maoyalu / zen_cc_search

Submission for Coding Challenge

Zensearch

Submission for Coding Challenge

---

Prerequisites

1. To run this CLI application, you should have Python 3.9(latest) and pip installed and configured to use.

   You can download it here:

   https://www.python.org/downloads/

2. git for downloading code and version control.

   You can download it here:

   https://git-scm.com/downloads

How to install

1. Clone and go into the root directory

git clone https://github.com/maoyalu/zen_cc_search.git

cd zen_cc_search

2. Restore project requirements

# If you have both 2.x and 3.x installed
# you might need to use pip3 instead
pip install -r requirements.txt

3. Run

# If you have both 2.x and 3.x installed
# you might need to use python3 instead
python -m zensearch

How to use

1. Configure cache settings.

   zensearch/config.py stores configs that enable reverse index. If it is set to true, additional spaces will be allocated for faster lookup, accelerating the search from O(n) to O(1).

   To make new configs effective, you need to restart the app.

2. Quit anytime.

   Type quit into the prompt and hit ENTER anytime, ane then the program will request your confirmation for exit.

   Y - Exit

   N (default) - Cancel and go back to the main page

3. Select options

   Option list will be provided above the prompt.

   Select the option number and hit ENTER. If the option number is invalid, an error message will be shown.

4. Search for the value

   Empty value is supported.

   Simply hit ENTER without typing anything.

5. Navigation

   Main
    |----- Search
    |       |----- Users
    |       |----- Tickets
    |
    |----- View Searchable Fields
   
   

Tradeoffs

1. In order to approach O(1) time as much as possible, additional spaces are allocated. Searching generally will take O(n) time so perform lookups to prevent searching if possible.

   This feature can be switched off or partially enabled depending on users' actual situation.

   By defulat, this feature is enabled.

Assumptions

1. Users are comfortable using CLI, but they are not that technical so interactions are designed in a humanly way instead of using command argument options.

2. Memory is sufficient for storing all the data.

3. No value of 'quit' exist in the data, otherwise it will contradict with the requirement 'Exit anytime'.

4. Optional fields are observed from the json files provided.

   User

   Field	Description
   _id	* required * unique * int
   name	* required * string
   created_at	* required * string
   verified	* optional * boolean * default False

   Ticket

   Field	Description
   _id	* required * unique * string
   created_at	* required * string
   type	* optional * string * default None
   subject	* required * string
   assignee_id	* optional * int * default None
   tags	* required * list of string

Scalability

Uses MVC structure to reduce coupling.

• Model - database.py

• View - cli.py

• Controller - __main__.py

• Constants - constants.py, config.py

1. No changes to CLI required for changes of searchable fields.

   You only need to update the enum in constants.py

2. Easy to switch to use a database in the future.

   You only need to update the implementation of database.py

3. Use constants to prevent hardcoded string

Future thoughts

1. Use date type instead of string for created_at

   It's ridiculous to use a date time format string to search.

   And date type allows more flexible searches like date range and provides better support for multi-timezone.

2. Cache on demand

   Another way to cache is storing them when they are searched along with an expiration time.

   It can reduce startup time and memory used in runtime comparing to the current cache strategy.

   This can be beneficial if the number of frequently searched items is very small.

3. Advanced search

   Current version only implements searching with one single field.

   Searches with a combination of fields generally take O(n) time.