Tutorial
You will need:
- Git
- Go (>= 1.1)
- A regular web browser OR curl/wget
mkdir tiedot && cd tiedot
export GOPATH=`pwd` # backticks surround pwd
go get github.com/HouzuoGuo/tiedot
./bin/tiedot -mode=httpd -dir=/tmp/MyDatabase -port=8080
tiedot HTTP server serves one database; collections in a database are automatically partitioned; the collections may use indexes to run queries.
Each new document has an automatically assigned unique ID (string) that will always stay with the document.
API requests (along with parameter values) may be sent using GET, POST or PUT HTTP method.
Create two collections:
> curl "http://localhost:8080/create?col=Feeds"
> curl "http://localhost:8080/create?col=Votes"
What collections do I have now?
> curl "http://localhost:8080/all"
["Feeds","Votes"]
Rename collection "Votes" to "Points":
> curl "http://localhost:8080/rename?old=Votes&new=Points"
Drop (delete) collection "Points":
> curl "http://localhost:8080/drop?col=Points"
Insert document:
> curl --data-ascii doc='{"a": 1, "b": 2}' "http://localhost:8080/insert?col=Feeds"
791206372389179361 # the new document's unique ID
Read document:
> curl "http://localhost:8080/get?col=Feeds&id=791206372389179361"
{"a":1,"b":2}
Update document:
> curl --data-ascii doc='{"a": 3, "b": 4}' "http://localhost:8080/update?col=Feeds&id=791206372389179361"
Delete document:
> curl "http://localhost:8080/delete?col=Feeds&id=791206372389179361"
Index is required by all lookup queries.
For an example: Put an index on a,b,c will help queries finding documents such as {"a": {"b": {"c": 1}}}.
Create some indexes:
> curl "http://localhost:8080/index?col=Feeds&path=a,b,c"
> curl "http://localhost:8080/index?col=Feeds&path=Title"
> curl "http://localhost:8080/index?col=Feeds&path=Source"
> curl "http://localhost:8080/index?col=Feeds&path=Age"
What indexes do I have now?
> curl "http://localhost:8080/indexes?col=Feeds"
["a","b","c"],["Title"],["Source"],["Age"]]
Remove an index:
> curl "http://localhost:8080/unindex?col=Feeds&path=a,b,c"
Prepare some documents:
> curl --data-ascii doc='{"Title": "New Go release", "Source": "golang.org", "Age": 3}' "http://localhost:8080/insert?col=Feeds"
> curl --data-ascii doc='{"Title": "Kitkat is here", "Source": "android.com", "Age": 2}' "http://localhost:8080/insert?col=Feeds"
> curl --data-ascii doc='{"Title": "Slackware Beta", "Source": "slackware.com", "Age": 1}' "http://localhost:8080/insert?col=Feeds"
Looking for the article "New Go Release":
> curl --data-ascii q='{"eq": "New Go release", "in": ["Title"]}' "http://localhost:8080/query?col=Feeds"
{"356740846970476516":{"Age":3,"Source":"golang.org","Title":"New Go release"}}
Looking for article "New Go release" and "android.com" feeds:
> curl --data-ascii q='[{"eq": "New Go release", "in": ["Title"]}, {"eq": "android.com", "in": ["Source"]}]' "http://localhost:8080/query?col=Feeds"
{"356740846970476516":{"Age":3,"Source":"golang.org","Title":"New Go release"},"8835531386221862775":{"Age":2,"Source":"android.com","Title":"Kitkat is here"}}
Looking for all but not feeds from "golang.org":
> curl --data-ascii q='{"c": [{"eq": "golang.org", "in": ["Source"]}, "all"]}' "http://localhost:8080/query?col=Feeds"
{"4530407170288349686":{"Age":1,"Source":"slackware.com","Title":"Slackware Beta"},"8835531386221862775":{"Age":2,"Source":"android.com","Title":"Kitkat is here"}}
Note that: "all" means "all documents"; {"c": [ .. ]} means "complement".
Looking for young feeds (age between 1 and 3) from "slackware.com":
> curl --data-ascii q='{"n": [{"eq": "slackware.com", "in": ["Source"]}, {"int-from": 1, "int-to": 3, "in": ["Age"]}]}' "http://localhost:8080/query?col=Feeds"
[{"Age":1,"Source":"slackware.com","Title":"Slackware Beta","@id":"12333034197694914883"}]
{"n": [ .. ]} means "intersect"; {"int-from": x, "int-to": y, "in": [ .. ] } is an integer range lookup.
There is also /count?col=foo API to return number of results.
Let's gracefully shutdown server:
> curl "http://localhost:8080/shutdown"