_name_folder_id" ON "folder_id" (
"given_name"
);
CREATE INDEX IF NOT EXISTS "INDEX_full_name_folder_id" ON "folder_id" (
"full_name"
);
CREATE INDEX IF NOT EXISTS "INDEX_nickname_folder_id" ON "folder_id" (
"nickname"
);
CREATE INDEX IF NOT EXISTS "SINDEX_file_as_folder_id" ON "folder_id" (
"file_as_localized"
);
CREATE INDEX IF NOT EXISTS "INDEX_file_as_folder_id" ON "folder_id" (
"file_as"
);
CREATE INDEX IF NOT EXISTS "keysindex" ON "keys" (
"folder_id"
);
COMMIT;
You would then need to create a bunch of inserts that follow the structure logic like this:
INSERT INTO "folder_id_phone_list" VALUES ('pas-id-5DBFDF5A00000000','+0000000000');
INSERT INTO "folder_id_phone_list" VALUES ('pas-id-5DBFDF5A00000000','+131545678');
INSERT INTO "folder_id_phone_list" VALUES ('pas-id-5DBFE7F200000001','+45646546565465');
INSERT INTO "folder_id_email_list" VALUES ('pas-id-5DBFDF5A00000000','first_mail@gmal.com');
INSERT INTO "folder_id_email_list" VALUES ('pas-id-5DBFDF5A00000000','home_email@gmail.com');
INSERT INTO "folder_id_email_list" VALUES ('pas-id-5DBFE7F200000001','asdf@fasdf.com');
INSERT INTO "folder_id" VALUES ('pas-id-5DBFDF5A00000000','2019-11-04T08:28:52Z(15)','a header mail, this','001-)71)/1KA)9?O79M��',NULL,'john smith','this','020-O79M�','a header mail','001-)71)/1KA)9?',0,0,0,0,'BEGIN:VCARD
VERSION:3.0
UID:pas-id-5DBFDF5A00000000
X-URIS:www.testing.com
FN:John Smith
N:a header mail;This;is;;
X-EVOLUTION-FILE-AS:a header mail\, This
REV:2019-11-04T08:28:52Z(15)
TEL;TYPE=VOICE,HOME:+0000000000
TEL;TYPE=CELL:+131545678
EMAIL;TYPE=PERSONAL:first_mail@gmal.com
EMAIL;TYPE=HOME:home_email@gmail.com
ADR;TYPE=WORK:p.o. box;456;Stree work;New York;New York;456465;USA
ADR;TYPE=HOME:p.o. box;456;Street;City;New York State;13676;USA
BDAY:2019-11-04
PHOTO;VALUE=uri:file:///home/osboxes/.local/share/evolution/addressbook/sys
tem/photos/pas_id_5DBFDF5A00000000_photo-file0.image%252Fpng
END:VCARD',NULL);
INSERT INTO "folder_id" VALUES ('pas-id-5DBFE7F200000001','2019-11-04T08:57:22Z(17)','sdaf','019-M/)3',NULL,'sdaf','sdaf','019-M/)3','','000-',0,0,0,0,'BEGIN:VCARD
VERSION:3.0
EMAIL;TYPE=PERSONAL:asdf@fasdf.com
TEL;TYPE=CELL:+45646546565465
ADR;TYPE=HOME:asdf;;asdfsdfasdf;;sadf;;asdf
FN:sdaf
N:;sdaf;;;
X-EVOLUTION-FILE-AS:sdaf
UID:pas-id-5DBFE7F200000001
REV:2019-11-04T08:57:22Z(17)
END:VCARD',NULL);
INSERT INTO "folder_id" VALUES ('pas-id-5DBFE7F800000002','2019-11-04T08:57:28Z(19)','next','014-C1WO',NULL,'next','next','014-C1WO','','000-',0,0,0,0,'BEGIN:VCARD
VERSION:3.0
FN:next
N:;next;;;
X-EVOLUTION-FILE-AS:next
UID:pas-id-5DBFE7F800000002
REV:2019-11-04T08:57:28Z(19)
END:VCARD',NULL);
INSERT INTO "keys" VALUES ('eds-reserved-namespace-is-populated','1','folder_id');
INSERT INTO "keys" VALUES ('revision','2019-11-04T08:57:28Z(20)','folder_id');
INSERT INTO "folders" VALUES ('folder_id',11,'email;prefix:phone','en_US.UTF-8','US');
What is important to note is the uid which is e.g. for the first record: pas-id-5DBFDF5A00000000 (which is defined as new HashMap<string, Field?>). I see it as fixed string pas-id-, then hex number 5DBFDF5A (appears a random one) and last is a counter 00000000 (next would 00000001).
I would personally go for the first option, but second is there if you would have issues creating the xml file. If you go with the manual inserts don't forget to be consistent with the data.
Edit - comment on the data exported out
It appears so attributes like file_as_localized, given_name_localized, and family_name_localized contain some "rubbish" data. However, these attributes are defined as TEXT so there should be some kind of readable data, but it is not (tried with different encoding but it did not help). I recommend reading the source code for these attributes. There maybe a possibility to enter empty value like it appears to be done on some account at family_name_localized with the 000- only. You can also try it with NULL value at the beginning and when you know it you can update it later on.
Conclusion
After performing the tests with the importer, I think the easier option is actually the option 2. The option 1 would include some programming in vala, if you do not want to import only names that is.
As for the SQLite tool I recommend DB Browser for SQLite where you can query your SQLite database, run queries and even edit by hand if needed.
